Timeline

Minutes         Progress Reports

Minutes

September
S	M	T	W	Th	F	Sa
			1	2	3	4
5	6	7	8	9	10	11
12	13	14	15	16	17	18
19	20	21	22	23	24	25
26	27	28	29	30

October
S	M	T	W	Th	F	Sa
					1	2
3	4	5	6	7	8	9
10	11	12	13	14	15	16
17	18	19	20	21	22	23
24	25	26	27	28	29	30
31

November
S	M	T	W	Th	F	Sa
	1	2	3	4	5	6
7	8	9	10	11	12	13
14	15	16	17	18	19	20
21	22	23	24	25	26	27
28	29	30

December
S	M	T	W	Th	F	Sa
			1	2	3	4
5	6	7	8	9	10	11
12	13	14	15	16	17	18
19	20	21	22	23	24	25
26	27	28	29	30	31

January
S	M	T	W	Th	F	Sa
						1
2	3	4	5	6	7	8
9	10	11	12	13	14	15
16	17	18	19	20	21	22
23	24	25	26	27	28	29
30	31

February
S	M	T	W	Th	F	Sa
		1	2	3	4	5
6	7	8	9	10	11	12
13	14	15	16	17	18	19
20	21	22	23	24	25	26
27	28

March
S	M	T	W	Th	F	Sa
		1	2	3	4	5
6	7	8	9	10	11	12
13	14	15	16	17	18	19
20	21	22	23	24	25	26
27	28	29	30	31

April
S	M	T	W	Th	F	Sa
					1	2
3	4	5	6	7	8	9
10	11	12	13	14	15	16
17	18	19	20	21	22	23
24	25	26	27	28	29	30

May
S	M	T	W	Th	F	Sa
1	2	3	4	5	6	7
8	9	10	11	12	13	14
15	16	17	18	19	20	21
22	23	24	25	26	27	28
29	30	31

 

 

 

 

              

September

9/17 EGR 410D meeting Minutes September 17, 2004 Friday 2:30pm – 4:00pm
Team meeting with liaisons, Location: Prof. Doughty's Office
Facilitator: Jie/ Recorder: Xiao-ning

• Elect Project Leader: Jie Zheng
• Plan tentative agenda for next meeting and meeting with liaison
• Traded contact info
• Visit SAL, register for SAL accounts
• Project idea was developed after a Smith student approached Jon Caris, need to contact her
• Come up with team name
• Read 2 GPS books
• Decide meeting time and location with liaisons: Thursdays at 4pm, make conference call

 

top of page

9/20 EGR 410D meeting Minutes September 20, 2004 Monday 2:30pm - 3pm
Team meeting with Professor Howe, Location: EGR 102
Facilitator: Jie/ Recorder: Xiao-ning

• Administrative businesses
  • 2 lockers available per team (EGR building under the 1st floor stairs)
  • register for assess to the space lab
  • complete readings
• Communication with the liaisons
  • Thursday afternoon at 4 in the Conference room in Bass
  • need to setup/bring the phone
• Due Dates
  • Progress Reports, Monday morning at 9 am
  • Proposal, including deliverables and presentation, Oct 8, 2004 Friday.
  • product/design show in March
• Budgets
  • ~ $ 2,000, additional funding available for apply w/ good justification
  • travel budgets ~< $1,000
  • look for used or donated equipments (e.g. Ebay)
  • funded by NCIIA Grants (the National Collegiate Inventors and Innovators Alliance)
  • fund is part of the Lemelson Foundation
    E-team Conference during spring break in San Diego, California.
• Consulting, 10 hrs w/ faculty in the EGR Department
• To be completed
  • Team meeting tomorrow (September 21, 2004), 6pm at Lamont dinning hall.
  • Set team name
  • Set future meeting schedules and locations
  • Set rotation schedule for facilitator and recorder

top of page

9/21 EGR 410D meeting Minutes September 21, 2004 Tuesday 6pm – 8pm
Team Reach Meeting, Location: Lamont
Facilitator: Jie/ Recorder: Xiao-ning

• Administrative Businesses
  • Team Name: TEAM REACH
  • Team Charter (see separate sheet)
  • Facilitator and Recorder Rotation
    • weekly, by the order of first name (Jessie, Jie, Monica, Xiao-ning)
  • Team Meeting Times
    • Monday, 2:30pm, EGR 202, w/ Professor Howe
    • Tuesday, 6 pm, at Lamont Dinning Hall, team
    • Thursday, 4 pm, at Bass Conference Room, w/ Liaisons
    • Thursday, after meeting w/ Liaisons, at Tyler, team (when necessary)
    • Sunday, 12:30 pm, brunch at Hubburd, team
    • Sunday, 12:30 pm, brunch at Hubburd, team
    • Class meeting times: 09/24F(design process), 10/01F(project management), 10/15F(proposal), 10/18M(open), 10/22F(open), 11/01(WP1), 11/15M(open), 12/03F(open), 12/06F(open)
• Technical Businesses
  • Setup team Website, group Sophia account, need to contact ITS
  • Test conference phone call functions in the conference room (Jie)
• Communications
  • Setup GPS training time w/ Jon Caris
  • Inquiry about the visual-impaired student on campus w/ Jon Caris
  • Inquiry about available hardware/inventory on campus w/ Jon Caris
  • Liaison phone call Thursday at 4 pm, Bass Conference
  • Contact Office of Disability Services
  • Leader, not facilitator, leads conference call w/ Andrew Faag
• General Preparations:
  • Read Fagg’s articles (linked from his email yesterday)
  • Conduct relevant researchs (market/budget; Database, technology assessable, voice activated software, PDAs, Hardware, wireless devices)
  • Look into available product on the market (e.g. VisuAids, Pocket PC to blind)
• Proposal / Progress Report
  • Prepare to start on Background and Introduction on Proposal
  • Decide/Divide Progress Report
  • Have reference books by Sunday meeting ( Sept. 25, 2004 Sun)
  • Recorder of the week writes the progress report for that week
• To be completed
  • Next meeting, Thursday, 4 pm (09/23), at Bass Conference Room, w/ Liaisons
  • Discuss plans for the proposal
  • Write Progress Report

top of page

9/23 EGR 410D meeting Minutes September 23, 2004 Thursday 4:00pm – 5:30pm
Team meeting with liaisons, Locations: Bass Lawn, Bass SAL
Facilitator: Jie/ Recorder: Jie

• Problems with conference phone, Team Leader on the phone with Professor Fagg
• Wearable computing design versus PDA handheld design?
• Consider interaction with the device and carrying it
• Wearable computing units have lots of connections: PCMCIA, USB, RS232, Keyboard… (not necessary in our design)
• PDAs don’t have large hard drives
• PDAs are easier to carry around, can have wifi headphones and mic
• What hardware is available to us?
• 2 Xybernaut MA4s, both are UMass property, can arrange for us to go see the equipment at UMass (8in x 4in x 3in)
• Do GIS databases need to be stored in hard drive or can it be accessed through the internet (Smith has 32.5MB Microsoft Access Geodatabase)?
• Cannot pack everything into mobile device, so make it an interface with connections. Voice-activated software should be stored in hard drive because of bandwidth limitations
• Suggestion: Think about complete system now, what kind of GIS receiver chip, wireless internet connection… with intent of focusing on individual components
• Think about whether we want a PDA that is compatible with many languages, software that can be edited
• Prof. Fagg will open up access to the UMass website with research materials
• Shared budget information and intent on approaching Office of Disability Services
• Meeting with Jon Caris
• Summary of meeting with Prof. Fagg
• Next Thursday after phone meeting, tutorial on GIS at 4pm in SAL
• Ask Prof. Howe tomorrow about conference phone
• Website up and running
• Smith hardware
• Trimble GIS receivers used for mapping, has 14 channels
• Team will borrow Trimble receiver soon to test coverage on campus
• DGPS at Smith
• Error of 3 meters or less
• Northampton DPW broadcasts signal via IP to cell phone
• WAAS increases accuracy
• To get sub-meter error, must work on base station calculating error and sending signal to receivers
• In-door, outside coverage, stick to outside
• Voice-activated software: Voice Insight
• Look into wireless connections
• 802.11b and hotspots on campus Voice-activated software: Voice Insight
• Focus on chips (hardware), figure out what we need, and connectors (i.e. USB)
• He will give us Tasha’s contact info, meanwhile, we will email Office of Disability Services to set up meeting
• Patent search, write patent up to emphasize that the product is unique to Smith College campus
• To be completed
• Talk to Prof. Howe about phone conference situation
• Get Tasha’s contact info, contact her, set up meeting time
• Email Office of Disability Services to set up meeting
• Decide time when team can use Trimble receiver to test campus coverage
• Jessie will watch Ebay for potential PDAs/GIS receivers that we can take apart
• Jie will write up minutes for this meeting
• Individual research using articles, websites, books
• Talk about Project Proposal next meeting and divide parts
• Decide time to watch GPS movie together
• Contact UMass about seeing their Xybernaut MA4
• patent searches

top of page

9/24 EGR 410D meeting Minutes September 24, 2004 Friday 2:30pm – 4:00pm
Team meeting, Location: EGR 201
Facilitator: Jie/ Recorder: Xiao-ning

• Design Process
  • Recognize the need – contact Tasha, Office of Disability Services
  • Define the problem
  • Plan the project – work on Proposal next meeting on Sunday
    • Problem Description – Everyone writes together, campus-wide range
    • Budget
    • Divide work
• To be completed
  • Next Thursday, Sept. 30th, meet in Tyler for conference call, Monica will bring phone
  • Go to ITS together to get webspace and email accout

top of page

9/26 EGR 410D meeting Minute September 26, 2004 Sunday 12:30pm - 4pm
Team meeting, Location: Hubbard Dining Hall and Wright Hall Computer Lab
Facilitator: Jessie/ Recorder: Jie

• Brunch/interview with Tasha, visually-impaired Smith student who originated the idea for the project
• Her suggestions were to chose between a receiver or receiver connected to laptop
• Chose between three software options:
• Jaws - Accessible, voice-activated, text-to-speech software for Windows
• Self-voicing application which uses recorded .wav files (i.e. a recording asks at the beginning “Menu, file, edit…”)
• Software coming out in January for PDAs with a text-to-speech engine that allows one to create a voice
• Consider the pocket PC and NOT the palm pilot
• Tasha uses a laptop with Kurzweil 2000, software that scans documents and transforms them into audio
• We asked Tasha what the product we are trying to develop should do
• She uses only one consistent route to get from one location to the next, which has the fewest hazards even if it is longer in distance. Problems arise when she gets off a route and orientate herself back onto the route (asking people does not help)
• GPS can help by allowing her to “record” certain routes and when while she is on a certain route, guide her by letting her know what buildings, streets or stable, inanimate objects are around and what she is approaching
• Questions and answer session next (paraphrased)
• Getting off-route—is this a daily or weekly problem?
• Daily. Problems arise when I leave from the wrong door to go somewhere or if a truck is parked somewhere in my route.
• Do you have access to a non-visual map of the Smith campus?
• No, I have only had access to maps for specific paths. I would like to be able to take a virtual tour of the campus using arrows on my laptop.
• Do you usually explore campus alone or with a companion?
• Depends. For unfamiliar places, I try to find a friend to go with me. I don’t explore much on my own. This device you will develop will help give me directionality and confidence to explore.
• For our handheld device, is a keyboard necessary? Regular or Braille?
• For a Pocket PC, a keyboard is necessary. Regular is fine. That’s the kind I use on my laptop.
• What about the Visu-aide handheld PDA for the visually-impaired?
• It is closed-architecture, so cannot add software. It has its own GPS software, Trekker. It is better for your project to use the most off-the-shelf software and hardware as possible.
• What would you prefer in terms of the product design?
• I carry my laptop everyday so it would be nice to have something pocket-sized
• Extra recommendations by student:
• Contact Freedom Scientific, who makes Jaws and a competing GPS software on proprietary PDA
• Visu-Aide
• She gave us her contact information
  
• Next, team met in Wright Hall
• Discuss developing a campus map for student before prototype for product is developed, would need annotation layer of GIS
• Began writing Project Proposal
• Setup the full format
• Wrote introduction and motivation
• After browsing through deliverables for other Design Clinic projects, developed list of deliverables and Gantt Chart
• Divided sections for background and research areas (* refers to main person in charge of research and writing the background)
• PDAs – Jie*
• GPS, DGPS – Jessie*, Monica
• GIS and databases – Jessie*, Xiao-ning
• Patents and market research – Xiao-ning*
• Available hardware and compatibility – Jie*, Monica, Xiao0-ning
• Voice-activated, voice-enabled software – Monica *
  
• Tasks to be completed:
• Sign charters tomorrow during meeting with Prof. Howe
• Individual research, think about budget
• Finish individual background sections by Thursday
• Decide to add to Charter: minutes for Sunday must be emailed to recorder who writes Progress Report by 7pm
• Meet at 5:45pm in Lamont on Tuesday. After dinner, work latest to 9:30pm on Proposal (budget, personnel, tasks…)
  

top of page

9/27 EGR 410D meeting Minutes September 27, 2004 Monday 2:30pm - 3pm
Team meeting with Professor Howe, Location: EGR 201
Facilitator: Jessie/ Recorder: Jie

• Suggestions from Prof. Howe regarding Progress Reports
• List meetings and corresponding locations
• Include more detail in the discussion section (distinguish between previously discussed and what will be discussed)
• List who does what, not just the task
• Add supplement in next Progress Report to clarify parts of this Progress Report
• Suggestion from Prof. Howe regarding facilitator role
• Facilitator prints out agenda for everyone or puts it up on the board ahead of time
• Leave space in agenda to take notes
• Send any documents that we want Prof. Howe to look at prior to meeting
• Discussion
• We determined our deliverables after consulting the other project summaries
• Paper design by Winter Break
• Work on prototype construction in Spring
• Revisit cost/benefit analysis in Spring (design trade-offs)
• We prepared Gantt Charts prior to the meeting but need to break down each task further into layers of sub-tasks using knowledge learned in class today
• Have a detailed Gantt Chart for this semester and outline a sketch for the next
• Add working presentations onto Gantt Chart, can use the first date listed (50 minutes each, no PowerPoint, interactive, discussion-based)
• Will also have visual and constraints presentation
• Upcoming 20-minute proposal presentation (can be PowerPoint)
• How will progress be monitored?
• Team will monitor it
• Review the Gantt Chart once a week
• Review list of tasks once a month
• Ask liaisons where we should be
• General suggestions from Prof. Howe
• Talk with constituents and consider design factors
• Get a group of testers and blindfold them
• Talk to more than one visually-impaired person
• Email Noho residents if there are not enough Smith students in this category
• Product should be design-informed, tested by end-users
• Tasks to be completed
• List meetings and corresponding locations
• Include more detail in the discussion section (distinguish between previously discussed and what will be discussed)
• List who does what, not just the task
• Add supplement in next progress report to clarify parts of this Progress Report
• Prepare questions to ask Office of Disability Services

 

9/28 EGR 410D meeting Minutes September 28, 2004 Tuesday 6:30pm – 7:45pm
Team meeting, Location: Lamont House
Facilitator: Jessie/ Recorder: Jie

• Administrative tasks
• Amendment to Charter: Recorder and Facilitator roles transition on Mondays
• Amendment to Charter: Facilitator should decide length of meeting in agenda for every meeting (can go up until the maximum length of time for meetings as stated in Charter)
• Worked on Proposal as a team effort
• Personnel
• Budget
• Started writing the Project Description
• Tasks to be completed
• Since Office of Disability Services replied to our team email, Xiao-ning will call them tomorrow and discuss a meeting time that definitely works for her and hopefully one other team member
• Software for visually impaired students?
• How are they currently helping visually impaired students on campus?
• Gather statistics on visually impaired students at Smith
• Not enough visually impaired Smith students probably, so ask them to help us get in touch with Noho residents who are visually impaired (we need ask them questions to help with our design specifications and also to test at the end)
• Testing the GPS receiver signal coverage on campus
• Jessie and Monica
• Plan to meet this Saturday, 10/2, at 2:30pm if Jon permits equipment usage during the weekend
• Assessing user needs and usability factors by blindfolding subjects and having them walk some chosen routes with commands as envisioned for prototype design
• Jie and Xiao-ning
• Will meet this Saturday, 10/2, at 10am in the CampusCenter to conduct tests on each other to improve design interface
• Team will work on Proposal in class on Friday (Xiao-ning will bring Technical Writing handbook to class)
• Revise Gantt Chart, Budget
• Continue writing Project Description
• Remember to attach Team Charter
• Team members will revise individual research background sections deadline to be Sunday, 10/3, not Thursday
• Jie will amend Team Charter

top of page

9/30 EGR 410D meeting Minutes September 28, 2004 Thursday 4:00pm – 6:30pm
Team meeting, Location: Tyler House/SAL
Facilitator: Jie/ Recorder: Monica

Telephonic conference with Prof. Andrew Fagg

• He will look over our list of deliverables and Gantt Chart and will comment on them by Wednesday evening
• He will look on the draft of the Project Proposal and let us know if we went far enough in our background research of different areas
• Prof Fagg will come in the area for a couple of times this year; dates are still unknown
• additional topics to be covered on Friday, October 1 st (Office of Disabilities meeting):
  • what they see as needs and how they would see the system performing
  • if they would be willing to help with the evaluation of the system, given the experience (which?) they have with visually impaired students
  • should we use # of steps or feet to communicate distance to the use
• Saturday testing of the availability of the GPS signal throughout the College Campus – Jessie and Monica
  • look into inertial systems if we have problems with the differential GPS system signal
  • Jon Caris – 2 types of GPS signals available – on campus (WAAS) and the system in Northampton (signals from our base transmit it onto cell phones); possibility: broadcast the signal through wireless net and check the hotspots on campus
• Saturday blindfold experiencing – Jie and Xiao-ning
  • suggestions for protocol: sketch the different dialogues that can take place between system and user. use this as a script
  • have a list of main menus and submenus
• don’t underestimate the time needed to integrate the hardware and make it compatible
• allocate some time during the first semester to test the separate components of our system (GPS device, voice interface) and consider the ways the users would communicate with them
  • user should have the ability to choose their path between two places
• signal when they are off track – how often does the GPS receive info (1/s?, GPS with compass)
• detect and announce landmarks’ location – for proximity alert the database has to be enriched – should also include garbage bins, fences, high curve cuts (post processing step possible)
• start by thinking about the big picture and be ambitious about the project, and later focus on having an interesting presentation
• take care of the routing at an early stage in the design process

To do list

• post Gantt Chart and description of tasks online by Monday morning as part of our Progress Report
• post Progress Report draft online on Sunday
• explore AI, Graph Theory, and GPS possibilities to scale the characterize the paths (by length, difficulty degree, etc)
• find a room where we can have a phone conference!!!

To be contacted

• Judy Franklin, Assistant Professor, CS Department, McConnell 211, ext# 3858
• Darryl Hampton, UMass CS Department, Vision Lab; he will recommend other people we can engage in the design phase (could only find contact for a so called Jeffrey D Hampton, Departmental Assistant!!)

GIS Tutorial Session given by Jon Caris

• Learn about ARCGIS using the free tutorial on ESRI Virtual Campus
• Create a user account first
• Open Arc Catalog – Explorer for GIS data
• Go to -> Q:/ drive -> shared -> campus -> ortho photos
• Open geodatabase (March 2003)
• Microsoft Access database
• Jon Caris will create multi-user
• Layers
• Buildings
• Hydrants
• Landmarks
• Manholes
• General pavement (road edges)
• Emergency phones
• River_poly (pond)
• Tennistrack
• Shape folder
• Has original shapes
• Find “paths2,” which represents all sidewalks
• Open 3D Analyst for 3D view of campus
• Open buildings
• Go to extrusion and set average height in meters
• To Do List
• Making CAD, GIS compatible (Physical Plant has all CAD files)
• Sheryl Obremksi at Physical Plant
• To convert, bring CAD and convert to GIS (add annotation, give intelligence to file)
• Model Impedances
• GIS
• Or using Graph Theory mentioned by Prof. Fagg
• Model GPS Coverage
• Above 15° is default threshold to meet coverage criteria
• modeling visibility analysis (Spatial Analyst 3D Analysis)
• UCSB Libraries and The Campus Community is developing wearable computing system
• Geo Department
• Keith Clarke – can email him
  

Dinner Meeting’s To Do-List

• Conference Room situation must be resolved. Jie will go to ITS tomorrow morning.
• Need to organize a hardcopy of all group documents and keep in Group Binder
• Xiao-ning and Jie’s experiment on Saturday
• Designed a menu/submenu for experiment
• Decided on testing 2 routes: Sage Hall to Art Library and Hubbard House to Campus Center.

top of page

 

October

10/01 EGR 410D meeting Minutes October 1, 2004 Friday 2:30pm – 4:30pm
Team meeting, Location: EGR and Office of Disability Services
Facilitator: Jessie/ Recorder: Jie

• Decided to develop Gantt Chart using a daily schedule
• Developed preliminary Gantt Chart for 1.5 hrs, but was lost due to computer error
• Team went to the Office of Disability Services to meet with Laura as was scheduled. Here are the questions and answers paraphrased:
• What kinds of services is the Office of Disability Services currently providing for Tasha?
• Whatever Tasha needs, we try to provide
• We help her get books for research, books will be scanned into computer
• Tasha uses
• Books are translated to Braille through an electronic device (Laura could not remember the name)
• Van rides around campus
• Mobility is the big issue
• Tasha has a mobility trainer, not a Smith employee. We could try to get in touch?
• She has a cell phone for emergencies
• What are the statistics on visually impaired students who have attended Smith? Our device, when fully developed, can attract more visually impaired students.
• (Laura answers based upon her knowledge from the 6 years she has worked at Smith)
• Besides Tasha, only one other was blind (graduate student)
• Some students are visually impaired and must use special magnifying glasses
• What legal documents provide guidelines for the development of Smith College so that it is accessible to the disabled?
• Section 504 of the Rehabilitation Act in 1973
• Americans of Disabilities Act in 1990 (Title 3)
• Make new buildings accessible
• Make programs accessible
• Do what’s already achievable (don’t have to redo old buildings)
• Requirements a combo of laws and state building codes
• Voice-activated Software
• Naturally Speaking
• There’s an adaptive computer lab in the library with this software.
• Contact Lisa Decaralis (x3099) at ITS for a demo of software
• Dragon Dictate
• Where can we find visually impaired people who might want to help out with our system design and also testing of the prototype?
• Stavros ( Amherst ) – Program run by people with disabilities
• School system in Northampton , High schools
• 5 colleges (UMass, Holyoke Community College )
• Should we use the number of steps or distance in standard units?
• Steps
• Also consider mobility training (would need markers)
• Various Contacts
• Mass. Assistive Technology Project
• AAAS (Rehabilitation Engineering)
• National Institute for Disability and Rehabilitation Research offer grants
• Hampshire Lemelson Center for Design
• Tasks to be completed
• Contact appropriate places as necessary
• Redo Gantt Chart
• Interview Tasha again
• Share interactive menu and ask for feedback
• Ask steps or distance?
• Find out the specific routes she takes and most commonly visited buildings
• Can she feel the difference between grass and pavement?
• Monica makes an appointment to visit adaptive computer lab
• Contact Tasha’s mobility trainer

top of page

10/03 EGR 410D meeting Minutes October 3, 2004 Sunday 12:30pm – 6:00pm
Team meeting, Location: Hubbard and EGR
Facilitator: Jessie/ Recorder: Jie

• Completed Tasks
• Jessie and Monica shared experiences with Jie and Xiaoning and vice versa about testing on Saturday
• Completed the first draft of the Proposal, except for the Abstract and References
  • Compiled separate background sections
  • Editted the entire draft
• Emailed Proposal to Prof. Howe and liaisons for suggestions
• Tasks to be completed
  • Add Abstract and References sections
  • Edit as necessary according to received suggestions
  • Prepare Proposal Presentation for Friday
    • Make a PowerPoint presentation
    • Rehearse individually and as a group

top of page

10/04 EGR 410D meeting Minutes October 4, 2004. Monday 2:40-3:20
Team meeting With Prof. Howe, Location: EGR Building
Facilitator: Jie/ Recorder: Monica

Discussion:

• Prof. Howe will find a room for conference; till then, some alternatives are:
  • Talk and get a room from Media Services
  • CDO room
  • Carla Cooke’s office
  • Prof. Grasso’s office
• Website – ITS will solve the problem
• Printing – keep track of it
• use Xerox in the EGR lounge
• printing cards for the course were requested by Michele
• will be reimbursed for the printing
  
• Progress reports
  • Send the reader to the discussion section after “we met x”
  • Tasks should be tighter and easier to check off
  • Remind at the beginning where we are currently on the task list
    
• Proposal
  • Write it as if to an investor
  • Motivation and background section
  • Place numbers before subheadings and separate the “Motivation and Background” into two sections (1. Motivation, 2. Background)
  • Include the importance of it being an entrepreneurship project & about Jessie
  • Include the possibility to expand this project to a larger geographical area in the motivation section
  • Section on hardware has to come first = it’s the main one
  • The section on patents has to be separate at the end
  • Project Description
  • Change the order of given information – separate the project description (what we want to create) from the process we follow (decisions we have already taken, why we choose the hardware we choose)
  • The second paragraph (on monitoring) should be a separate section, not part of the project description
  • Third paragraph (on challenges) should be a separate section
  • Include a more detailed project description using the bottom level of the Gantt chart
  • Paragraph on design process should be separated from “Project description and scope”
  • Need to make a more detailed estimate on the budget (include travel, printing, meals)
  • Task List
  • Transform the task list into a Gantt chart with which it is easier to work
  • Change in text “development” with “planning” to be consistent with the task list (planning includes then ordering and figuring out components)
  • Go into more detail (1-2 levels more) about construction, troubleshooting, and other tasks for the spring semester (no dates needed, but all should be 1-2 weeks tasks)
  • No weighted objective tree needed; should just say what we chose and why
  • All our write-ups of activities related to the project can be included as appendices to the project proposal and report as supporting evidence for the process (include a description of what we actually did)
  • Lab/Space for deconstructing
  • Send Prof. Howe an e-mail to register for a key to the EGR lab
  • Check if we can find space in the machine shop
  • Keep record of the time spent when consulting professors & write about the professors consulted outside the EGR department
  • Next meeting with Prof. Howe: in class on Friday, October 15 th

Tasks:

• Find a conference room
• Rearrange and complete the progress report
• Work on the progress presentation
• Send Prof. Howe e-mail for EGR lab key

top of page

10/05 EGR 410D meeting Minutes October 05, 2004 Tuesday 6:30pm – 9:00pm
Team meeting Location: Lamont House & SAL
Facilitator: Jie/ Recorder: Monica

• Completed Tasks
  • The sections in the proposal were reorganized according to Prof. Howe’s comments
  • The sections were assigned for editing and improving as follows:
    • Jessie: Intro, Motivation, References, Challenges and Restrictions
    • Jie: Project Description, Design Process - Product Development
    • Monica: Scope, Design Process - Research
    • Xiaoning: Budget, Gantt Chart, Design Process - Communication
  • Each team member will edit their own part of the Technical Background
  • Jie will arrange with Media Services for us to have a conference room for Thursday
• Discussion
  • Regular meeting times from now on Tuesdays are starting 6:30 instead of 6:00 (after dinner) in SAL
• Tasks to be completed
• PowerPoint presentation for project proposal – Jessie will have the slides prepared for Thursday at 10:00pm
• Editing of each separate sections
• Everybody will have edited their sections by Thursday 6:30 pm
• Jessie and Xiaoning will meet Thursday at 6:30pm, read through the entire document and edit it
• Jie and Monica will meet at 8:00 and read a second time through the entire document
• Team will meet at 09:30pm in SAL and go through the PowerPoint presentation

top of page

10/07 EGR 410D meeting Minutes October 07, 2004 Thursday 4:00pm – 5:00pm
Team meeting with liaisons Location: Media Services
Facilitator: Jie/ Recorder: Monica

• Completed Tasks
• Timeline updated for the Spring semester
• Integrated in the project proposal part of the comments that Prof. Fagg and Jon had sent to us through e-mail
• Discussion
• We have described to Prof. Andrew Fagg what changes have been done to the project proposal since the version posted online, and what comments Prof. Howe had about it
• Narrated in detail the “Wizard of Oz ” experiment carried by Jie and Xiaoning and their conclusions
• Need to account for the temporary hazards that wouldn’t be in the permanent GIS database – can be done by the use of a guide dog or cane, except for the cases in which they drastically change the path that has to be taken
• Guiding landmarks need to be included in the database
• Developed the menu interaction
• Measure distance in number of steps and direction through three angles: straight ahead, 45 degrees left or right, 90 degrees left or right
• Prof. Fagg will talk with his visually impaired friends, one of which works in developing technology for the blind, and will provide us with contact information
• Told Prof. Fagg about the GPS signal check around campus that Jessie and Monica did
• External antennae, worn on top of the head for the handheld device would most probably increase the received signal significantly (there are flat antennae of approx 2x2x0.5 inches)
• Problems encountered with receiving a signal at the entrance and along some building might be due to the signal being distorted and lost though the multiple reflections from the surrounding buildings – consider using inertial GPS
• Example of wearable computer with the antennae onto the shoulder – what is important is that the body wearing the device does not stay between the signal and the antennae
• Tasks to be completed
• Get in touch with media services to arrange the form of payment for our conference
• Arrange for a video conference for next Thursday ( linux client, 4323)
• Use group e-mail after Fall Break
• All members of the team should get personal sophia accounts; procedure: put the request in through the User Support Center (x4487) by Friday, October 15 th
• Include all of Prof. Fagg’s comments
• Meet with Tasha and share with her the findings of the “Wizard of Oz” experiment and get her comments (Jie and Xiaoning are not skilled in using the cane, and might have encountered other types of problems that a visually impaired person would)
• Test the difference in signal receiving for different positions and heights of holding the Trimble receiver (antennae); account for the frequencies used; also test the signal when using an extra antennae

10/07 EGR 410D meeting Minutes October 07, 2004 Thursday 10:00pm – 12:30am
Team meeting, Location: SAL
Facilitator: Jie/ Recorder: Monica

• Completed Tasks
  • All team members put together the edited proposal
  • All put together the PowerPoint slides and rehearsed the presentation

top of page

10/12 EGR 410D meeting Minutes October 12, 2004 Tuesday 08:00pm – 08:30pm
Team meeting, Location: Campus Student Center Cafe
Facilitator: Jie/ Recorder: Monica

• Completed Tasks
  • Jie showed the other members how to use the Proxy option for the team e-mail account; each member has their own folder as part of the account
• Discussion
  • Each team member will be posting write-ups of their own research online, so that the other members can easily refer to it
  • From now on, our meetings will rely less on an agenda prepared by the facilitator; we will go once around the table, and each member will describe the progress being done in their research
  • Thursday, October 14 th meeting:
    • As Jie has still not gotten any answer from Media Services, we will postpone our video conference with Prof. Fagg for a future date, and hold a phone conference as usual on Thursday, October 14 th (in the Engineering Building)
    • Each team member will describe the state that they have reached in their research
• Tasks to be completed until Thursday, October 14 th:
  • Announce Jon Caris about the place of this Thursday’s phone conference
  • Divide the contacts list and assign contacts to each team member
• Other tasks:
  • Jessie will work on adapting the Smith map to our project, task which in out timeline is listed for the time interval 10/2 – 10/16; in case she needs help, she will contact other members of the team
  • Update the timeline – all team members think about ways to update the timeline, especially the sections assigned to them (ex: inertial GPS, tasks related to headphone and microphone)
  • At our Sunday meeting, we will review our state when compared to the Gantt Chart (and timeline)

top of page

10/14 EGR 410D meeting Minutes October 14, 2004 Thursday 4:00pm – 5:00pm
Team meeting with liaisons, Location: Engineering Building, Room 102
Facilitator: Jie/ Recorder: Monica

• Discussion
  • Voice interface/recognition
    • Speech API (Application Program Interface, with Windows as operating system) has all parts integrated, so we would only have to add the new words and maybe enrich the list of grammatical structures to be recognized (“find…”)
    • Xybernaut MA5 (with DSP processor) is compatible with voice recognition software
    • There are noise cancellation microphones
    • The voice recognition software and audio information will be one of the determining factors in choosing the computer platform (performance and speed criteria)
    • Using a simple cell phone like voice input would make the system user dependent, and one would have to retrain the system for each new speaker
    • Voice recognition software can use a different grammar at each stage of the interaction, sometimes more limited, according to the situation; it is thus less CPU demanding and works faster
  • Do we want to build every part of the device from scratch
    • Using already available hardware components will lead to a not-so-elegant/small size/reduced power consume device, but it might be the only way to reach a reasonable prototype at the end of this school year and have a demo
    • Sizing it down would be left as a future improvement
    • We could already use a laptop to test the functioning of the GPS system, interacting with the voice interface, etc
  • How will the software be distributed between a base station and the user’s PDA
    • The navigation database and the software determining the different routes could be installed on a desktop machine that would act as server and send already processed information to the PDA through wireless network
    • Such a server would have a better storage capacity and would work faster
    • The database would be in this case much more easy to keep updated, and the user wouldn’t need to periodically make sure it’s up to date
    • However, there are currently only few hot spots for the wireless network on campus; we are still waiting for a reply from ITS regarding an official map of these (apparently, it would include: the Campus Center, Neilson, Science and Arts Libraries, Stoddard Hall)
    • We could limit our project to the available spots, and use our final device as an argument for the need of further expansion of the wireless network
    • We should also look into using the cell phone wireless network (another frequency) (Verizon, 1.40 KB/s; data exchange with computer; cell phones with digital standards)
  • We should consider more closely the specific computation capabilities that our system needs to have because of using a voice interface and GIS database
    • Lay out the different pieces in the design, and specify what and how they would communicate
    • GIS can be carried around on a laptop, but not on a PDA (in Access the campus database has 30MB); which could still be portable (if we cannot use the server alternative)
  • We will arrange for a video conference when we have something to show
• Tasks to be completed
  • We will be using our team reach e-mail account from now on
  • On Sunday we will update the timeline, which will be reviewed weekly
  • Add to the proposal ideas and tasks that we had left out because of time constraints
  • Jessie will bring in class tomorrow a laptop working in Windows 95, and we will try to use it as a first tester
  • Rewrite the tasks on the timeline to accommodate the more coarse approach to specific hardware included in our project
  • Discuss a possible reassignment of team members to different areas (more on software; keep mic/phone for hardware)
  • Look for the network analysis program already existent on campus (related to the distributed software PDA-base)
  • Resend an e-mail to ITS: official map of hot spots on campus; plans for expansion; explain to them why we need it

top of page

EGR 410D meeting Minute October 15, 2004 Friday 2:30-4:00 pm
Team Reach Class meeting and meeting with Professor Howe Location: EGR 203

Facilitator: Monica        Recorder: Xiaoning

Updates on Available Resources

• Jessie’s Laptop
  • 2 PCMCIA slots
  • 1 printer port, 1 serial port
  • 16 MB RAM
  • Pentium Processor, 0.87 GB hard drive
  • 12.1” LCD Screen
  • Broken battery, working AC/DC input
  • Floppy Drive
  • Windows 95
• Laptop from ITS
  • 1 PCMCIA slot (in use by Ethernet Port), 1 USB 1.1 port
  • 1 Ethernet Port
  • 1 printer Port, 1 serial port
  • 3 modem port (regular phone line)
  • 1 microphone input
  • 12.1” LCD screen
  • Functional Battery, working AC/DC input
  • Floppy Drive, CD-ROM
  • Pentium II 233 MHz Processor, 10 GB hard drive
  • Windows 98
• Wireless Assess Points in the Science Quad
  • 1 in the Science library, 1 on Bass 1st floor, 1 on the Engineering building 1st floor, 1 on McConnell 1st floor, 1 on Burton 1st floor, 1 on Burton 3rd floor, 2 on Sabin-Reed 1st floor, 1 on Sabin-Reed 2nd floor, and 2 on Sabin-Reed 4th floor. (802.11b in all except 802.11b in EGR building
  • complete coverage is the Engineering building/Science Library/Bass 1st and 2nd floor area, all others spotty
  • Coverage area will expand

Technical Specification Discussed

• Antenna v. Plate receiver
  • Field strength
  • Health risk
  • Amplification power
  • GPS receiver
  • GIS receiver
  • Signal processing
  • Server (802.11b 2.4 GHz)

Proposal Discussion with Professor Howe

• Mention the Design itself more directly
• Address refinement to prototype (make suggestions base on the survey results)
• Address Patent in the scope briefly, so it will not be abrupt at the end
• Add description of the map in the report, print the map in color
• Update Timeline and G Chart, with Spring semester included
• How does the product stands out from other devices/resources currently available at Smith

Presentation feedback from Professor Howe

• Need a short summary at the beginning to catch the attention of the audience
• Define technical earlier before using them or specify that they will be defined later
• Diagrams are small, find better diagrams
• Give more Pictures of the GPS system, give the audience and sense of how much room does a GPS system take up
• Provide images of devices current in used, to inform the audience of the size and usage (e.g. a blind person holding a small box and walking)
• Professor Howe commented on everyone’s presentation styles
• Need more focus on the introductory part; the presentation does not have to be too technical
• Address questions well, clarified aspects not addressed during the presentation

Plan of Actions/Tasks

• Jessie will schedule meeting with Tasha Next Sunday (October 24, 2004) for Brunch
• Sunday team meeting (October 17) in SAL at 1 pm
• Revise proposal
• Work on new timeline, redistribute research area to adjust to the current plan (starting with a laptop rather than from scratch)
• Write Thank you email to Bill from Technical Services for the donation of the laptop
• Review current progress (where we are on the timeline?)
• Visit ITS to fill out Sophia account application and ask for used hardware components (right after the meeting)

top of page

EGR 410D meeting Minute October 17, 2004 Sunday 1:00-4:00 pm
Team Reach meeting Location: SAL        Facilitator: Monica        Recorder: Xiaoning

Completed Tasks/Actions

• Updated each other with current available resources (laptop from ITS)
• Listed specification of the laptops
• Update everyone w/ the current Assess Points (response from Eric Brewer)

Technical Issues Discussed

• Team website encryption needed to protect the content
• Allow specific users to assess the team website
• Possibility of starting a new server to broadcast data
• Adopt removable storages
• 38.3 MB GIS data base is on Q:\Share\Campus\Geodatabase\campus

Discussion on Design Process

• Redistribute research area to fit new design process
• Start with a laptop as the fundamental processing unit for the prototype
• Adopt wireless adaptor
• Install voice software
• Install GIS software
• Install DGPS processing software
• ArcPad 6.0, preliminary system requirement, 31.1 MB hard drive and 64 MB RAM
• Microsoft SDK (voice software, recommended system requirement: 266 MHz CPU, 128 MB RAM
• Need to find the specifications of the respective software requirements

Proposal Revisions Progress

• Timeline is partially updated, with the tasks listed. We’ll need to add the dates and distribute the tasks

Plan of Tasks/Actions

•  Continue to revise the timeline during the Monday (10/18) class time (All)
• Distribute proposal revision parts by Monday (10/18)
• Meeting during Monday class time in EGR 201. We’ll move to a lab if necessary
  

top of page

EGR 410D meeting Minute October 18, 2004 Monday 1:00-2:30 pm

Team Reach meeting   Location: EGR 203 briefly, then SAL

Facilitator: Monica        Recorder: Xiaoning

 

Completed Actions/Tasks

• Brief meeting with Jon Caris regarding setting up a server
• Divided parts for the proposal for revision
  • Jie (Abstract, Design Process, Project Monitoring)
  • Jessie (Introduction, Motivation, Project Description, Technical Background, add description to the picture, format the proposal)
  • Monica (Project Scope, Voice Interface)
  • Xiaoning (Challenges and Restrictions, Progress, Budget, G chart, add few more dates to the timeline)
• We’ve compared our progress to the timeline and made appropriate adjustments.

 Administrative Changes

• There will not be a meeting this Tuesday (09/18/2004) to allow more time for members to work on revising the proposal

• The future Tuesday night meeting time and location are changed to 6:30pm at the Spatial Analysis Lab (SAL), instead of the original meeting time and location at 6pm at Lamont Dinning Hall.

 Meeting Summary with Jon Caris

• Eric and Jon are awared of the possibility of allocating space for us to setup a server
• The server is operated jointly by the Smith Environmental Science department (SAL, geology, etc) and the town’s Department of Public Work
• The server will not simultaneously broadcast signals, because this will need FCC approval
• Limited broadcasting is a possibility
• Or, we can use a device to connect to the server’s IP and read the data
• We’ll need to look into commercial wireless connections since the smith wireless is spotty.

 Design Process Revisions Discussed

• The timeline is mostly done. Xiaoning will add the dates for spring assignment due dates according to the syllabus.
• The parts are divided, so each member can work on their respective part during the week, keeping in mind that there is also a regular EGR 410D due.

Proposal Revisions Progress

• Timeline is partially updated, with the tasks listed. We’ll need to add the dates and distribute the tasks

 Plan of Actions/Tasks

• Invite Eric Brewer to join our Thursday meeting with liaisons
• Work on revising the assigned proposal parts
• Refer to timeline for individual tasks and write summery for completed tasks
• Setup password protection for the team website
• *Whoever is taking the GRE (> $100 registration fee), please inform the other members; we may take it with you.
• Amend the new Tuesday meeting schedule (6:30pm in SAL) in the charter.

 Recommendations

• Get used to check the Reach email account
• Now that Sophia accounts are created, we should learn to update the website ourselves, without relying on Jie all the time; she has a lot of work to do=)

top of page

EGR 410D meeting Minute October 21, 2004 Thursday 4 pm – 2 am

Team Reach meeting     Location: SAL, 4-5 with Jon Caris and Eric Brewer, then with the team

Facilitator: Monica        Recorder: Xiaoning

 

Completed Actions/Tasks

-          Meeting with Jon Caris and Eric Brewer regarding setting up a server

-          Revise Proposal

o       Jie (Abstract, Design Process, Project Monitoring)

o       Jessie (Introduction, Motivation, Project Description, Technical Background, add description to the picture, format the proposal)

o       Monica (Project Scope, Voice Interface, Research)

o       Xiaoning (Challenges and Restrictions, Progress, Budget, Gantt Chart, add few more dates to the timeline)

o       Compare our progress to the timeline and make appropriate adjustments

 

Administrative

-          We are not having a phone conference with Professor Fagg today (10/21)

-          Proposal revision editing time allocation

o       5-6 pm, Jessie and Xiaoning type, finishing up their parts

o       6-8pm, Monica and Jie compile and edit the parts while Jessie and Xiaoning go eat.

o       8-9, Monica and Jie go eat, and Jessie and Xiaoning compile and edit the parts

o       9-2am, the team compiles and finalizes the revision.

o       Jie will print the proposal and the team will sign it tomorrow (09/22 F) when meeting right before class at 1 pm.

 

Meeting Summary with Jon Caris and Eric Brewer

-          Use the current server vs. allocate space from current server for our usage

-          The choice between WiFi (~5 MB /s) and WAP. Eric anticipate future assess point expansion with WiMax, the successor of WiFi on campus, that one antenna covers the entire campus

-          We can ask Professor Howe to apply for Grant from Smith CET to WiFi expansion on campus.

-          The data output from the base station needs to be more often, the current rate is hourly. Jon will look for the base station manual.

-          WAS and GPS, without DGPS, can also give us a accuracy of up to 3 meters

-          Ask Physical Plant to give the updated versions of maps, when different events impose a change on the geography (e.g. temporary blocked path, new construction area) of the campus

-          Decide how often synchronizations need to be done

-          Need to establish server assess accounts

-          Path finding program (store the path as one walk)

-          Voice Assist for ArcPad is hard to use (e.g. current commands include “zoom in” and “zoom out,” very visual terms that are difficult for the visually impaired to understand)

-          Sequo Server, software that allows multiple users to edit the data base simultaneously. Our team has to decide whether we need it or not.

 

Plan of Actions/Tasks

-         Print the proposal (Jie) by Friday, before 1:00 pm

-         Sign the proposal (All), when meeting right before the class at 1:00 pm

-         Schedule meeting with Tasha (Jessie if she sees Tasha, Jie through email)

-         Sunday (10/24) meeting at 1 pm, in SAL

-         Contact John Crowly and Bill W. regarding how to set password to our website (Jie).

-         Write up the individual research summaries, if not already completed by the set deadline, by the Sunday (10/24) meeting

-         Write the Progress Report

top of page

EGR 410D meeting Minute October 23, 2004 Sun 1pm -3pm

Team Reach meeting     Location: SAL

Facilitator: Monica        Recorder: Xiaoning

 

Completed Actions/Tasks

-          make adjustments to the Chronological Task List

o       add task: email Sherril Obremsky. Oct 24-Oct24 (JM)

o       edit task: obtain most current CAD file Oct 24 – Nov 6, originally Nov 08 – Nov 20 (Jessie)

o       add task: email Professor Howe regarding CET funding to expand WiFi coverage on campus (Jie, XX bringing up in meeting on Mon) Oct 25 - Oct 25

o       add task: test WiFi coverage on campus w/ Xiaoning’s laptop Oct 29 – Oct 29 (Jie)

o       add task: find out the file format used in different GPS software bundled with different GPS units. Oct 25 – Oct 25 (Jie, XX)

o       edit task: research compatibility w/ complementary devices (pedometer, Braille keyboard) Oct 25-Oct 25 (XX)

o       edit task: market review on Trekker,(JZ, JM)

o       add task: research forward and backward communication among GPS, GIS and Voice interface. Oct 24 – Oct 31 (XX)

o       add task: prepare a budget binder to keep track of expenditure. Oct 24 – Oct 31 (XX)

o       add task: research on path finding software, Oct 24 – Nov 6 (JM, XX)

-          Compare our progress to the timeline and made appropriate adjustments.

-          Briefly exchanged summaries of individual researches

 

Administrative

-          Budge Binder, to keep track of expenditure

-          Need lock for a team locker (JZ)

-          Proposing purchasing a laptop hard driver enclosure (XX)

-          Rotate team roles. Facilitator: Xiaoning. Recorder: Jessie

 

Brief Summary of Individual Research Findings:

-          Microsoft SAPI enables programming in VB and C++

-          Microsoft SDK offers framework to connect to target software

-          Microsoft will release SAPI 5.1 tomorrow (Oct 5 M)

-          Voice Insight by VQL Technology, act as a bridge software between voice recognition software and a targeted software

-          WASS (up to 3 m) vs. DGPS (need to find out how to transfer the signal, or both, to improve accuracy

-          A “guy” in California specialized in DGPS signal transfer, keep an eye on his publication and contact him when necessary

-          Running software on both Windows 98 and Windows CE

 

Plan of Actions/Tasks

-         Meeting with Professor Howe on Monday, 2:40 pm, (all)

-         Read guidelines for working presentation by Tuesday (Oct 26, all)

-         Schedule meeting with Tasha (JZ, XX), right after the meeting, 3 pm

-         Incorporate the task changes to the Chronological Task List by Tuesday (Oct 26 XX)

-         Write the Progress Report by Monday (Oct 25, XX)

-         Complete the tasks on the Chronological Task List by the set deadline

top of page

10/25 EGR 410D meeting Minutes October 25, 2004 Monday 2:40pm– 3:10pm

Team meeting with Professor Howe, Location: EGR Computer Lab

Facilitator: Xiaoning/ Recorder: Jessie

 

• Discussion:
• Team lock and locker available
• Adding more money to Print card: Enquire at OneCard office
• Obtain reimbursements using Expenditure Request and Reimbursement form-check Design Clinic fund code and account code for specific task
• Software licenses are available at Smith via group license, so purchase will not be necessary
• Suggestions from Professor Howe:
• The CET (Committee on Educational Technology) has a standard form for online funding requests, the sections of which need to be filled out by the team. Alternatively, Eric Brewer could be asked whether the form process may be circumvented
• Send email request for Working Presentation Day
• For future progress reports: Reiterate Chronological Task List tasks in Plan of Action/tasks
• Revised Proposal:
• Good revision in focus on software additions to laptop
• More positive focus needed in paper
• Design process section needs better organization; include Communication as main part
• Redundancy in Research & Product Development
• Consider revised order for Design Process section (see comments on paper); select software, then design conceptual model
• Refine Flowchart in Appendix D
• Good work on Chronological Task List/Project Timeline
• Tasks:
• Obtain team lock and locker—Locker #5, lock 12, combination: (see email)
• Check out CET funding form online at: http://www.smith.edu/cet/cet_tech_init.html
• Send email request for last possible Working Presentation day
• Prepare for Tuesday meeting

 

top of page

10/26 EGR 410D meeting Minutes October 26, 2004 Tuesday 6:30pm– 7:30pm

Team meeting, Location: SAL

Facilitator: Xiaoning/ Recorder: Jessie

 

• Discussion:
• Chronological Task List-updated and changed
• Research and Task updates
• XX-
• Complementary devices, pedometers
• Self-contained or attachable?
• Working Presentation data set to Nov 12th
• JM
• Emails: to Cheryl Obremski about CAD files, from Andy Kuether and Bob Newton re:-base station-now forwarded to team account
• Pathfinding algorithms online-will research more by Sunday
• MM
• SDK &. SAPI: on-campus resources, who to contact
• JZ
• DGPS and WAAS-product research
• Team administrative tasks
• Website encrypted with password protection
• Team locker selected (#5), lock combination distributed
• Round table discussion
• Questions/Comments-misc. (off-topic)
• Meeting effectiveness J
• Tasks:
• Working Presentation 1 (WP1)
• For Nov. 4 th-Think of ideas for ‘issue’ to be discussed during WP1
• For Nov 7 th-Contact friends to schedule audience sit-in for WP1 practice and brainstorm session
• For Nov 11 th-Prepare for run-through at liaison meeting and check into video conferencing possibilities for Prof. Fagg.
• General
• Continue with tasks listed on Chronological Task List (Henceforward referred to as CTL)
• Prepare for next meeting

top of page

10/28 EGR 410D Meeting Minutes October 28, 2004 Thursday 4:00pm – 5:15pm
Team meeting, Location: EGR 102
Facilitator: Xiaoning/ Recorder: Jessie

Telephonic conference with Prof. Andrew Fagg

Discussed:

• CET funding
• New research sections, JZ, MM, JM, XX
• WAAS & GPS enabled receivers
• Possibility of running a cable between two units
• Engines & trackpads
• Serial port over WIFI
• Blutooth technology
• Use TRIMBLE-brand receiver—NMEA format (standard that defines an electrical interface data protocol for communications between marine instrumentation)
• 2 serial-port receiver?
• Location accuracy- 3 meters deemed reasonable, will ask Tasha Friday at 4 pm
• Prototype: Connections by wire now, possibly wireless in future
• Campus paths in ArcGIS map—to be buffered out to 3m to check system requirements
• Windows CE now loaded on dual-boot computer (donated laptop)
• How does one query the ArcGIS system?
• GIS layers-paths, roads
• Shortest path algorithm already in place, needs ‘smart network’ to work with
• Send latitude-longitude value in, return data: buildings, elevation, etc.
• VoiceAssist for ArcPad—made by Voice Insight Co.
• VQL applications website examined
• VQL connects ArcPAD with voice software (as shown in website diagram)
• Voice Insight is a Xybernaut partner company
• In future, email website links ahead of time
• How will the path-finding system be adapted for ArcPAD?
• Path-finding: Graphs, nodes, links and intersections
• Must make network-aware path layer—is this possible for ArcPAD?
• Must find paths and routes, adapt algorithms for pedestrian path-finding

Working Presentation rehearsal on Nov. 11 th- videoconference to be arranged.

top of page

10/31 EGR 410D Meeting Minutes October 31, 2004 Sunday1:00pm-2:00pm
Team meeting, Location: SAL
Facilitator: Xiaoning/ Recorder: Jessie

Discussed:

• Review of Chronological Task List-adjust dates
• WIFI coverage testing
• Campus map updates
• Individual Research Reports
• JZ,XX
• Reviewed flow chart with Tasha, made adjustments
• Terminology: ‘path’ is subset of ‘route’
• 3 meter accuracy okay
• User testing on Tuesday
• JAWS-screen reader
• JZ
• Trimble GPS receiver and others
• MM
• Inertial navigation systems
• System available on Ebay which contains GPS and GIS
• INS/GPS integrated system
• JM
• Base station books
• Obrymski-CAD file updates in progress
• GIS stuff-campus map updates
• Pathfinding
• XX
• OS loaded, WIN 2000
• WIN CE on free trial
• Large RAM requirements
• Working Presentation:
• Videoconferencing
• Topics for next week
• Happy Halloween! J

top of page

November

11/01 EGR 410D meeting Minutes November 1, 2004 Monday 2:00pm – 2:30pm
Team meeting with Professor Howe, Location: EGR
Facilitator: Jessie/ Recorder: Jie

• Completed Tasks
• Prof. Howe handed out Critique summaries of our Proposal presentation given by classmates
• Strengths: clear context, good organization, conveyed “big picture,” good Q&A session, strong grasp on knowledge
• Future Improvements: define acronyms, clearly state final deliverages, clarify technical background, adding visuals, practice presenting, budget slide was unnecessary, do not read from slides, add final prototype image, discuss hardware/software
• Prof. Howe returned Progress Report 6
• Each team member provided a summary of up-to-date progress
• Xiaoning – Partitioned the hard drive so that the laptop has 3 operating systems (OS): Windows 2000 for file transfer, Windows 98 (the original OS) and Windows CE 2.0
• Jessie – Needs to get in touch with Phys Plant to update files. Obtained general understanding of path finding software and the two kinds available: vector and raster based.
• Monica – For speech interface, has understood that the components necessary may include SAPI, SDK and VQL. Inertial GPS (INS) comprises 3 gyroscopes and accelerometers in each of the axis.
• Jie – Gained understanding on WAAS system. Called Garmin about GPS receiver products. Searching for appropriate GPS receiver from Trimble that meets criteria of WAAS, DGPS and also INS. Has found one that meets the first two criteria (BD950).
• Discussion
• Future Progress Reports
• Do not just copy and paste from minutes, difficult for reader to understand the progress made, should be reorganized
• Administrative tasks do not need to be mentioned
• Discussion points are most important (could add table for it)
• Better to mention discussion points rather than list tasks completed or to be completed from the Work Budget Sheeting
• Would have been very helpful if research summaries were included in Progress Report
• Working Presentation
• Used to get classmates to help us with issues
• Will focus more on user-related issues than designer
• Revise our flowchart, as it was developed only with input from five people
• Do not need to limit ourselves to current issues, can also mention issues coming up in the future
• Can provide interactive activities!
• Nothing too technical unless we can describe the technical background very clearly and thoroughly
• Opportunity to practice explaining things before the technical presentation
• Class will not really be an audience. They will provide feedback through discussion.
• Tasks to be completed
• Jie will prepare the CET proposal in the specified online format and email those parts to Prof. Howe by Thursday night. This proposal will include a concrete request for more access points (i.e. it will address how many and where they should be installed—this info will be obtained after WiFi testing).
• Jie and Xiaoning will meet conduct WiFi coverage testing in the Science Center at 4pm tomorrow using a Campus Map. Trimble Explorer is not needed for extreme accuracy. Just a general idea is good enough.
• Jie and Xiaoning will meet with Tasha at 5pm tomorrow to test out the flowchart using two routes chosen with the help of Tasha in the Science Center. They will use Trimble Explorer to chart the paths and landmark points.
• Jie will read Trimble manuals on the base station located at Smith College. She will learn about transfer of information via the internet and the DGPS data format the base station sends out.
• Xiaoning will install ArcPad on laptop.
• Xiaoning will investigate path-finding software options available in ArcGIS.
• Jessie will call Physical Plant about the CAD files as the email request she sent last week has not be replied to. Then, she will update those files.
• Jessie will investigate the path-finding options available in ArcGIS.
• Monica will establish a firmer understanding of SAPI and SDK applications (their differences, similarities, usage) through talking with a Comp. Sci major and/or calling companies.
• Monica and Jie will search for commerical INS products reasonable for our use. Monica will look for independent components, and Jie will look for GPS receivers that are INS-enabled.
• Members will post research blurbs up online

top of page

E L E C T I O N D A Y
11/02 EGR 410D meeting Minutes November 2, 2004 Tuesday 6:30pm – 7:45pm
Team meeting, Location: SAL Facilitator: Jessie/ Recorder: Jie

• Completed Tasks
• At 4pm today, Jie and Xiaoning conducted WiFi coverage testing in the Science Center using Xiaoning’s WiFi-enabled laptop. Their findings will be marked on a map.
• At 5pm today, Jie and Xiaoning met with Tasha to walk 2 routes
• Original plan was to guide Tasha using Flowchart, but problems with being able to read the menu in the dark and also potential incompatibility with Tasha’s navigation style
• Therefore, we decided to have Tasha talk us through 2 routes so that we could observe her navigation style, the landmarks she uses, potential difficulties. Also, we used the Track logger to record one route
• Hubbard House to Wright Hall
• Hubbard House to Bass Hall
• Xiaoning created the research sections online for each member.
• Discussion
• WiFi Testing
• Walkway from between McConnell and Sabin-Reed to end of Burton first tested
• From Burton to Bass to Neilson and finally Washburn tested
• Only part not covered was the Neilson tunnel
• Two kinds of connections discovered
• Network to Computer (access points established by Smith)
• Computer to Computer (likely to be established by students with permission)
• Tasha Interview
• Jie called Trimble two times already, but she has not been able to get in touch with tech support yet. They left a message saying that they will call tomorrow.
• Path Finding software – 3 options have been explored by Jessie
• ArcNetwork (application most closely related for our use)
• License about $80
• Allows user to build a smart network of routes and locations
• ArcPad Streetmap built for ArcPad
• Has own database of roads, streets
• Built-in routing function
• If we choose this, must add more routes
• ArcGIS Program Extensions
• Free website downloads for ArcGIS written by various users
• Uses Arc objects, plot and info layers
• Members discussed on-site (saved on laptop) vs. off-site storage (GIS server) of path-finding software
• Software for 2 applications
• A) Shortest distance when user is off-path
• B) A-to-B route-generation
• B) is most important for our purposes at this point, and should be stored in Server due to processing required and size.
• A) should be stored in laptop to save time and would not required much processing power or hard drive space.
• Will mention at Thursday’s phone conference
• Is Inertial GPS necessary for our project at this point?
• Considering probably the high cost and sophisticated usage of INS by aircraft, do we need to include this into our one-year project?
• Must compare INS-enabled GPS receiver versus separate units. Jie will look at former, and Monica will look at latter.
• Working Presentation
• Eliminated practice workthrough presentation session for this Sunday
• Jie suggested having an interactive demo of our device with one member pretending to be blind and the other acting as the machine to jumpstart the presentation
• Tasks to be completed
• Jie
• Contact Media Services about the details of the computer-to-computer WiFi connection coming from the Alum Gym on Wednesday.
• Email the sections for the CET proposal to Prof. Howe by Thursday evening. Deadline is Friday, so Prof. Howe can email it in by deadline.
• Write up summary of findings from 2 nd meeting with Tasha by Sunday
• Plot out WiFi coverage testing points by Thursday
• Read Base station manuals
• Get in contact with Trimble and ask about suitable INS-enabled GPS receivers
• Jessie
• Call ITS to find out why she and Jie does not have access to the Research folder for our website by Thursday. This problem is preventing members from uploading and posting their research blurbs.
• Call Sharon at Phys Plant by Thursday to inquire about files
• Further investigate ArcNetwork by Sunday
• Is it compatible with ArcPad?
• Processing power required
• What existing info is required?
• If GIS data can be used, what format must the data be in?
• Scroll through Q:/ drive to find more current versions of files and copy them over. Currently, files date from 3/2003 to present. Changes in sidewalks since files last updated.
• Eventually must add landmarks (grasslines) to database, but team must determine easiest way to plot out landmarks (use pre-existing files OR use track logger???). Should do very soon, so team can ask Prof. Fagg for advice.
• Xiaoning (I know you have more tasks, in addition to those on the task sheet, but I don't know what they are. please feel free to add on!)
• Write up summary of findings from 3 rd meeting with Tasha by Sunday
• Monica
• Investigate separate INS units (not built into GPS receiver units) by Sunday
• All team members
• For working presentation, each team member should prepare 2 issues that can be brought up. Can be a current issue relevant to member’s research area or an upcoming foreseen problem that we are not dealing with head-on currently
• Post research blurbs online by Sunday

top of page

11/04 EGR 410D meeting Minutes November 4, 2004 Thursday 4:00pm – 5:30pm
Liaison meeting, Location: Media Services Videoconference room Facilitator: Jie/ Recorder: Jessie

• Completed Tasks
• Discussion
• Summary of Meeting with Tasha (see Tuesday minutes)
• Pathfinding-the most important aspect will be the user finding their way back to the path (shortest distance to path algorithm)
• The path algorithms may need to be server-based
• Algorithms will compute distance to line segment
• The ArcNetwork application needs to be further investigated
• JAWS-The Jaws program is a screen reader which converts text to voice
• For our project, converting speech input to commands will be more important
• Tasha currently users a normal keyboard for inputs
• Our project doesn’t need a full speech engine; Microsoft applications have a text reader already
• Dual boot WIN 98/ WIN CE OS system
• Triple-partition already set up on laptop
• Will WIN CE be able to support path-finding software?
• WiFi testing—two types of accessibility options found
• Access points set up by the college
• Router points set up by individuals
• GPS receiver
• Trimble contacted by Jie
• Lassen iQ starter kit recommended
• Approx. $695
• Power range, wireless accessiblity unknown
• Uses external power converter
• According to contact, WAAS is generally less effective than DGPS
• Autonomous GPS has a location accuracy of 5m, WAAS 3m, and DGPS 1-2m
• Garmin products should also be looked into
• They may provide a direct connection to the laptop instead of a USB cable
• Should be NMEA and WAAS-enabled
• Base station output data is in CMR+ format, RTCM protocol
• Is there a CMR+ à NMEA conversion program
• Need to confirm w/Trimble whether receiver accepts CMR+ protocol, whether software works on system
• For our project, we should probably start with GPS/DGPS, and leave Inertial Navigation System (INS) for a later add-on to the system
• When the user is close to a building she doesn’t need great location accuracy
• The system should guide the user to a point nearby the entrance, not to the door
• WiFi access points are generally set up inside buildings and readily accessible near buildings
• INS/GPS handheld systems are currently under research
• USC has a research project involving a robot helicopter with roll, pitch, and yaw control.
• Inertial Sciences has an ‘IMU’
• Research papers recommended by Prof. Fagg:
• Kalman filters à use current position and rate of travel to predict future location, deal with imperfect sensor information
• It might be possible to set up a program to use GPS data to predict future location using Kalman filters and linear pathways—we should talk to Prof. Pau Atela
• We want our project to be used in a controllable situation, in which the user will not be led into harm’s way
• The IRB board needs two weeks notice to approve petitions for human subject testing
• The Working Presentation will be reviewed during next week’s conference
• WiFi connection for project should be set to log on to strongest signal until the signal is lost
• There is a time delay to switch signals
• Perhaps we can use API to search for a strong signal
• Linux software has this capability
• Lower connection bars mean more background noise and a greater possibility of losing the signal
• For the GPS receiver, there is a USB to computer adaptor. The computer draws poer at 500mA and 12 V.
• We need a DC to DC power converter
• Prof. Fagg’s friend at the U. of Minnesota would like to be contacted by the team
• He runs the disability services program and can provide feedback on technical problems
• Team members should think of questions to ask via email
• The WP review next week will last approx. 25 min.
• Each team member should come up with 2 issues for the WP by Sunday

top of page

 

11/07 EGR 410D meeting Minutes November 7, 2004 Sunday 1:00pm – 4:00pm
Team meeting, Location: SAL/EGR
Facilitator: Jessie/ Recorder: Jie

• Completed Tasks
• Jessie discussed her findings on route-finding software. The options:
• ArcNetwork with ArcInfo (desktop) – Since ArcNetwork is an extension of ArcInfo, we must purchase ArcInfo. ArcInfo is a fancier version of ArcGIS. It is only compatible with desktops and would have to be placed on the server due to its high processor/OS requirements.
• ArcGIS with Map Objects (desktop) – Mentioned by Jon. Need to find out if can be used separately from ArcInfo.
• ArcPad with Streetmap (handheld) – Has routing capability and comes with database. We select stops and choose routes. We can set impedance values, but program defaults to shortest length. Need to find out if database that it comes with can be modified.
• ArcGIS Network Analyst (desktop or handheld?) – Coming out in the first half of 2005. Seems to be exactly what we need, but cannot depend on it due to its uncertain release date.
• ArcLogistics Route – Has routing possibility, but could not find evaluation version online
• Installed Map Objects on desktop and Arc Streetmap on handheld
• Had difficulty with getting sample Map Objects and Streetmap files, found online, to work. Need administrative permissions to access them.
• Xiaoning
• In terms of voice software, IBM Voice provides speech-to-text, which should be the focus of our project. Not sure about how this would be connected with commands.
• Contacts for path-finding software
• Pau Atela – MTH 204, is familiar with various software and may have existing algorithm)
• Judith Cardell – Comp EGR, can help with programming in Visual Basic (used for ArcGIS)
• Nick Howe – Comp Sci, on sabbatical
• Currently running trial version of ArcPad on laptop. Need to find serial number.
• WiFi campus connection
• Internal setting for refresh rates (currently, there is at least a 3s delay).
• Each access point has log-in process, which is time-consuming. Will contact ITS about ways around this issue.
• Monica
• Confirmed that we are not trying to find a single GPS, DGPS, INS unit
• INS will be add-on if we find something appropriate
• Found an INS that can serve as an add-on part
• Rate Gyroscope GR 203 by Silicon Sensing System Japan (29mm 2) <http://www.spp.co.jp/sssj/silicon-e.html>
• Outputs voltage through pins
• Concluded that Kaulman filter will not serve as a good model since it only works for linear paths, which does not apply in our case
• Met Poornima along with Jie to get a better general understanding of SDK, SAPI and VQL
• SDK (software development kit) is umbrella of libraries and functions. It can be used for many applications, and speech is just one of them. It encompasses SAPI. SDK and SPI each have their own libraries.
• Use VQL as a communication tool to connect SAPI database with commands.
• Jie
• Met with Jon last Friday since Jon could not make Thursday meeting
• Realized that tracklog data from Tuesday with Tasha was not saved!!!
• Need to establish area of testing and focus on adding landmarks into the GIS database within that area of test
• Two ways to add more info
• Use Ortho maps
• MassGIS, which has not been updated for about a year so inaccurate
• Sharon O. at Phys Plant knows a guy who takes close aerial photos for fun. If we could obtain photos, we could make them ortho using software
• Field testing using receivers and tracklogger
• We must choose the landmarks to include in new GIS layers
• Grasslines
• Fences
• Walkway edges
• CMR and RTCM data formats for DGPS – Did not find conversion software.
• From base station manuals, Trimble receiver hardware is either RTCM-enabled or consists of RTK (realtime kinetic) to accept CMR. From this assumption, we would need RTK equipment.
• Discussion
• Xiaoning, Jessie and Jie all experienced difficulties uploading files online. Team members should ensure that file permissions are set so that all can access every file and that file paths in links are correct.
• Xiaoning addressed whether we should buy WiFi adaptor now.
• Specifications of this do not depend on other components
• Team agreed that we could buy one that 802.11b.g card now
• Xiaoning expressed concern over the adaptability of path-finding software. She encouraged team to use software protocol that imports commands.
• Every Sunday, team members will email Xiaoning the changes that need to made to the Timeline, and she will edit it by the next meeting.
• GIS Day coming up on Wednesday, 11/17 th, and events will be held in the Campus Center. Team members will try to attend some of the events.
• Team planned out Working Presentation (WP) and divided up parts
• Interactive System Demo
• W/ team member (Monica acts as blind person and Xiaoning acts as machine) (3 min)
• Gives class overview of what our device does
• Allows class to see potential problems with system
• Monica will take a pre-planned route across the class from Start to End passing 2 “landmarks.” She will purposely go off-path so we can show alert function.
• Xiaoning must walk by Monica closely while giving commands so her voice does not indicate location and simulates worn or handheld device
• W/ class volunteer (Find volunteer to be blind person and Jie acts as machine) (6 min)
• The volunteer will be guided with instructions down a different route from the first demo
• Blind-folded and turned in circles several times before beginning.
• From Start to End, she will pass 2 “landmarks.”
• Flowchart Discussion (Jie) (10 min)
• Technical Overview (Jessie)
• Board drawing of the basic components of system with arrows: device, GPS, DGPS base station, server with GIS database and route-finding software and user (to show voice interface) (5 min).
• Show and tell of actual laptop. Basic explanation of hardware and software that needed to be added (5 min).
• User-testing (Monica) (10 min)
• How to go about testing users to evaluate system?
• Communication (Xiaoning) (10 min)
• How to make communication better between Phys Plant, Public Safety, Noho DPW and us?
• Tasks to be completed
• Jessie
• Need to find out if Arc Streetmap database that it comes with can be modified.
• Can Map Objects be used independently from ArcInfo?
• Call Sharon at Phys Plant
• ArcLogistics Trial CD can be ordered for free http://www.esri.com/software/arclogistics/download.html
• Prepare the Technical Overview part of WP
• Jie
• Buy phone card
• Email MassGIS to find out how often ortho photos of Smith are updated and whether they can take a current photo
• Email Prof. Howe to request to go 2 nd on Friday
• Call Trimble
• Ask about CMR and RTCM conversion
• Call Garmin
• Ask about CMR and RTCM conversion
• INS GPS products?
• Prepare volunteer demo of interactive menu and Flowchart discussion of WP
• Xiaoning
• Email to make appointment to test Dragon Dictate and Naturally Speaking with Monica
• Find ArcPad serial number
• Prepare team demo of interactive menu and communications sections of WP
• Decide whether handout is necessary, will prepare if necessary
• Email ITS to find out more about WiFi logon
• Monica
• Email or call Silicon Sensing System Japan to find spec sheets
• Prepare team demo of interactive menu and user testing sections of WP
• Decide whether handout is necessary, will prepare if necessary
• Ask UMass Comp. Sci students and professors for more info on SAPI, SDK, VQL

top of page

11/08 EGR 410D meeting Minutes November 08, 2004 Monday 2:45pm – 3:15pm
Team meeting with Professor Howe, Location: Engineering Building, Rm. 102

Facilitator: Jie/ Recorder: Monica

• Completed Tasks
• Preliminary plan of our working presentation submitted to Prof. Howe
• Discussion
• Progress Report – good amount of detail; should end or begin with a short summarizing narrative that extracts the key points of the week’s activity to help our liaisons have an overview and us compile the midyear and final report
• Working Presentation
• Handouts would be useful to use
• We present after the mid-class break, and thus have time to move the furniture back the way we need it
• Include question at the end that makes them repeat what they’ve understood from the technical presentation of the device and see if it worked better this time
• Keep an eye on what we see works better and include those parts in future presentations
• Have clear in mind what we want to get out of this presentation

• Help with constructing our flowchart; get input on possible improvements
• Provide out classmates with a better understanding of what the progress involves in terms of technical background and functionality of final product
• Get them brainstorm about other ways to reach possible users from the community and get more people to test our prototype

• Where do we see us in the process: by the end of this semester, we should have a list with all components to be purchased, and be sure they are compatible; we can purchase part of them before and part immediately after Winter break
• March (during Spring Break) NCIIA conference in San Diego: all members can come but Jessie; Prof. Howe will be there anyway; she will see if there is any paperwork we need to do, since we get funded by them (there is a section where such groups present); check online at www.nciia.org (9 th annual conference)
• Tasks to be completed
• Include in the minutes what “Tasha talked us through two routes” means
• Said what landmarks she was reaching while walking
• Pointed out problem areas (bifurcations of the path in front of neilson Library)
• Jie – resend e-mail to Mass GIS, specifying coordinates of Smith campus
• This week we will be working on

• Jessie – path finding algorithms
• Jie – contact Mass GIS to inquire about the possibility that they provide orthomaps for the Smith campus more often than once a year; contact Trimble to find out about the different corrections capabilities (? compatible with an inertial system)
• Xiaoning – test Naturally Speaking and Dragon Dictate software on campus (with Monica); contact Prof. Pau Atela about path finding algorithms, contact UMass Amherst about the association for the visually impaired; buy wireless adaptor card
• Monica – find professor in the Com Sci department at UMass that has working knowledge on SDK, SAPI; get in contact with the Japan company producing the only separate inertial system we’ve found, to see if it would be compatible with the Trimble GPS receiver
• Other people to contact: Prof. Nick Howe

top of page

11/09 EGR 410D meeting Minutes November 09, 2004 Tuesday 6:30-7:45pm
Team meeting, Location: SAL
Facilitator: Jie/ Recorder: Monica

• Completed Tasks
• Jie found out from MassGIS that the earliest date we can expect a new orthophoto of the campus to be taken is Spring; as a result, we need to go out with the Trimble receiver and record the landmarks ourselves
• Discussion
• Access of our website: we will notify all team members before giving the password to new contacts; their names will be written in our binder; if needed, we can later set passwords for different levels of access
• Jessie is still trying to contact Phys. Plant
• Xiaoning has talked to Prof. Pau Atela about the path finding algorithms we could use
• She will share the results of the meeting with us after the video conference on Thursday
• She will schedule a meeting with him for Tuesday at 4:00 pm, where all team members will be present
• Working Presentation 1:

• What to bring in class: Laptop that we will use; scarf, meterstick, prize for volunteer and audience members that talk
• Jie will have the flowchart prepared; all blocks included, but with less detail in each of them
• We will go after the break => set up the furniture in the room, of needed, place plates for all four start and end points, have flowchart drawn on the board but hidden, write agenda on the board
• Have prepared two of each type of handouts per group (14 copies of each handout at least +prof howe+us =>no more than 20 copies!)
• Ask them to get together according to their design teams

Jessie	Jie	Monica	Xiaoning
	Have flowchart drawn on the board but hidden at the beginning
Set uo furniture in the room	Write agenda on the board	Set up furniture in the room; ask them to get into groups according to their design teams	Place plates for all four start and end points, distribute all handouts
Say intro
		1 st Blindfold presentation	1 st Blindfold presentation
	Ask for a volunteer
	Tell volunteer to say “end” for destination, blindfold her, spin her, bring her to the new starting point	Tell them we’re looking for ideas; we will not collect handouts, but have a discussion at the end
	2 nd Blindfold presentation
	Give prize to volunteer
	Tell them what we’re looking for, and let them discuss in teams
Go and discuss with teams	Go and discuss with teams	Go and discuss with teams	Go and discuss with teams
Write on the board suggestions that come out of the discussion	Conduct flowchart discussion, ask for ways to improve
	Transit to Jessie’s part
Present section on technical background – draw on the board the different components of the device: what goes in the system (+WiFi), what goes out
Conclusion: this is what we work with: show laptop
		Transition from technical section
		Conduct user feedback and testing section – tell them what we’re looking for
Go and help groups, if needed	Go and help groups, if needed	Let groups talk about it, go and help groups, if needed	Go and help groups, if needed
		Get feedback
			Transition from user-feedback section
			Conduct communication section – work with Jessie’s map (? Flow of events)
	Conclude

• Tasks to be completed
• Jie and Monica will set up a time to decide on landmarks in the Central Campus area and go out to record them until the meeting on Sunday (landmarks right now include: grass lines, fences, cement/blacktop paths)
• Jessie has to pull out a GIS file for Tuesday’s meeting with Pau Atela, so that he can see how the layers’ data is stored
• Task 2
• Next meeting: 4pm Media Services, video conference with presentation of the WP1 to Prof. Fagg and Jon Caris and get their feedback; practice the WP1 once again as a team; hera about Xiaoning’s meeting with Pau Atela

Meeting on Sunday: 8pm in SAL!!!!!

top of page

11/11 EGR 410D meeting Minutes November 11, 2004 Thursday 4:00pm – 7:30pm
Team meeting – preparation for WP1, Location: Media Services, EGR 201, Tyler Dining Room
Facilitator: Jie/ Recorder: Monica

• Completed tasks
• Video connection with Prof. Fagg could not be established
• Team practiced working presentation and got feedback from Jon Caris
• Team members practiced for a second time in EGR 201
• Team members received input from one another
• Team decided on handouts to be distributed to class
• Tasks to be completed
• Jie will print the handouts
• Also to be brought in class tomorrow: walking stick, blindfolder

top of page

11/14 EGR 410D meeting Minutes November 14, 2004 Sunday 8:00pm – 10:00pm
Team meeting, Location: Tofu Restaurant, Northampton
Facilitator: Jie/ Recorder: Monica

• Completed Tasks
• Feedback from Prof. Howe and classmates was read and distributed to team members according to their sections
• Tasks to be completed
• Meeting with Prof. Howe: Monday at 1:10pm
• See Prof. Pau Atela on Tuesday at 4:00pm as a team
• Use team time on Monday to go over what Prof. Atela told Xiaoning at their previous meeting
• Discuss feedback from classmates; did we meet our goal with WP1?
• Help with constructing our flowchart; get input on possible improvements
• Provide out classmates with a better understanding of what the progress involves in terms of technical background and functionality of final product
• Get them brainstorm about other ways to reach possible users from the community and get more people to test our prototype
• All team members will use the rest on the class time to record landmarks across Central Campus using the Trimble receiver

top of page

EGR 410D meeting Minute November 15, 2004 Monday 1:10-3:00 pm

Location: Engineering Lounge    Meeting w/ Professor Howe and team meeting

Facilitator: Monica   Recorder: Xiaoning

 

Administratives

-         Our presentation video is not ready yet

-         We can check out other team’s video for the doing the critique

-         E-tem conference

o       NCIIA, registration in February

o       Grants available to continue the work

o       Schedule in Spring to discuss about the continuing gof the work

o       www.nciia.org

-         To have a section on progress report including items purchased, items finalized and items arrived

 

    Working Presentation

-         People were really engaged

-         Consider a handout with key terms defined

-         Discuss what did we got out of the WP1

-         Flowers, not completely gave out

-         Professor Howe’s Comment

o       Goals not fully articulated at the beginning, could announce ahead so people can think ahead how to fill out the worksheets

o       Good schematics

o       Good review of the design breakdown (fall/spring)

o       Demonstration kept the audience alert

o       Consider picking just important questions when running out of time

 

    Plan of Actions/Tasks

-         Get Landmarks (today)

-         Paul Atela meeting (tomorrow, T, Nov 16, 2004)

-         GIS Day

-         Type up comments from the class for WP1

top of page

EGR 410D meeting Minute November 16, 2004 Tuesday 4:00-5:00 pm

Location: Burton third floor    Meeting w/ Professor Atela

Facilitator: Monica   Recorder: Xiaoning

 

Maps

-         Reconstructing possible path using the otrhomap, connecting only the points on the sidewalk

-         Using Cad files with high resolution

-         Identify doors using Cad files, only one door is necessary per building

-         Obtain Cad files from physical plant, presumably constructed using the building plan

 

Path-finding

-         Use only the sidewalks

-         Two kinds of algorithm necessary, one is for finding the entire route and one is for getting back on the route when strayed off

-         Identify and test for practical error on Smith campus through experiment to narrow down possibilities

-         Explore or adapt MapQuest and its algorithms

 

Plan of Communications

-         GIS day (Nov 17, 2004 W) (all)

-         On GIS day, talk to Jon Caris, Don Clark, and ask questions (all)

-         Contacting the Geology Department (to be determined)

top of page

EGR 410D meeting Minute November 18, 2004 Thursday 4:00-4:00 pm

Liason meeting with Professor Fagg and Jon   Location: EGR 102

Facilitator: Monica      Recorder: Xiaoning

 

Reporting to Professor Fagg

-          WP1, incluing the user testing, communication, flow charts

-          Recent researches

-          GIS Day, Andy Kecther

-          Analyst software requested, anticipate to get it for free

 

Technical Specifications Discussed

l         GPS receiver in general

-          CMR, more compact format than RTCM

-          RTCM (need to request for setup)

-          Cellar modem connections

-          TCPIP Protocol file transmiison

-          802.11 vs. cellar modem

-          Use another base station to transmit correction signals

-          MIT research group has studies on data transmission

-          Find out price of Garmin relative to Trimble

-          Garmin: 2 serial ports

-          WiFi through Laptop vs. radio transmission

l         Trimble

-          Data logger (small computer)

-          Need host name/IP address/Port number

-          Eye Liser,

-          Use serial connection to issue correction signal

-          Need to buy separate Digital compass

-          Acurracy. Trimble 1-2 m. Garmin 3m.

-          Outputting radios

-          DGPS signal (where does it go?)

l         Signals

-          Coast Guard Signal (Beacon Signal)

-          WASS, Uncorrected signal (test for accuracy)

-          Use geodata marker to conduct ground testing on campus

 

Administrative Business

l         Video Conference in the future with Professor Fagg

l         Check on Local resellers for donation and hosting base station

-          Way Point Technology (Company)

 

Plan of Actions/Tasks

l         Call Trimble and Garmin (Jie and Monica)

l         Visit Neilson for voice software (Monica and Xiaoning)

l         Test Path-finding Software (Jessie and Xiaoning)

l         Meeting Sunday (November 20) at 2:30 pm in SAL

l         Find out the IP address to the correction data online

top of page

EGR 410D meeting Minute November 19, 2004 Friday 1:40-3:00 pm

Team Meeting       Location: EGR 203

Facilitator: Monica  Recorder: Xiaoning

 

Plan of Actions/Tasks

l         Order Trimble Starter Kit using Engineering Credit card), about $700 (Jie and Monica)

l         Visit Neilson for voice software after Tuesday or Wednesday after Lunch (Moincia and Xiaoning)

l         Meeting Sunday 2:30-4 pm in SAL (November 20, 2004)

l         Setup a log for items finalized, items purchased and items arrived (Xiaoning)

l         Type up Class feedbacks for WP1 (Monica)

l         Type up Critique feedbacks for WP1 (Xiaoning)

top of page

EGR 410D meeting Minute November 21, 2004 Sunday 2:30--3:30 pm

Team Meeting       Location: EGR 203

Facilitator: Monica  Recorder: Xiaoning

 

Technical updates

l         Need to reinstall software to reconfigure to RTCM format

l         Get a tour of the base station, take picture for WP2 and mid year report

l         Follow-ups to the inertial system; we’re still undecided on whether to drop this topic all together

l         Incorporating the use of a digital compass (Jessie, Xiaoning)

l         Voice Software in Nelson (Monica and Xiaoning)

 

Plan of Actions/Tasks

l         Meeting Tuesday 6:30 pm in Campus Center Cafe

l         Watch DVD for WP1 (all)

l         Compress DVD video and upload it to the website (XX)

l         Setup a tour for the base station

l         Follow up on CET funding (Jie)

l         Read Midyear report guidelines by Tuesday

l         Fix GIS landmarks currently showing in Atlantic Ocean (Jessie)

l         Setup a time and method for GPS unit error testing

l         Purchased Wireless headset (Jie) by end of the break

l         Purchased Wireless adaptor and Ram (Xiaoning) by end of the break

top of page

EGR 410D meeting Minute November 23, 2004 Tuesday 6:30--7:30 pm

Team Meeting       Location: Campus Center Cafe

Facilitator: Monica  Recorder: Xiaoning

 

Completed Actions/Tasks

l         Monica emailed the person in charge of the voice software and found out that someone new is taking over the responsibility for that. Therefore, we will setup a new time to visit Neilson for the software.

 

Administrative Tasks

l         Monica’s contact number during Thanksgiving: (570)898-6240.

l         Xiaoning’s contact number during Thanksgiving: (415)812-3446.

l         Jie and Jessie will be in Friedmen Apartment during Thanksgiving.

l         Purchases

-          Mike from Trimble may call Jie anytime tonight (Nov 23, 2004 T) or tomorrow by 10 am eastern standard time (Nov 24, 2004 W)

l         Include Bob Newton to our future cc. list for outgoing emails.

l         CET funding for the expansion of Wi-Fi assess points not approved, we need to reconsider the option of using a cellular modem.

 

Plan of Actions/Tasks

l         Meeting with Professor Howe (all, 2:40 pm, November 29, 2004 M in EGR 102)

l         Watch DVD for WP1 (all, by November 3, 2004 F)

l         Compress DVD video and upload it to the website (XX, by Nov 24, 2004 W)

l         Send an update email to Professor Fagg (Jie, by Nov 23, 2004 T)

l         Fix GIS landmarks and juxtapose them on the orthomap (Jessie, by Nov 23, 2004 T)

l         Email to request a tour for the base station (Jie, by Nov 28, 2004 S)

l         Email to request for RTCM file setup (Jie, by Nov 28, 2004 S)

l         Read Midyear report guidelines by next meeting with Professor Howe (all, by November 29, 2004 M)

l         Purchased Wireless headset (Jie, by November 29, 2004 M)

l         Purchased Wireless adaptor and Ram (Xiaoning, by November 29, 2004 M)

l         Find out the price and speed for cellular internet connection (Jie, XX, by November 29, 2004 M)

l         Visit ITS for possible equipment donations (XX, 3:10pm, Nov 29, 2004 M)

l         Visit Neilson Library for Dragon Dictate and Naturally Speaking voice software (MM, XX, afternoon, November 30, 2004 T)

l         GPS unit error testing without any correction signals (Monica and Jessie, December 6, 2004 M)

l         Digital Compass Research (Jessie, XX, by December 6, 2004 M)

l         Discuss landmark width during out next Tuesday meeting (all, Nov 30, 2004 T)

top of page

EGR 410D Meeting Minutes November 29, 2004 Monday 2:45—3:15 pm

Meeting with Prof. Howe       Location: EGR 102

Facilitator: Xiaoning  Recorder: Jessie

Agenda (with comments from meeting):

• Purchases—(Excel list made, bring to meeting Friday)
• Review of progress for Fall—(Team still happy about progress)
• Peer evaluations—(Read over and bring to meeting on Friday)
• Questions—(N/A)

Topics discussed:

• No CET funding for WIFI access points L
• Excel purchase sheet completed—bring to meeting on Friday
• Progress Report, Comments from Prof. Howe:
  • Pathfinding-back on path programming aspect-still being researched
    • Receiver testing today for errors and feasibility check
  • Wireless headset—Bluetooth purchase decided upon but yet to be performed
    • For future project development, keep track of how decisions are arrived at
  • Digital compass—what’s going on?
    • Which direction the user is facing is important for getting back on path, and the beginning of the directions list
  • Project should include a visual chart of components--JM
  • Task load for upcoming week—approved by team
• To date: Hardware investigated and purchases decided upon, software investigated—team members will need time to familiarize selves with software
• Voice-enabled computer lab visit tomorrow (11/30) MM & XX
  • Check out program memory consumption and site license availability
• Testing should begin immediately in Spring
• For Spring: should think about which software pieces will be most troublesome (and schedule extra time), and which components can be eliminated if necessary: digital compass, for example
• Base station and cell modem communication need to be investigated
• Next meeting Tues. at 6:30 in SAL

top of page

EGR 410D Meeting Minutes November 30, 2004 Tuesday 6:30--7:30 pm

Team Meeting Location: SAL

Facilitator: Xiaoning Recorder:Jessie

Agenda:

• Individual Task review
• Progress Review for Fall
  • Software
  • Hardware
  • Integration
• Mid-year Presentation
• :), Good luck

Under discussion:

• Excel purchase sheet updates
• Individual Task Review:
  • XX, JM: Pathfinding—MapObjects uses a developer library-further exploration not recommended
  • XX:
    • WiFi adapters sold out at ITS
      • Looking at purchase of a wireless adapter (B/G), and OS compatibility
    • Expansion ports for laptop should include serial port, USB port, docking port
    • Voice lab visit
  • JM:
    • Error testing-SAL testing indicates WAAS signal and uncorrected GPS signal have comparable accuracies
    • Tracklog taken by MM &JM will be analyzed further
    • Terrasync program used by the Geo dept.—email Bob Newton to ask for demonstration
    • Digital compass research and Visual project chart still in progress
  • MM:
    • Voice lab visit-XX and MM will go separately due to scheduling (Thurs)
    • Also visit with ITS person separately
  • JZ:
    • Email to ESRI rep (later noted that he had already contacted Jon Caris)
    • Emailed Andy Kuether re: outputting RTCM format from the base station
    • Add component to Bob Newton email re: “BOB” GPS unit
• WP recording should be scanned through by Friday
• Mid-year report draft should be done by next Monday, Individual sections should be done Saturday by 10pm for compiling—think about illustrations for your section
  • Mid-year report assignments:
    • MM: description of work done
    • XX: Work remaining
    • JM: Challenges, compiling
    • JZ: Other sections
• J-term programming classes available—the team should brush up their skills

top of page

December

top of page

EGR 410D Meeting Minutes December 2, 2004 Thursday 6:30--7:30 pm

Liaison Meeting Location: Media Services

Facilitator: Xiaoning Recorder:Jessie

 

Topics of Discussion:

• Discussion of errors-tracklog and timing
  • Points gathered at certain times should be in front of specific buildings
  • Analysis of data points-Finish by Thursday
  • PF—should take into account drift due to mis-estimation of where satellites are
  • Testing should be done on different days & times
• Difference between WAAS and DGPS corrections
• SAL testing reveals little difference between WAAS and DGPS corrections
• Receiver unit ordered is not WAAS-enabled
• Possibility of testing and returning units to company
• How to get ‘cold’ GPS correction data on laptop (from base station)
• Should put more energy into testing equipment and software: voice, pathfinding
• Voice Lab visit (XX, JZ)
  • Diagram online à Screen Reader & comment recognizer
  • Complete software package needed, not just separate components
  • PF: Linux programs for screen reading
  • XX: Possible to delete everything else on the desktop, make screen reader easier to use
  • PF: Dragon Dictate: input list of commands
    • Menuing interfaces
    • Other interface for set of fixed phrases or sentences
    • SDK software development kit
  • Two programming levels
    • Goto--Goto--Goto
    • Goto X, X is list of commands
  • Via-IBM competitors software
• Digital compasses
  • XX: determine direction facing—use two receivers in parallel, coordinate system is on a rectangular grid
• Magnetic interference—should do testing
• Compass sensor
• I 2C interface
  • synchronous serial protocol
  • High-speed signal
  • Windows driver for parallel ports -- I 2C
  • Pick Microcontrollers for interface module (available at UMass?)
  • Define serial protocol
  • XX: ready made adapters available?
• Read analog signal directly from sensor
  • Requires programming on student side—laptop data conversion
• Mid-year Report draft due on Monday—return comments by Friday
• Terrasync demo Tues. at 1:00
• ESRI Network Analyst—communication w/company
• Prioritize task list à Voice, pathfinding

12/6 EGR 410D meeting Minutes December 6, 2004 Monday 1:00pm – 3:00pm
Team meeting first and then later with Professor Howe, Location: EGR Lounge
Facilitator: Jessie/ Recorder: Jie

• Completed Tasks
  • Divided Mid-Year Presentation parts
    • Title, Overview, Intro - Jie
    • Motivation, Background, Project Description - Jessie
    • Work Completed - Monica
    • Work Remaining - Xiaoning
    • Challenges, Conclusion, Acknowledgements, Q&A - Jie
  • Jessie made a visual diagram showing how various hardware/software components fit into our project
  • Jie prepared a draft of handout to be used for our Mid-Year Presentation
  • Submitted 1st draft of Mid-Year Presentation soon following this meeting to Prof. Howe and technical liaisons.
  • Kindly requested feedback from them by Friday.
• Discussion
  • Progress Report for the week was comprehensive and good
  • Mid-Year Presentation coming up this Friday!
    • Prof. Howe stated that we need to make sure we get across what the main point of our project to the general audience, since our Proposal Presentation did not convey this point clearly
    • She encouraged us to use visuals
    • Using a video clip of how the user and machine would interact by Xiaoning. This is better than having a live demo, as things can go wrong and it would be difficult to maintain a timely presentation.
    • Xiaoning inquired as to whether RealPlayer was installed on the computer in Neilson (our WP video clip is only compatible with this program). Prof. Howe was not sure and suggested that we just Xiaoning's laptop for the presentation and that the 3 teams going after us should have their PowePoint files stored on it as well.
    • We shared our idea for the handout, which includes the visual diagram of how our system works (created by Jessie). Prof. Howe said that we should only print out 10-15 and make them available at the end to anyone who's interested. It would be too distracting to pass them out before our presentation.
    • We decided that having a giant poster on an easel with this visual diagram would be really effective in conveying our project. We will set this up during the break session.
    • We confirmed with Prof. Howe that for the report and presentation, we chose an appropriate stopping point (yesterday, 12/5). For these two deliverables, we will not discuss progress made beyond that date.
    • Prof. Howe commented on the WAAS/DGPS tests conducted by Jon Caris and his class and said that they confirm that DGPS is indeed better. This should be good news, even if we've ordered a receiver without WAAS and means doing the work to setup the base station.
    • Overall, Prof. Howe thinks that we are on task and should into Winter Break feeling good about the progress we've made.
• Tasks to be completed
  • Jie
    • Revise Handouts
  • Jessie
    • Make the Giant Poster
  • Xiaoning
    • Find an appropriate demo video clip to use for our presentation
  • All members
    • Email Jessie our respective Powerpoint slides by Wednesday, 10pm so that she can compile them.

top of page

12/8 EGR 410D meeting Minutes September 20, 2004 Wednesday 10:00pm – midnight
Team meeting, Location: SAL
Facilitator: Jessie/ Recorder: Jie

• Completed Tasks
  • Jessie combined all of our slides into one PowerPoint Presentation
  • We rehearsed through the presentation. We stopped at the end of each slide to give feedback on how to improve the slide, other info to give while explaining the part and other general suggestions.
  • Xiaoning made a wonderful animation showing the user/machine interaction which we placed near the beginning of the slides.
  • After running through the presentation once, we rearranged the slides into a order which we think makes our project clearer to an audience who've never heard about it before.
  • Emailed Prof. Howe to see whether we could submit 2 drafts to her for feedback before turning in a final copy
• Tasks to be completed
  • Jessie will revise the visual diagram and its corresponding presentation slide and the Giant Poster (add title, team name, project description) by Thursday evening
  • Xiaoning will add "eyes" onto the main character in the animation by Thursday evening
  • Plan for tomorrow:
    • Liaison meeting at 4pm...run through presentation and discuss upcoming tasks
    • Right after liaison meeting, group members will meet briefly to ensure that we are on task and will be ready with our presentation slides and parts by 7pm
    • From 6-7pm, dinner break!
    • At 7pm, we will meet in SAL to update the presentation as necessary with revised slides
    • We will rehearse numerous times timed to ensure that we can present in about 18 minutes
  • Print out Giant Poster and handouts Friday morning
  • Mid-year presentation Friday afternoon!!!

top of page

January

EGR 410D meeting Minute January 25, 2005 Tuesday 4:00-4:30 pm

Team Meeting                       Location: Campus Center Café

Facilitator: Jie                      Recorder: Xiaoning

 Follow-ups on Equipments

• Trimble Lassen iQ package is incomplete and will be refunded. Jie will follow up on this matter. We’ll start researching on appropriate GPS packages from Garmin.
• To Date, we have not receive Network Analyst.

 Plan of Actions/Tasks

• Jie will keep in touch with Mike to get refunds for the Trimble Lassen iQ package. (by the next 1-2 weeks)
• We will attend the entrepreneurial luncheon on January 31, 2005.
• Email Professor Fagg and Jon Caris (Jie, by this week)
• Go over the Comments for the report and the presentation (All, within a week)
• Research on Garmin packages (all, within two weeks)
• Contact Network Analyst (?)
• Friday Meeting (Jan 28, 2005 F) after design class
• Sunday Meeting (Jan 30, 2005 Sun), to be scheduled in the Friday Meeting
• Decide dates to attend the E-team conference in San Diego (individually?)
• Establish new meeting times (all, by Friday)
• Email most updated schedule to the Reach account by Friday (all)

  top of page

EGR 410D meeting Minute January 28, 2005 Friday 4:20-4:50 pm

Team Meeting                       Location: Engineering Lounge

Facilitator: Jie                      Recorder: Xiaoning

 Completed Actions/Tasks

• Email most updated schedule to the reach account

 Follow-ups on Equipments

• Trimble Lassen iQ package is incomplete and will be refunded. Jie will follow up on this matter. We’ll start researching on appropriate GPS packages from Garmin.
• To Date, we have not receive Network Analyst.

 Plan of Actions/Tasks

• Sunday Meeting (Jan 30, 2005 Sun) at 6:30 PM in Hubburd House (All)
• Inquire CS student about obtaining a copy of Windows 2000 (Jie and Monica by Next Monday)
• Obtain Voice interface Software (JM and XX by next Wednesday)
• Install the Operating Systems and Voice Software (JM and XX  by next Friday)
• Jie will keep in touch with Mike to get refunds for the Trimble Lassen iQ package. (by the next 1-2 weeks)
• Buy Garmin unit (Jie by next Wednesday)
• Find a way to export GIS data continuously and automatically from the GIS database (Jie and Monica by next Friday)
• We will attend the entrepreneurial luncheon on January 31, 2005 Monday. (All)
• Email Professor Fagg and Jon Caris (Jie, by this week)
• Research on Garmin packages (all, by next Tuesday)
• Contact Network Analyst (Monica, by next Tuesday)
• Sunday Meeting (Jan 30, 2005 Sun), to be scheduled in the Friday Meeting
• Decide dates to attend the E-team conference in San Diego (by next Friday)
• Establish new meeting times (all, by next Monday?)

top of page

EGR 410D meeting Minute January 30, 2005 Sunday 6:30-7:10 pm

Team Meeting                       Location: Hubburd Dinning Hall

Facilitator: Jie                        Recorder: Xiaoning

 Completed Actions/Tasks

• Jessie had showed Monica and Jie the GIS database files
• Monica and Jie had looked at the output file from GIS database and will investigate into outputting the database continuously and automatically
• Monica and Jie had email the company developing the Network Analyst to request for the software
• Xiaoning had downloaded the Platform Builder to her computer and the Speech SDK Library and started experimenting with them
• The laptop intended to emulate PDA is formatted and all Operating Systems are erased to provide for a clean start.

Items to be Discussed

• www.windowschallenge.com/about_contest.html, please take a look at the contest information and decide whether we want to enter. The last day to register is February 6, 2005
• Possibility of getting another outdated laptop solely to run a modern version of windows, windows 2000 or XP, so that the current one can be solely devoted to the embedded system development emulating PDA

 Plan of Actions/Tasks

• Finalize the meeting schedule with Professor Howe and the Liasons (all, by January 31, 2005 M)
• Contact Liaisons and setup the final meeting schedule with them (JZ
• Request for a PC (desktop or laptop) with high processing power solely devoted to Operating System, Voice Interface and application development, using developer kits provided by Windows (all, by Feb 1, 2005 T)
• Obtain a copy of Windows 2000 from available resources (All, by February 2, 2004 W)
• Request for Visual C/C++, Visual Studio and Visual Basic to be installed on the developing PC (JM and XX, by Feb 1, 2005 T)
• Prepare a simple test version of Windows CE build with Platform Builder on the Targeted Laptop (the old one obtained from ITS (JM & XX, by Feb 4, 2005 F)
• Prepare a simple test version of speech application, with the function of recognizing speech and screen reading in the Windows CE OS using Platform Builder (JM & XX, by Feb 11, 2005 F)
• Install Windows 2000 on the laptop (JM and XX, by Feb 9, 2005 W)
• The team will all look at the Platform Builder and get familiar with using it, since features can be add through it (all, by Feb 6, 2005 Sun)

top of page

February

 

Missing minutes for 1/30 - 2/8. Xiaoning, please post them here.

 

top of page

EGR 410D Meeting Minutes February 8, 2005 Tuesday 3:30-4:00 pm
Meeting with Prof. Howe                       Location: EGR 104a
Facilitator: Monica                                  Recorder: Jessie

Topics discussed:

• Future meeting times with Prof. Howe fixed at Tuesdays from 3:30 to 4:00
• Prof. Howe's comments on Progress Report:

• Have teammates read over report for grammar and spelling before submission
• Points should be worded as conrete tasks--do not say "Explore
• Are we scaling down too much?  (If components are going to be added when 'baby         steps' are complete, we should say so.
• Include 'Spring Task List' in next report

• Naturally Speaking and Dragon Dictate
  • The program function is mainly speech recognition
  • The command recognition functions are limited and not useful for our project
  • Currently working with an API (open source program) and talk-back program to construct a useful short code to open and run programs
  • the speech library will be limited to certain commands in order to speed up processing and make GoPAL! user-friendly
• WP2
  • Gear presentation towards technical content, get into details
  • MITRE will present first in SAL, Reach second, then snack break.
  • Computer presentations are okay, lists of bullet points are not     
• Network Analyst
  • The ArcView program is being learned--the campus geodatabase will have to be modified in order to be useful
  • The ESRI development team have asked us to be 'unofficial' beta testers
• Laptop purchase-set for today, Dell computer, to be donated to program when finished
• Project modified overall to start with simpler steps and accomplish more

top of page

EGR 410D Meeting Minutes February 10, 2005 Thursday 4:00-5:15

Liaison Meeting                       Location: Media Services, Video Conferencing Room

Facilitator: Monica                        Recorder: Jessie

 

Topics discussed:

• Progress Update

  • Network Analyst software on the way

  • NCIIA conference-team will be in attendance, but will not present

  • external antenna of GPS receiver broken-hopefully the warranty will cover a replacement

    • PCMCIA adapter ordered

  • Laptop ordered from Dell

  • SAPI and SDK-working with two basic programs to modify code

    • plan to implement restricted grammar--> limit library of recognized words so that all speech leads to command recognition

    • will test system first to make the program open and run, then perhaps work on 'training'

    • "Menu" program downloaded from Microsoft, which opens window in Windows

  • Increment goal for Network Analyst (ArcView)~set start and end point on campus path, find least 'costly' route

  • read out a set of instructions

  • campus pathways layer needs to have lines merged

• PF recommends that we put more work into software interaction and integration--looking at how these programs will work together

• Monica plans to use string codes in ArcView to interface easily with the voice software

• Baby steps:

  • Use speech control to open Network Analyst; test code to open program, run software

  • Monica will look at the source code for ArcView, get line commands to JM & XX

top of page

EGR 410D Presentation Feedback February 11, 2005 Friday 2:00-2:30

WP2                      Location: SAL

Presenters: Monica and Jessie

 

Team Comments:

• Great/good use of visuals/More visuals when explaining the two types of correction techniques would have been helpful and made it clearer/The visual aids were very helpful in understanding the details of the project
• Kept attention
• Clear, well organized information/Explanations were clear and easy to follow for the most part/Clear explanation of shortest path software/Both Monica and Jessie spoke with great clarity/Jessie did a good job of explaining how the GIS system works. Monica was very clear and articulate when she spoke
• Content: Good GPS, GIS info/The team appears to be exremely comfortable with their material and GIS in general.
• Solid presentation overall--stated problem and walked the class through process of plotting grass lines. Truck driver & San Francisco-explained software really well through example and talked about how it fits (into) Team Reach's project--voice enabled, so (you) can tell it where you want to go
• The best part of the presentation was actually going through the example (i.e. find the shortest path in San Francisco)
• Could have set the presentation into the scope of the project more as in where what you showed us will help you move on with your project

Individual Comments:

• Jessie
  • You presented well.  Because this presentation was done without text visuals, it was a little hard to absorb all of the information--I didn't understand GPS and WAAS the first time around, so do use visuals whenever you can
  • seemed slightly nervous, but good introduction
  • watch hands
  • spoke slowly and clearly, but could speed up tempo a wee bit
  • lot calmer when describing the difference between the two systems;good speed, clear
• Monica
  • Very clear and well spoken/clear answers/good volume and clarity
  • good start by building off what MITRE talked about
  • good explanation about the Route finding software
  • When clicking around ArcView it seemed like you were unsure of some of the functions/buttons you wanted to show

top of page

EGR 410D Meeting Minutes February 13, 2005 Sunday 1:00-2:30

Team Meeting                       Location: SAL

Facilitator: Monica                Recorder: Jessie

• Progress Update

  • Dell laptop arrived in the mail.  Username: teamreach, Password: <return>

  • SDK 5.1: Compiled and ran text recognition program

    • ASCII to UNICODE conversoin

    • need text file with Network Analyst instructions for real test

    • Locate where 'Directions' file is stored in memory

    • Find a way to move output directions into input for text-to-speech file

• Topics under discussion

  • The 'old' ArcGIS programs and Network Analyst will be run on the computer in the Sandbox, while the beta copy version will be installed on the new laptop

  • The geodatabase needs to be updated before relocation to the laptop

  • database on SQL server?

  • We will focus on the central area of campus: the intersection in front of Neilson will be the first pathfinding test

  • The 'passing by' function could be represented by a field in the attribute table of the pathways layer

  • The team needs to create a layer containing campus building doorways

• Task List

  • Ask Jon Caris about keys to Bass 105--JM

  • Ask ITS about team copy of Visual Studio (team task, accomplished at meeting)

  • Ask Prof. Howe or Voss about keys to Circuits lab to store laptop

  • Install software components on laptop (Visual Studio, ArcGIS 9.1 with Network Analyst software)

  • Flight information should be emailed to Prof. Howe (done in meeting for those flying out of Bradley)

  • Team members need to create a list of tasks and deadlines by tomorrow

• Individual Tasks

  • JM & XX: Speech to text applications, libraries limitation, limiting vocabulary on engine level

  • JZ: output to Network Analyst, establish SQL server, enquire into receiver antenna replacement, work with receiver when installed (with JM)

  • MM: output directions, become expert on Network Analyst

top of page

2/15 <3 EGR 410D meeting Minutes February 15, 2005 Tuesday 3:30pm – 4:00pm
Team meeting with Professor Howe, Location: Professor Howe’s Office
Facilitator: Xiaoning/ Recorder: Jie

• The current state of the various project components
  • Campus geodatabase (Monica)
    • Monica – Currently updating the campus pathways layer so that all the lines are connected and intersections are present
  • Voice interface (Xiaoning & Jessie)
    • The text-to-speech conversion is working. Xiaoning found the code online for Visual C++ and made minor modifications.
    • Difficulty with speech-to-text conversion. Most code is in Visual Basic, with which team members are not as familiar. They have decided to limit the voice library as much as possible to achieve successful implementation (i.e. just use two words that sound very different)
  • MSAccess (Jie)
    • Only familiar with how to query data internal to the database, but unsure on how to query the Access database with imported data from the GPS receiver. Is this a function built into Access? Will ask technical liaisons on Thursday.
  • Teletype GPS Receiver (Jie)
    • The CompactFlash PCMCIA adaptor card that Xiaoning ordered has been successfully installed. The GPS receiver driver file .inf available online cannot be identified by our laptop, even though the online directions were followed closely.
    • Will ask John and Gregg in the Machine Shop for help on fixing the GPS receiver interface. Meanwhile, contact Teletype about exchange/repair policy since we have 90-day warranty.
• Team’s current goal is to develop Go!PAL to a stage where the below goals are achieved by March 15 th for the NCIIA conference.
  • GPS and GIS integration
    • Latitude and longitude signals from the GPS Receiver will be imported into the GIS database.
    • The GPS coordinates will be compared with the coordinate information of buildings in the database at a defined threshold level using conditional statements to identify which building it corresponds with
  • Smith College database and path-finding software integration
    • Using Network Analyst, a path connecting two points was successfully generated on a San Francisco map included in the tutorial package
    • The Smith database will be modified as necessary so that it is compatible with Network Analyst
  • At the conference, we should be able to present
    • The device’s successful identification of the user’s current location on the Smith campus
    • The device’s successful generation of a path on the Smith campus based upon entered start and end locations
    • Since we will not be able to complete the above demonstrations in San Francisco, we could save files and make a short video of these tasks while we are at Smith right before the conference
• Team members should keep a log of any questions that come to find and email technical liaisons with them as they come to maintain communication
• A good workspace and location to store laptop is the SAL. We will contact Jon regarding access to a cabinet and keys in the SAL.
• To be completed
  • Finish connecting campus pathways in ArcMap (Monica)
  • Find more sample code online for Visual Basic and Visual C++ for speech-to-text conversion, choose one that is simple to modify for our needs (Xiaoning and Jessie)
  • Find out whether Access can import GPS coordinate information (Jie)
  • Contact Teletype to inquire about how to install the GPS receiver driver (Jie)
  • Install Visual Studio on Go!PAL (Jie)
  • Install SDK on Go!PAL (Xiaoning)
  • Waiting for ArcGIS to arrive from ESRI so that we may install it on Go!PAL and test out the new Network Analyst software.

top of page

EGR 410D Meeting Minutes February 17, 2005 Thursday 4:00-5:15

Liaison Meeting                       Location: Media Services, Video Conferencing Room

Facilitator: Jie                         Recorder: Jessie

• Progress Update

  • Programs installed on laptop (Visual Studio, SDK)

  • Voice interface software (SDK) installed on Sandbox computer

  • Beta testing software received from ESRI

  • Campus paths layer cleaned up
• Topics discussed
  • Trigonometry-Curvature of the Earth
    • conversion between latitude and longitude
  • Landmarks display fixed in Network Analyst (beta)
  • GPS receiver-replacement
  • Geodatabase stored in Microsoft Access format
    • Query format--> external info-->into program
    • Program should pull in GPS coords, generate query
    • receiver generates NMEA data format
    • need connection to data stream
    • Open bi-directional file in C
    • Sample code for talking to receiver?
    • SDK -->(download) open device and query device
      • TTM sample: code locations
    • ArcGIS--read data stream, convert to layer, get position
    • ArcObjects developer kit
    • ArcIMS, GPS toolbar-->get code?
  • GPS receiver installed, software needed to read GPS data
  • SDK SAPI 5.1 download
    • Ask PF how to read in random words
      • grammar, how program will fail
      • how to test program (Text to speech)
    • Spech to text program
      • VB vs. VC++?--where is other development taking place
      • Strong visual components, need to be removed
      • how to strip down program
      • send PF sample code--.zip, API documentation
    • No tech support
• Tasks to be completed
  • Programs to be installed on SAL #4
  • Beta software to be installed on laptop
    • Check whether programs can be installed on two machines at once
  • Naming convention for campus paths needs to be worked out
  • Building entrances need to be mapped as a layer and installed as nodes in Network Analyst
  • Need concrete deadlines
  • PF is out of town March 10th: meeting recheduled to Tues. 4-5
  • Sunday-coffee shop program installation (voice software

top of page

2/20 EGR 410D meeting Minutes February 20, 2005 Sunday 1:00pm – 2:30pm
Team meeting, Location: SAL
Facilitator: Xiaoning/ Recorder: Jie

• Administrative tasks resolved
  • Rotations of Facilitator and Recorder positions on Tuesdays
  • Keys (for SAL and the cabinet where Go!PAL is stored) will be stored in our Locker in the Engineering Building
  • The 3 rd Working Presentation will be given by Xiaoning and Jie, hopefully on the requested date of Monday, March 7th
• Tasks completed
  • Monica
    • Connected the lines in the pathways layer that fall in the central campus area.
    • Used Network Analyst to successfully generate a path on modified parts of this layer
    • The directions that were provided were not accurate. The number of turns was incorrect
    • Some of the lines in ArcMap could not be selected or modified, which seems to be bug in the program.
  • Xiaoning
    • Followed the ESRI instructions to generate a Loopback on Go!PAL so that the beta ArcGIS could be installed but could not do so since the key generator is still missing.
  • Jessie & Jie
    • Completed 4 tutorials on using Queries in MSAccess.
    • Then, they constructed a small database of campus buildings specified by latitude and longitude coordinates (that were obtained from tree data). Next, they successfully designed a query to find a building within the database using inputted X and Y coordinates as well as a set tolerance level in units of seconds. These “inputted coordinates” will in the end be dynamically imported from the GPS receiver.
    • Found a module in Visual Basic to calculate distance using the Great Circle Distance, which we will need to determine the threshold radius level around any location (http://www.experts-exchange.com/Databases/GIS_GPS/Q_20383758.html). Will work on adding this to our database.
• Tasks to be completed until March 15 th
  • ** Install Beta ArcGIS ** ASAP (Xiaoning)
  • ** Install driver for Teletype GPS Receiver ** ASAP (Jie)
  • Jessie
    • Program text-to-speech conversion, conduct testing and modify as necessary
    • Figure out dynamic input of GPS coordinates to GIS software
    • Conduct building entrance point collection using handheld receiver
  • Jie
    • Figure out dynamic input of GPS coordinates to GIS software
    • Generate correct directions using the new Network Analyst with the Smith database
    • Decide on a naming procedure for Campus paths and then proceed to assign all paths in central campus with names
    • Choose an appropriate method to store buildings as points (i.e. points along the perimeter spaced out 1m) and do conversions so this data is stored in database
  • Monica
    • Finish modifying the pathways layers
    • Generate correct directions using the new Network Analyst with the Smith database
    • Prepare a new layer of building entrances and conduct point collection using handheld receiver
    • Decide on a naming procedure for Campus paths and then proceed to assign all paths in central campus with names
    • Choose an appropriate method to store buildings as points (i.e. points along the perimeter spaced out 1m) and do conversions so this data is stored in database
  • Xiaoning
    • Text-to-speech implementation using the older Network Analyst-generated directions based upon the San Francisco map from the tutorial
    • Speech-to-text testing and improvements
    • Integration of voice interface with GIS software

top of page

 

2/24 EGR 410D meeting Minutes February 24, 2004 Thursday 4pm - 5pm
Team meeting with technical liaisons, Location: Media Services Conference Room
Facilitator: Jie/ Recorder: Monica

• ESRI will send the hardware key for the Beta version; Jon will keep in touch with them; right now we will only be able to use it on one machine at a time
• GPS antennae was sent back through mail
• Jon will assist us in installing ArcView 3.3 for now, so that we can test the GPS receiver
• SAPI in SAL on computer #4
• laptop will get locked in the cabinet in SAL

Professor Fagg:

• While we wait to hear back from the ESRI, we should install ArcGIS and see how data is streamed from the GPS receiver
• Speech to text:
  • We could try to use the tutorials from SAPI - however, we couldn't figure out what all the inputs of the functions used there mean
  • We currently work from already written code
  • We should try to improve the accuracy by limiting the vocabulary using the grammar interface (this is a good way to improve speed and performance=>through it try to limit focus)
  • Unfortunately, there's apparently no tutorial for this
  • "Coffee" tutorial seems to link the dependencies we couldn't find
    • The module is very visual; we'll have to the out the visual part
    • We've found one program working without visual interface, but we couldn't find the file where we'd specify the grammar
• Check at SDK/samples/utilities -> grammar compiler (specify a very small grammar)
• We have not yet tried to run the grammar compiler (tutorial extremely abbreviated)
• We might have to convert files before they can run (.xml extension to .cfg);
• Network Analyst
• Even if we cannot look at the modules, we might get an idea about how the directions are generated and what the requirements are for the data tables if we make a small subset of paths and test how it works in diffeent simple situations
• After this, we can decide how the different paths should be divided into different objects (sections); we might have to insert "nodes" at windings in order to get the right directions (for all turns)
• First use the new shape file sent by Jon and connect all campus paths (this file should not make problems anymore)

top of page

2/27 EGR 410D meeting Minutes February 24, 2004 Sunday 1pm - 2pm
Team meeting, Location: SAL
Facilitator: Jessie/ Recorder: Monica

• GPS Unit
  • GPS unit works; Xiaoning and Monica went out last night (Saturday) and could use the tracklog function to record the readings; since we did not have the database and layers on the laptop, we could not superimpose our tracklog on the campus map; consequently, we do not know how accurate they are;
• Imported tracklog data onto the SAL computers was projected in the NAD 83 MA Mainland System and superimposed on the campus map; accuracy is not great; looks like we’ve walked as far as the fountain on the other side of Burton Hall and on the roof of McConnel; real path: front of McConnel.
• Only after we can verify the readings are accurate, can we send the receiver back to the distributor, so that we can use it with the antennae
• Jie and Xiaoning took the laptop out on Sunday afternoon and tried to get readings and check the accuracy of the data; they did not get any readings, since the outside temperature was below freezing point, the minimum temperature at which the GPS receiver functions
• Laptop
  • Currently installed on the laptop:
  • Running: ArcView 3.3, old version of Network Analyst, ArcPad
  • Not running: ArcMap, ArcCatalog (the licence allows us to only use them while connected to the network);
  • Jon might have a functional license for the Network Analyst Beta version; will have to be tested on the laptop
• Speech-to-Text recognition:
  • Xiaoning is working on the turorials suggested by Prof. Fagg
  • There is a way to have all actions be recorded in VB so that we would not have to program similar steps from scratch, but use the generated code as a reference
• Database
  • The campus paths layer sent by Jon does not have any issues anymore; 80% of the paths are connected (Monica) and the layer has been tested in Network Analyst, and the shortest path function works well;
  • Jessie and Monica could not make the new entrances layer because the GPS unit does not get any readings when closer than 6 meters to an entrance; a special method will have to be used, in which a Laser Range Finder calculates the difference between the current position of the GPS receiver and the spot on which it is shun (in our case, a building entrance); Jon will show us how to use it
  • A copy of the campus database in access format and orthomaps of the campus are copied onto the laptop
• WP 3
  • Topics will be discussed on the Tuesday meeting with Professor Howe
  • Jie and Xiaoning will be presenting Monday, March 7th; Monica and Jessie will give advice, feedback, and help along the rehearsals
  • A list of ideas provided by Jie and Xiaoning is given
• To do list
  • Push to get a license for individual use of ArcGIS software, since we need mobility!
  • Xiaoning continue working on speech-to-text software following Prof Fagg’s suggestions
  • Jessie and Monica record the building entrances layer (Jessie pick up the Laser Range Finder from Tony Caldanaro (Geology Department, 1st Floor of Burton Hall) on Tuesday)
  • Jie and Xiaoning try installing the beta version on the laptop using the new license
• Arrange meeting with Jon to learn how to use it
• Next meeting time: Tuesday 7 pm on the first floor of Bass Hall to go through the first rehearsal and discussion on WP3

 

top of page

March

 

Missing minutes for 3/1 - 3/11. Xiaoning, please post them here.

 

3/11 EGR 410D meeting Minutes March 11, 2005 Friday 3:30pm – 4:00pm
Team meeting with Prof. Howe, Location: SAL
Facilitator: Monica/ Recorder: Jessie

Topics of discussion:

• NCIIA funding—funds over limit?
• Andrew Fagg-Final presentation, some trip costs covered
• San Diego
  • March Madness
  • Meals covered
  • Dinner reception optional
  • Dress code-business casual
• Revise index card
  • Setup-funding
  • Technical details
  • Go!PAL again
• Prepare elevator pitch
• 200-300 people; faculty, business
• Goals for conference?
• GPS receiver testing
• Plan Post-San Diego
  • Key features to focus on & wrap-up
  • Main components-what can be put off until later
• WP3
  • Transitions
  • Drawing
• WP4
  • Visual
  • Audio for video fine
  • Drawing
• Howe leaving Tuesday morning

top of page

3/15 EGR 410D meeting Minutes March 11, 2005 Tuesday 3:30pm – 4:00pm
Phone call by Jie with Prof. Fagg, Location: SAL
Facilitator: Jie/ Recorder: Jie

• Suggestions from Professor Fagg on preparing for the NCIIA conference
  • Network with college professors and administrators (graduate school options)
  • Network with people in industry (find a job)
  • Bring resume
  • Ask e-teams how they are carrying out user-testing, learn about their projects
• Progress with Jon Caris
  • Installation of Tracking Analyst, which can be used to achieve the real-time collection of GPS positioning data
  • Team obtained manual on Tracking Analyst

top of page

3/17 - 3/19 NCIIA Conference: Dynamic Learning: Changing Models for Changing Times
San Diego, California :)

top of page

3/24 EGR 410D meeting Minutes March 24, 2005 Thursday 4pm – 5:30pm
Meeting with technical liaisons, Location: Media Services
Facilitator: Jie/ Recorder: Monica

• Shared impressions on the conference
  • Most projects displayed had been developed for more than two years
  • Many teams had not considered subject testing as being part of their project, and are leaving this to the companies that would manufacture or market it
  • Many teams had members form different industries and departments other than engineering (psychologists, biologists, business majors)
• Database status
  • Building entrances layer done manually by comparison with the campus buildings layer
  • Layer contains as additional, non-automatically formed fields, the following:
• Name of the corresponding building (when more entrances were recorded for the same building, the names are of the form: “McConnell Entrance Facing Burton”)
• Comments – this field would be able to contain any additional comments that the user might want to check. Currently, some of the objects contain the comment “Limited hours” for the doors that are closing before or right after dinner;
  • Campus paths layer has to be modified to have nodes at each intersection in order for network analyst to be able to make turns (Monica with the help of Jon Caris)
  • Name field has to added for the paths so that it can be tested how directions are generated and whether there are other object fields that network analyst requires to generate them (Monica)
• GPS receiver
  • This was tested using Tracking Analyst in ArcView and real time data was obtained. Due to weather conditions, the testing time was extremely short and only two points were collected. The accuracy of these points has to be verified on the orthophotos. (Jie and Xiaoning)
  • Avenue (programming language coming with ArcView) works with the interaction between Tracking Analyst and ArcView and the data flow between them
  • We do not need to get into the GPS API in order to get the coordinates for the database query. The coordinates are already in the tracklog readings=> we only work with them within the GIS software
  • Not yet sure whether we will have to deal with averaging GPS readings over time in order to avoid false off-path alerts; the tracklog, once beyond the first couple of minutes, settles to a consistent reading; we need to check in the technical specifications of our receiver the time of first fix = time required until it gets the first reading)
• Network Analyst
  • We have previously found no way of automatically outputting the directions through speech, since we do not have access to the script behind network analyst; the directions are automatically outputted in a pop-up window on the screen
  • Using the programming language that comes with ArcView (Avenue), we should be able to establish communication between network analyst and the voice interface
  • Using modules in Avenue, we might be able to output the directions from network analyst into a text file to be read, or feed them directly to the voice software (try to find in the little script visible the section that opens the pop-up window; according to Jon, the “little script” might be enough to provide us with some information on the GIS route object)
  • We need a reference to the route created by network analyst so that we can continuously compare the current position with the route and alert the user in case she gets off-route. Problem: hard to get the reference if we do not have access to the network analyst script
  • Alternative: at regular intervals check where the user is and create new set of directions keeping the same destination. Only output through speech the first set of directions. Constantly check the current position compared with the path she is supposed to be on as in the “directions” text file
  • For getting the user back on track we will use network analyst, since we assume she will stay on a path, even if on the wrong one, and not go on grass.
• To be completed
  • Baby step – make it work: one cycle of the process, namely get the current position, record the destination, find the route, output directions, read the first set of directions, check if the user is on-route, after a while find a new shortest route between the current point and the destination.
  • Make tree for categories of words to be recognized by the speech to text software; to be used for limiting the library of words recognized

top of page

3/25 EGR 410D meeting Minutes March 25, 2005 Friday 3pm – 3:30pm
Meeting with Prof. Howe, Location: Prof. Howe's Office (Xiaoning not present)
Recorder: Jie

• Clarifications
  • Projection problem with ArcPad: incompatible projection settings between the ortho photos/campus layers and the tracklog data. We are now clear on the necessary step of specifiying the projection for the tracklog prior to taking the data, so that it can be shown over the background campus image layers.
  • Avenue is a programming language similar to VB that works with ArcView
• Comments by Prof. Howe
  • Interesting observation about how most of the other e-teams were not focusing on user-testing for their products
  • We should think about which team members can learn which aspects Avenue and divide tasks most effectively
  • For example, Xiaoning can work on outputting the Network Analyst-generated directions from text to speech while others can work on linking the dynamic GPS feed with the geodatabase
• Discussion
  • Ideas for Working Presentation 4 next Friday
• Use still images at the different stages of the movie with notes on how the various function pieces are functioning
• Flowcharts and diagrams are very effective. Perhaps add more detail to the flowchart we developed at the beginning of the school year.
• Chart to illustrate accuracy of GPS receiver
• Image to illustrate accuracies of DGPS and WAAS technologies
• Express the sequential progress of the project through a web. Sart by showing a black box, and with each new image, add another component
• Our main goal: to get one functional cycle of the system working assuming that the user stays on path, the user uses location names within our library, and the speech is translated into correct text
• The next step after getting this cycle working is to make the system ready for off-path conditions. This includes directions on getting the user to a known location that is close and then looping back to the earlier system step where directions are generated based upon the current detected user location and the earlie specified one.
• User-testing will not be the main goal and probably not be carried out this year given that we will be working on the technical developement

top of page

 

3/27 EGR 410D meeting Minutes March 27, 2005 Sunday 1pm – 2pm
Team Meeting, Location: SAL (Xiaoning not present)
Recorder: Jie

• Jessie shared cartoon-like images she drew based upon slides we used for Working Presentation 3
• We decided that it would make the most sense to develop images for WP4 that we would be able to incorporate in our final report and presentation.
• First part of WP4: Design Process Web
  • Handheld PDA -> laptop -> various technology components fly in and out with every click (i.e. GPS receiver, DGPS, WAAS, etc.) -> laptop with GPS receiver, geodatabase, GIS software (including path-finding)
• Second part of WP4: Operation of system
  • Begin with an animation with both audio/video much like the one we used for our Mid-year Presentation
  • Then, show 7 steps of the animation with corresponding "what's going on in the system" diagram explanations
  • Steps are shown below:
  • Turn on/location recognition
  • Destination location
  • Directions generation
  • Output of first part of directions
  • Guiding directions (user stays on path)
  • New directions (user strays off path)
  • Destination reached!
• We will use team time tomorrow in class to decide upon the division of tasks for our scheduled working hours starting this week:
• Monday evenings: Jessie, Monica
  Wednesday evenings: Jessie, Jie, Monica, Xiaoning (sometimes)
  Friday evenings: Jie, Xiaoning

top of page

3/28 EGR 410D meeting Minutes March 28, 2005 Monday 8:30pm – 9:30pm
Team Meeting, Location: SAL
Facilitator: Jie Recorder: Jie

Completed actions/tasks

• We went through with Xiaoning the ideas we came up for WP4

Items discussed

• Division of tasks regarding Avenue
  • Jon sent us an ArcView Avenue Help File, which we will consult
  • Tutorials available through the files provided in the manual
  • Difficult to divide tasks now, since we still need to get a better sense of the capability of Avenue through getting familiar with it. Everyone will go through tutorials and learn about the different functions relevant to her divided tasks.
  • Jessie and Monica: GPS <-> Geodatabase (Buildings, Houses, Route)
    • Once a user has embarked on a route, what does her “current” position need to be compared with so that the device decides whether she is on path or off path
    • This presents problems since it seems as if Avenue is still not the most accessible from a programmer’s point-of-view.
    • Xiaoning suggests that a function might exist which compares the two layers of track log data and route data
  • Xiaoning and Jie: Arcview <-> Voice Interface
    • Speech recognition: User’s voice command (i.e. Neilson Library) needs to be translated into text and then feed into the destination location of the Network Analyst route-finding function
    • Creation of the library of text to be recognized
    • Speech output: Automation of obtaining the text in the directions and translating the beginning portion of this from text to speech
  • This division may not be balanced, since the bulk of the work lies in programming the on-route, off-route functions, which falls under the GPS <-> Geodatabase part. We can reassess and divide tasks again in a week with clearer understandings of Avenue
• Team Reflections
  • Monica: Thanks to Jessie and Xiaoning for giving WP4.
  • Jessie: Will not present in the final presentation (maybe Intro slide), but will participate in the Q&A session.
  • Xiaoning: It’s good that we are looking towards the end but anticipating some frustrations.
  • Jie: Waiting for good weather to test our device (snow and now rain). Realizing that we will be balancing technical progress with final deliverables soon. We should all communication more frequently to teammates so that when we share decisions with our liaisons, everyone is on the same page. We have experienced times when members seem to make decisions single-handedly that do not necessary reflect group opinion. Team will meet for 15 minutes before technical liaison meetings from now on to gather thoughts

Plan of action/tasks

• Jessie and Xiaoning will meet Tuesday afternoon following our meeting with Professor Howe to finalize the sections/slides for WP4 and perhaps figure out tasks that Monica and Jie can help out with
• Team will meet this Wednesday from 8 – 8:40pm to do a run-through of this presentation
• Jessie and Xiaoning will give Working Presentation 4 on Friday
• Xiaoning and Jie will meet Friday evening to work on integrating Arcview <-> voice interface using Avenue
• Team will meet Sunday as a session to experiment with Avenue

top of page

3/29 EGR 410D meeting Minutes March 29, 2005 Tuesday 3:30pm – 4pm
Team Meeting with Professor Howe, Location: Prof. Howe's Office
Facilitator: Xiaoning Recorder: Jie

Completed actions/tasks

• Professor Howe likes our WP4 ideas, which were presented in our Progress Report

Items discussed

• Team Reach wins the Grinspoon Award for Entrepreneurial Spirit! $1000 for the team!
• Invitation to the Grinspoon Dinner next Thursday: competition of best visual display
• We can design a poster to use for this dinner and as our Final Poster deliverable for class

Plan of action/tasks

• Jessie and Xiaoning will deliver WP4 on Friday
• They will meet this afternoon following this meeting to finalize the sections/slides for WP4
• Team will meet this Wednesday from 8 – 8:40pm to do a run-through of this presentation

top of page

 

3/30 EGR 410D meeting Minutes March 30, 2005 Wednesday 8pm – 8:40pm
Team Meeting, Location: SAL
Facilitator: Xiaoning Recorder: Jie

Completed actions/tasks

• Jessie and Xiaoning practiced WP4 in front of Monica and Jie, whom provided feedback and suggestions.

Items discussed

• Comments for Jessie
  • Nice animated web!
  • Don't say "accurate positioning" in reference to GPS
  • Explain that the digital compass was only needed to identify the starting orientation of the user (the direction she's facing). Since this is one small component of the project, we decided to focus our energies on the more essential elements of the project (i.e. voice interface, dynamic collection of GPS data, path-finding)
  • Mention that Tracking Analyst and Network Analyst are both extensions of ArcView, which falls under GIS software
• Comments for Xiaoning
  • Whimsical and pretty animation--very entertaining!
  • Explain our most basic goal for the year to developing a functioning loop without considering the off-path situation and then how we would explore that situation time-permitting
  • Do not focus on the preferred route situation, since it is not a main component of what we are currently working on
  • Include a flowchart explaining the logic of what's going on to reinforce the situation to viewers

Plan of action/tasks

• Jessie and Xiaoning will complete and revise the powerpoint slides for WP4 and practice their parts for WP4 on Friday

 

top of page

4/4 EGR 410D meeting Minutes April 4th, 2005 Monday 5:40pm – 6:15pm
Team Meeting, Location: Tyler Dining Hall
Facilitator: Xiaoning Recorder: Jie

Completed actions/tasks

• Directions generation by Xiaoning and Jon. Using an Avenue script in Arcview, they outputted the directions for a calculated route to speech. With a pre-specified layer of two points (the current and final destinations), the shortest route between the two is calculated and text directions are generated automatically and appear on the screen. They are then stored and saved in a text file, and read aloud by the voice software.
• Testing by Xiaoning and Jie with Tracking Analyst and ArcPad on Friday. The goal was to collect and log data at real-time with Tracking Analyst so that it may be compared with coordinate information of objects in the geodatabase. After spending more than an hour with the testing unit, only two points visible in Arcview was collected. Viewing this data magnified reveals multiple data points. This needs to be analyzed in further detail to determine whether this data represents locations on campus. Projections . They attempted to collect data on receiver accuracy using ArcPad but were unable to save the track log, even though their positions were visible and map projections were set correctly.
• Monica determined that we could not activate the function in Network Analyst that would take in locations by their address. Instead, it would only work if clicked on points on the map. Tthis is problematic, since the destination location will be identified in an address form (i.e. Building name). She experimented with the San Francisco map, since route generation by the "address function" worked for WP2. Monica and Jon determined that some connections have to be added to the San Francisco shape file in order for it to be used to do address-based route generation. She made necessary changes, and the file is now working.
• The voice library text file of campus buildings in our testing area completed by Jie. The names are organized in alphabetical order with ";" as dividers. The major possible ways one could refer to a building were included in this file (i.e. bass hall, bass). This list organized as it is will faciliate more more accurate and convenient speech-to-text conversion. The first letter of a voice command will be identified, and a search in this text file will be conducted by that letter.
• Also, all this progress with directions generation has been made with the a map layer of streets for San Francisco provided in an ArcView tutorial, which is more developed than our campus pathways layer. We need to store addresses in the table format. Jessie: campus pathways development and naming

Items discussed

• For further development of directions generation, the next step is to form an initial layer of two points automatically by address, instead of having to do it by hand and feeding it to a script function
• Confusion over our Sunday meeting due to Daylight Savings time and lack of communication led to our team not holding a meeting. Addition to Team Charter: In the case that a member does not show up to a team meeting, the present members must try to contact her to find out why she's not present and/or remind her of the meeting. The reasons for this include that it is very difficult to schedule meetings that work for our schedules, and if we can try our best to ensure that everyone is present, then it's better for everyone. Sometimes members accidently forget a meeting or confuse the meeting time, and simply need to be contacted to show up.
• Clarification of the scope of our testing area, which helps us determine the amount of work in naming campus pathways and voice library text files. We decided to include the Central Campus area as well as Bass and McConnell in the Science Quad.
• WP4 Comments
  • Start slide
    • use looping and make sure it flows, use text that is understandable without explanations, too much detail, make Go!PAL larger, need greater balance (empty on top, bottom crowded)
  • Design Process Web
    • only show components that we are using now. Perhaps include short text as each component comes up to explain it, use straight arrows
  • Image of Go!PAL user and Animation
    • correct spelling of cemet"e"ry, directions do not tell the user to turn, a little busy, label device components on Go!PAL , make the person bigger
  • Flowchart and maps
    • show simplified case first and then future expansions, don't use partial images, look off-balanced to the left, make sure text is readable horizontally, too busy, add titles for slides (building entrances), doesn't always make sense to use both flowchart and campus map on one slide, use labels for the maps, flowcharts hard to follow (use more spacing and arrows), make everything easy to see (color, size considerations), option for user to specify and save preferred option?, no skewing of images, better quality images (no pixellation)

Plan of action/tasks

• Monica and Jessie will work on preparing the poster. A draft will be completed the latest by the Tuesday meeting with Professor Howe prior to the Grinspoon Banquet.
• We will all start to document any progress we make in group or individual sessions with short email summaries to the rest of the team. Thanks to Monica for starting the trend.
• Name campus pathways
  • what had to be done: network analyst has to be told for each path object where to look for the addresses on the left and right sides of each of the extremes of each street. in the table associated with the san fran shape file, they are already stored in four columns, corresponding to the street numbers: left side of one end, right side of the same end, left side of the other end, right side of the other end. i only had to link those to the corresponding variables that the "address function" is looking for. it's very easy to do that, even if it sounds complicated. this means however, that our smith campus shape file has to have the same four columns, with two addresses for each end of a path. we'll have to talk about how we would do that.
• Compose separate text files containing building entrances and campus pathways. These campus layers should each have their own corresponding text file.

 

top of page

EGR 410D meeting Minutes April 12, 2005 Tuesday 3:30pm – 4:40 p.m.
Team Meeting, Location: EGR building
Facilitator: Jie Recorder: Xiaoning

 Items discussed

• Tracking Analyst. At the software menu interface, we cannot use tracking analyst to extract data from the tracking theme layer. However, we can achieve this at the programming level using Avenue script.
• Poster suggestions
  • Too much text. Need more bullets, less text and bigger font
  • Limit each section to one sentence
  • Think of the poster as if it is a PowerPoint presentation
  • Use the one sentence description
  • Spell out the acronym Go! PAL
  • Specify the liaisons’ affiliations, Jon from Smith College and Professor Fagg from Oklahoma University
  • Shorten the background section, place emphasize on that the project was proposed by visually-impaired student (the client)
  • Split Design Process into Design Elements and Design Integration
• GPS unit exchange. So far, Teletype wants us to return our unit for exchange. However, we will like to receive the new units before return the current units.
• Assignment #7
  • Level of end user involvement
  • Decide when direction reading should end and at what pace
  • The future direction of the project. We will need to decide whether we will recruit students to continue the project.

 Plan of actions/tasks

• GIS competition applications (Monica and Jessie by the April 15? Deadline)
• Greenspoon Dinner (April 15, all, meet at 4 p.m. in front of McConnell and Sabin-Reed). Please bring or complete the following
  • Laptop, GPS unit, headphone, antenna cap and all accessories stored in the cabinet in SAL (Monica)
  • Business Cards, Speakers (Jie)
  • Update and print the poster (Jie, Jessie)
  • Poster Stand/Easel, Candies (Jessie)
  • Powerpoint (Jie, Xiaoning)
• GPS realtime location filtering (Jie)
• Name field check, so that we can use them with Network Analyst (Monica, Jessie)
• Speech Recognition program from Coffee02 (Xiaoning)
• Complete respective parts of the final report (All)

top of page

EGR 410D meeting Minutes April 14, 2005 Thursday 3:45pm – 5:00 p.m.
Team Meeting, Location: Media Services
Facilitator: Jie Recorder: Xiaoning

Liaison meeting with Professor Fagg and Jon

• We reported our current status of our geodatabase to the liaisons.
  • Currently, the unit measurements are set in meters and directions are reported in meters.
  • The connection pin for the receiver is broken again, we are in the process of obtaining an exchange
• Professor Fagg will not be attending our final presentation. He has difficulty planning a family trip here due to tight schedules.
• Voice Interface
  •  Jessie and Xiaoning tested the voice recognition program during the meeting. The recognition rate was decent, but the microphone seemed to be too sensitive. For example, a tap on the table translated to "Burton."
  • Professor Fagg and Jon also tested the program. Professor Fagg got a 100% accuracy on his computer with his laptop's built-in microphone while Jon got about over 80% on our laptop.
  • We will need to adjust the volume of the microphone to optimal level.
  • We suspect there may be differences in recognition accuracy for male and female voices and native and non-native voices.

Plan of action/tasks

• Meeting Sunday, April 16, in SAL (all)
• Investigate in tracking analyst scripts to process real time GPS locations (Jie, Monica, Sun, April 16)
• Expand voice recognition options (XX, Sun, April 16).
• Divide tasks for final presentation and report (all, April 16).

top of page

EGR 410D meeting Minutes April 16, 2005 Sunday 1:00pm – 3:00 p.m.
Team Meeting, Location: SAL
Facilitator: Jie Recorder: Xiaoning

(will be updated soon today, here are the information requested)

Items discussed

• Preliminary final presentation format
  • Title / Introduction             
  • Overview                                
  • Video showing Go! PAL          4 minutes, Jessie
  • Background / Motivation
  • Objectives                              
  • Constraints                               4 minutes, Monica
  • Design decisions                       8 minutes, Monica, Jie
  • Results / Deliverables
  • Ethics                                      
  • Future Implementation              7 minutes, Xiaoning
  • Conclusion                               1 minute, Jessie
• Final Celebration. We will like to  invite Jon and Professor Howe for a meal. We will set the schedule at later meetings.
• Professor Fagg will not be able to visit us.

Plan of action/tasks

• Final report parts. (all by Thursday evening 7 p.m.
  • Design / Decision making process. (all for corresponding parts)
  • Final results /deliverables with user manual (XX)
  • Ethics (Jessie)
  • Future implementation (Monica)
  • Read over and update the remaining parts (Jie)
  • Final report editing starts at 7 p.m. Thursday with Jie and Jessie. Monica will edit from 10 – 1 a.m. Xiaoning will join Monica at 11 p.m. and possibly participate partly during the first session.
  • Final report draft due Friday
• Email Merrilyn Lewis to setup practice presentation schedule (Jie)
• Presentation practice times
  • All presentation slides will be emailed to Monica by noon on Thursday, April 2.
  • Team practice time, thursday, April 21, from 2 - 4 p.m. in SAL
  • April 25, 2005 W, 5-6 team practice.
  • April 25, 2005 Monday with Merrilyn Lewis if possible
• Check out video equipments to make the final presentation video (Jie)
• Email Merrilyn Lewis to setup practice presentation schedule (Jie)

top of page

Plan of action/tasks

• April 22, Friday, 4-6 p.m. in SAL, practice presentation.
• April 23, working on Go! PAL
• April 26, Tuesday, presentation run through 3-4:30 Professor Howe, 4-5 Professor Fagg.
• April 28, Thursday, 12 -1 p.m., final presentation run through, EGR 102
• April 28, Thursday, 4-6 p.m. poster session.

   Make invitation list:

1.  Paul Atena

2. Loren Roch.

3. Bill Kregar

4. Aisha Gabriel

5. Tash Chemel

6. Julie Tomson

7. Hampshire assistive technology people (lemelson)

8. Bob Newton

9. Andy K.

10. Harold Grinspoon

11. NCIIA guy

12. Hampshire guy / volunteer tester

13. Jie's 3 roommates

14. EGR professors (voss, ellise, doughty, mikic, guswa, riley, cardell)

15. Pfabe's

16. Joe

17. Piotre

18. Christopher

19. Mahnaz Madhavi

20. Lewis (WFI)

21. Niel Anderson

22. Desu

23. Barbara Lattanci. (digital media)

24. Lewis Davis.

25. Sandra D. (advancement)

26. guy took the picture

 

April

Thursday, April 28th (4:30-6 PM in McConnell Foyer) - FINAL POSTER SESSION

Friday, April 29th - FINAL PRESENTATION

top of page

May

Monday, May 2nd 3pm - FINAL REPORT DUE

top of page

Progress Reports

1 9/17 - 9/26	2 9/27 - 10/3	3 10/4 - 10/12	4 & 5 10/13 - 10/25		6 10/26 - 10/31	7 11/1 - 11/7
8 11/8 - 11/14	9 & 10 11/15 - 11/25		11 11/29 - 12/5	12 1/25 - 1/30	13 2/8 - 2/15	14 2/15 - 2/20
15 2/24 - 2/27	16 3/1 - 3/11	17 3/12 - 3/26	18 3/26 - 3/29	19 3/30 - 4/5	20 4/5 - 4/11	21 4/12 - 4/18
22	23	24

Team Reach
Progress Report Week 1
Jessie McCartney, Jie Zheng, Monica Marinescu, Xiaoning Xu
Dates: September 17, 2004 to September 26, 2004

 Completed actions/tasks  

• We were assigned this Design Clinic Project. (9/17)
• Jie was elected Team Leader. (9/17)
• The members exchanged contact information with each other and with liaisons. (9/17)
• We chose the group name, Team Reach. (9/21)
• We wrote a charter, which defined the roles and rotation schedules (in the order of first name) of the facilitator and the recorder, and the duties of team members. It also addressed how to resolve potential conflicts and listed weekly meeting times. (9/21)
  • Mondays, 2:40pm - 3:10, EGR, w/ Professor Howe
  • Tuesdays, 6 pm - 9pm, at Lamont House, team
  • Thursdays, 4 pm - 6pm, at Bass Conference Room, w/ Liaisons
  • Thursdays, after meeting w/ Liaisons (6pm - 9pm), at Tyler House, team (when necessary)
  • Sundays, 12:30 pm - 4pm, brunch at Hubbard House, team
  • Class meeting times: 09/24F(design process), 10/01F(project management), 10/15F(proposal), 10/18M(open), 10/22F(open), 11/01(WP1), 11/15M(open), 12/03F(open), 12/06F(open)
• We held two phone conference meetings with our technical liaisons, Jon Caris and Andrew Fagg. (9/17 and 9/23)
• Everyone on the team registered for a SAL account to access GIS databases. (9/23 - 9/29)
• Everyone read the two GPS introduction booklets provided by Jon Caris. (9/23 - 9/29)
• Monica collected references books from Smith and UMASS to use for writing the progress report. (9/26)
• Monica found a movie about GPS for the team to watch. (9/26)
• Jie found a PDA for the visually impaired by Visuaid (www.visuaid.com). (9/22)
• Jie set up a website for the team with weekly minutes, relevant documents and research links: (http://sophia.smith.edu/~jzheng/EGR410/Index.html). (9/21)
• The team received a tutorial from Jie regarding how to update the team website using Dreamweaver and FTP. (9/24)
• The team decided on the preliminary deliverables of developing a prototype of the GPS navigation system, regardless of size and weight, by the end of the year.
• The team gained a basic understanding of GPS, DGPS and GIS through research and interaction
• The team contacted the Office of Disability Services to make an appointment. (9/24)
• The team began writing the introduction and motivation sections of the Proposal. (9/26)
• The team had an interview with Tasha, the only visually impaired Smith student on campus currently and the girl who originated the idea for the project (9/26)
• We learned that it has been challenging for her to stay on paths and that once off-track, it was difficult to get on-track again
• Clear land marks indicators are most helpful
• She wanted a product that could be programmed to remember paths she would choose
• She uses the speech-enabled software, Jaws and Kurzweil 2000

Items discussed

• We decided to make our device a handheld rather than a wearable computing system, because the former is more compact and has sufficient size for our needs. (9/23)
• Jie will contact Visuaid, Freedom Scientifics and other companies with technologies that might aid us in the design of our product. (9/26)
• We mentioned the possibility of first developing a campus map (assessable through computer, compatible w/ Jaws, to give a virtual tour of the campus) prior to the prototype. This will need an update of the annotation layer of GIS on campus. (9/26)

Items for discussion

• We should determine the needs of the visually-impaired student on campus
• We should determine how to consult with Tasha and how she could help use with the design and testing of the device
• We need to decide on which interfaces to include for our device.
• We need to discuss when and where to test the current signal strength of the GPS navigation systems available.
• We need to research the possibility of broadcasting through the wireless 802.11b interface available on campus.
• We need to discuss the product’s compatibility of software as well as its programmability.
• We should determine if and when we would want to visit UMass and their GPS-enabled Xybernaut MA4 wearable computing unit.

Plan of actions/tasks

• Jie will contact ITS
  • Try to resolve the weekly conference phone situation by establishing a permanent meeting location that allows for off-campus conference calls.
  • Get web space and a team email account
• Jessie will watch Ebay for potential PDAs/GPS receivers that we can take apart.
• We will watch the GPS related movie in Jie’s room.
• Xiaoning will conduct preliminary patent searchs. Xiaoning, with the help of the team, will eventually write up a patent emphasizing its unique quality of designed for usage on the SmithCollege campus.
• Prof. Fagg will open up the access to the UMass website with research materials so that each member can browse or download relevant information.
• We will continue doing research and writing up our individual Technical Background sections for the Proposal. The research is divided as follows (* refers to the main person in charge of that section):
  • Market research – Jie*, Xiaoning
  • GPS, DGPS – Jessie*, Monica
  • GIS and databases – Jessie*, Xiaoning
  • Patents – Xiaoning*
  • Hardware and compatibility – Jie*, Monica, Xiaoning
  • Voice-activated, voice-enabled software – Monica *
• Each member will give short summaries of she has researched on the Tuesday evening meeting in Lamont (09/28). Their respective writings should be completed by Thursday.
• We will sign Team Charters during the meeting with Professor Howe. (09/27)

top of page

Team Reach
Progress Report Week 2
Jessie McCartney, Jie Zheng, Monica Marinescu, Xiaoning Xu
Dates: September 27, 2004 to October 3, 2004

Completed actions/tasks

• We determined a list of deliverables after consulting the other project summaries and their corresponding Fall semester deadlines (9/28).
  • Project proposal (10/8)
  • Market review (10/30)
  • Design features, schematic designs of hardware and packaging (12/10)
  • Project budget (12/10)
  • Cost and benefit analysis, cost estimates (12/3)
  • Mid-year presentation and report (12/13)
  • Prototype presentation (Spring)
  • User guide (Spring)
  • Draft report (Spring)
  • Final poster, presentation, and report (Spring)
• We decided on developing a paper design by Winter Break and working on prototype construction in the Spring. (9/28)
• We made amendments to the Team Charter regarding facilitator and recorder roles and responsibilities. (9/29)
• We were given an overview by Jon Caris of the GIS database and associated programs (located in the SAL) as well as how to operate and record paths using the SAL’s Trimble GPS receiver units. (9/30)
• We decided to organize a Team Binder containing the hardcopies of minutes, deliverables, handouts, research documents, and other useful documents. (9/30)
• We met with Laura Rauscher, the Director of Disability Services at Smith. (10/1)
• Jie and Xiaoning conducted testing on the effectiveness of an interactive menu they developed through simulation of the user/machine communication (10/2)
  • One person acted as the user was blindfolded and given a guiding stick, while the other acted as the machine and recited the menu.
  • Two routes were chosen and used during testing: Sage Hall to the Campus Center and the Campus Center to Hubbard House.
  • Determined that a pedometer and density detector may be necessary
  • Determined that GIS databases must be updated with guiding landmarks
• Jessie and Monica conducted GPS campus coverage testing by walking around campus using the SAL Trimble Receiver to chart areas of coverage and non-coverage (10/2)
  • Determined that coverage was quite comprehensive except at a few building entrances
  • Chartered only the central and lower Elm campus areas, where there is a higher concentration of academic buildings and residential houses
• We finished the first draft of our proposal. (10/3)

Items discussed

• After meeting with Prof. Howe, we decided that our progress would be monitored mostly by the team through reviewing the Project Timeline once a week and the general list of tasks once a month. We will also ask our liaisons for constant feedback on our general direction and rate of progress. (9/27)
• Laura gave us a lot of information by answering our questions. (10/1)
  • Smith provides services to Tasha based upon her needs (i.e. books are scanned and translated into Braille or outputted through audio, van rides)
  • Tasha has a private mobility trainer who may be able to help us in developing an appropriate design
  • In terms of the statistics of visually impaired students at Smith, in the past six years, there has only been Tasha and a blind graduate student
  • The legal documents that provide guidelines for the development of Smith include Section 504 of the Rehabilitation Act in 1973, the Americans of Disabilities Act in 1990 (Title 3) and state building codes
  • Smith has an adaptive computer lab in the library with voice-enabled software, Naturally Speaking and Dragon Dictate. Need to contact Lisa Decaralis for a demo of software.
  • We could find visually impaired people possibly interested in helping out with our system design and/or testing of the prototype
    • Stavros ( Amherst ) – Program run by people with disabilities
    • Public school system in Northampton
    • 5 colleges (UMass, Holyoke Community College )
• We discussed whether it was better to have directions in terms of number of steps or distance units (such as feet or meters).
  • Laura suggested steps as the visually impaired do not have a good sense of a physical unit (10/1)
  • Jie and Xiaoning both agreed on steps after their simulation session, because it is easy to lose track of distance units while steps can be counted (10/2)

Items for discussion

• We must establish how best to model impedances.
  • Temporary and permanent, although the former would be more difficult
  • Use density detector?
• We must determine how to find the “best” paths from location A to location B on campus. (9/30)
  • Prof. Fagg suggested exploring Graph Theory or AI.
  • Jon Caris suggested using a GIS capability to do this.
  • We could interview Tasha and ask about her preferences.
• We must decide which potential contacts and resources we should seek out and develop corresponding specific requests or questions.
  • UCSB Libraries and The Campus Community is developing wearable computing system Geo Department Keith Clarke – can email him
  • Judy Franklin, Assistant Professor, CS Department, Smith College
  • Darryl Hampton, UMass CS Department, Vision Lab
  • Tasha
    • Share designed interactive menu and ask for feedback
    • Ask if she prefers lengths given in steps or distance units?
    • Find out specific routes she takes and most commonly visited sites
    • Ask if she feels the difference between grass and pavement?
  • Tasha’s mobility trainer
  • Mass. Assistive Technology Project
  • AAAS (Rehabilitation Engineering)
  • National Institute for Disability and Rehabilitation Research offer grants
  • Hampshire Lemelson Center for Design

Plan of action/tasks

• Prof. Howe will contact ITS to help us find a permanent space for weekly conference calls, to get webspace on Smith’s server for our website and to open a team email account
• Jessie and Xiaoning will work on making GIS and CAD files compatible
  • Contact Sheryl Obremski at Physical Plant to obtain current CAD files
  • Convert CAD to GIS and add annotation to file
• Jessie and Monica will continue to work on modeling GPS coverage
  • Learn about modeling visibility analysis with Spatial Analysis 3D Analyst
  • Conduct campus coverage testing using Trimble receivers on days of varying weather
• Jie and Xiaoning will continue to contact various companies and read datasheets (i.e. Trimble, Freedom Scientific) concerning hardware components
  • Microprocessors
  • GPS receivers
  • Density Detectors (also consider whether to include in design)
  • Pedometers (also consider whether to include in design)
• Each individual team member will continue to conduct research in her respective areas as necessary
• We will continue searching Ebay for used GPS/GIS receiver units that we can purchase and deconstruct.

top of page

Team Reach
Progress Report Week 3
Jessie McCartney, Jie Zheng, Monica Marinescu, Xiaoning Xu
Dates: October 04, 2004 to October 12, 2004

Completed actions/tasks

• Delivered the proposal and presentation in class (10/08)
• Found a permanent conference room – Thursdays 04:15pm, Engineering Building, Room 102
• We have a group e-mail account

Items discussed

• We will get refunded for printing => keep a folder with hardcopies of major documents (10/04)
• Printing cards for the course were requested by Michele (10/04)
• Regular meeting times from now on Tuesdays are starting 6:30 instead of 6:00 (after dinner) in SAL (10/05)
• Narrated in detail to Prof Fagg the “Wizard of Oz ” experiment carried by Jie and Xiaoning and their conclusions (10/07)

• Need to account for the temporary hazards that wouldn’t be in the permanent GIS database – can be done by the use of a guide dog or cane, except for the cases in which they drastically change the path that has to be taken
• Guiding landmarks need to be included in the database
• Developed the menu interaction
• Measure distance in number of steps and direction through three angles: straight ahead, 45 degrees left or right, 90 degrees left or right
• Prof. Fagg will talk with his visually impaired friends, one of which works in developing technology for the blind, and will provide us with contact information (10/07)
• Told Prof. Fagg about the GPS signal check around campus that Jessie and Monica did (10/07)
• External antennae, worn on top of the head for the handheld device would most probably increase the received signal significantly (there are flat antennae of approx 2x2x0.5 inches) (10/07)
• Problems encountered with receiving a signal at the entrance and along some building might be due to the signal being distorted and lost though the multiple reflections from the surrounding buildings – consider using inertial GPS (10/07)
• Example of wearable computer with the antennae onto the shoulder – what is important is that the body wearing the device does not stay between the signal and the antennae (10/07)
• Each team member will be posting write-ups of their own research online, so that the other members can easily refer to it (10/12)
• From now on, our meetings will rely less on an agenda prepared by the facilitator; we will go once around the table, and each member will describe the progress being done in their research (10/12)
• The timeline will be checked and updated once a week, on our Sunday meetings, when we will also compare our progress (10/12)
• Voice interface/recognition (10/14)
• Speech API (Application Program Interface, with Windows as operating system) has all parts integrated, so we would only have to add the new words and maybe enrich the list of grammatical structures to be recognized (“find…”)
• Research on: Xybernaut MA5 (compatible with voice recognition software), noise cancellation microphones
• Using a simple cell phone like voice input would make the system user dependent, and one would have to retrain the system for each new speaker
• Voice recognition software can use a different grammar at each stage of the interaction, sometimes more limited, according to the situation; it is thus less CPU demanding and works faster
• Using already available hardware components will lead to a not-so-elegant/small size/reduced power consume device, but it might be the only way to reach a reasonable prototype at the end of this school year and have a demo (10/14)
• The navigation database and the software determining the different routes could be installed on a desktop machine that would act as server and send already processed information to the PDA through wireless network (10/14)
• Such a server would have a better storage capacity, would work faster, easier to update
• Currently only few hot spots for the wireless network on campus; we are still waiting for a reply from ITS regarding an official map of these (apparently, it would include: the Campus Center, Neilson, Science and Arts Libraries, Stoddard Hall)
• Using the cell phone wireless network (another frequency) (Verizon, 1.40 KB/s; data exchange with computer; cell phones with digital standards) might be another possibility
• We should consider more closely the specific computation capabilities that our system needs to have because of using a voice interface and GIS database (10/14)
• Lay out the different pieces in the design, and specify what and how they would communicate
• GIS can be carried around on a laptop, but not on a PDA (in Access the campus database has 30MB); which could still be portable (if we cannot use the server alternative)
• We will arrange for a video conference when we have something to show (10/14)

Items for discussion

• Add to the timeline the tasks we did not include before the presentation because of time constraints
  • Include all of Prof. Fagg’s and Jon Caris’ comments
  • All team members think about ways to update the timeline, especially the sections assigned to them (ex: inertial GPS, tasks related to headphone and microphone)
  • Accommodate the more coarse approach to specific hardware included in our project
• Divide the contacts list and assign contacts to each team member
• At our Sunday meeting, we will review our state when compared to the Gantt Chart (and timeline)
• Discuss a possible reassignment of team members to different areas (more on software; keep mic/phone for hardware)

Plan of action/tasks

• All members of the team should get personal Sophia accounts; procedure: put the request in through the User Support Center (x4487) by Friday, October 15 th
• Send Prof. Howe e-mail for EGR lab key
• Get in touch with media services to arrange the form of payment for our conference
• Meet with Tasha and share with her the findings of the “Wizard of Oz” experiment and get her comments (Jie and Xiaoning are not skilled in using the cane, and might have encountered other types of problems that a visually impaired person would)
• Test the difference in signal receiving for different positions and heights of holding the Trimble receiver (antennae); account for the frequencies used; also test the signal when using an extra antennae
• Jessie will work on adapting the Smith map to our project, task which in out timeline is listed for the time interval 10/2 – 10/16; in case she needs help, she will contact other members of the team
• Look for the network analysis program already existent on campus (related to the distributed software PDA-base)
• Resend an e-mail to ITS: official map of hot spots on campus; plans for expansion; explain to them why we need it

top of page

Team Reach
Progress Report Week 4 and Week 5
Jessie McCartney, Jie Zheng, Monica Marinescu, Xiaoning Xu
Dates: October 013, 2004 to October 25, 2004

Completed actions/tasks

• All members had setup Sophia accounts.
• We will group email account from now on.
• The team met with Jon Caris and Eric Brewer (Oct 21), and decided to use the current Smith Server for geodatabase storages.
• The team submitted the revised proposal in class (Oct 25 Friday).
• Jie and Xiaoning obtained a used laptop from ITS (Specifications: P2 266 MHz Processor, Win98, 64 MB RAM, Ethernet connection enabled)
• Jie and Xiaoning sent an thank you email to Bill from ITS (Oct 14 Friday), who donated the laptop.
• Jie and Xiaoning emailed Tasha to setup an appointment this coming weekend (Oct 24 Sun)
• We have obtained information regarding WiFi wireless internet coverage in the Science Quad from Eric Brewer in terms of number of assesses point.
  • (1 in the Science library, 1 on Bass 1st floor, 1 on the Engineering building 1st floor, 1 on McConnell 1st floor, 1 on Burton 1st floor, 1 on Burton 3rd floor, 2 on Sabin-Reed 1st floor, 1 on Sabin-Reed 2nd floor, and 2 on Sabin-Reed 4th floor. Out door coverage remains spotty around these buildings, but the coverage area will expand))
• The team amended the new Tuesday meeting schedule (6:30 pm in SAL) and Sunday meeting schedule (1 pm in SAL) into the charter. The original schedule is 6 pm in Lamont Dinning Hall on Tuesdays and 12:30 pm in Hubburd Dining Hall for Sundays.
• We compared our progress to the Chronological Task List and made appropriate adjustments.
  
   

Items discussed

·        Improve our presentation base on Professor Howe’s comments. Jie is thinking about recording future practice presentations.

·        Add password protection to the team website, allowing only specific users to assess it

·        Redistribute the research area to fit our new design process; the focus will shift to software integration since we are building the prototype from a laptop.

·        Adopting the existing GIS server. The server is jointly maintained by Smith Science Department (Geography department, SAL) and Northampton Department of Public Works.

·        We’ll install the software on both Windows 98 and Windows CE operating system on the laptop. The laptop hard drive will be partitioned to allow dual boot. Under windows 98, all software is expected to run normally. Win CE may need some programming, but it resembles the interfaces used on PDAs.

·        Communication between the DGPS server and the navigation device’s receiver. Option 1: the server broadcasts a signal to the receiver. Option 2: the navigation device connects to the server through wireless internet to read the correction data. Jie found a website published by a Californian that provides methods to synchronize the corrections between the server and the receiver.

·        WASS vs. DGPS. These two forms of the correction signals can greatly increase the accuracy of positioning the user. WASS alone offers an accuracy of up to 3 meters. We need to determine whether to use one of the two corrections or both of them.

·        Decide how often synchronizations need to be done. The current Smith Server processes the DGPS data and compiles the information and output it as a file every hour. We’ll need to decide how often this process has to occur to ensure on-time synchronization with our navigation device.

·        Choosing between WiFi (~5 MB /s) and WAP. Eric Brewer anticipated future assess point expansion with WiMax, a successor to WiFi technology on campus, that allows one antenna to cover the entire campus. Intel announced to integrate the WiMax technology into their chips by February 2005.

·        Path-finding software. We’ll store user path in two ways. One is to walk the path once and while tracking it with the GPS unit and store the route as a path. Another is to find/write and algorithm that finds the safest path between two locations.

·        Sequo server. Sequo server allows multiple users to edit the GIS database on the server simultaneously. This is not necessary for now because: 1) the members can edit it at different times. 2) Jessie is the only one familiar with editing the database, so the other members will play supporting roles, rather than actively engaging in updating the database.

·        Voice Software. Microsoft SAPI enables programming in VB and C++ to customize the voice interface. SAPI 5.1 will be released on Oct 25, 2004 M. Microsoft SDK offers the framework to connect SAPI to target software. Voice Insight by VQL Technology also acts as a bridge software between voice recognition software and a targeted software

Items to be discussed with Professor Howe

·        Applying for CET (Committee on Educational Technology) grant to expand wireless internet coverage on campus. We have obtained the information from the meeting with Eric Brewer and Jon Caris during the Thursday meeting, that CET provides funding for Curricular Initiatives in Educational Technology. The application deadlines for this academic year are November 5, 2004 and March 25, 2005. The application must be submitted by a faculty.

·        Budget Binder. We will put together a budget binder to keep track of the expenditures and save all the receipts. Is there a preferred format and how often should we turn request for reimbursements?

·        Copy Card. Currently there are 40 dollars on the card, which can only be used at labs with print stations. What will be user name and password if we want to print documents from labs without the print stations? Is the team going to add money to the print card, or will this be done through the engineering department?

·        Locker. We will need a lock and a locker for our team.

·        Working Presentation Date Request.

Plan of action/tasks

·        Read guidelines for working presentation by Tuesday (Oct 26, all)

·        Incorporate the task changes to the Chronological Task List by Tuesday (Oct 26  Xiaoning)

·        Contact John Crowly and Bill W. regarding how to set password to our website by Oct 27 (Jie)

·        Prepare a budget binder to file receipts and record expenditures by Friday (Oct 29, Xiaoning)

·        Update the website for meeting minutes, progress reports and individual research summary (all). The minutes are to be posted a day before the next meeting and the progress report is a day before the meeting with Professor Howe. Refer to the Chronological Task Line for due dates of individual tasks (the Chronological Task Line is available on the website under Visual Charts).

top of page

Team Reach
Progress Report Week 6
Jessie McCartney, Jie Zheng, Monica Marinescu, Xiaoning Xu
Dates: October 25, 2004 to October 31, 2004

Completed actions/tasks

• Jie and Xiaoning met with Tasha to discuss and update the interactive flowchart menu for users
• Team locker set up at #5 with budget binder and reimbursement forms
• Individual research tasks completed
• Chronological Task list reviewed and dates updated

Items discussed

10/25

• Team lock and locker available
• Adding more money to Print card: Enquire at OneCard office
• Obtain reimbursements using Expenditure Request and Reimbursement form-check Design Clinic fund code and account code for specific task
• Software licenses are available at Smith via group license, so purchase will not be necessary

10/26

• Chronological Task List-updated and changed
• Research and Task updates
  • XX-
    • Complementary devices, pedometers
    • Self-contained or attachable?
    • Working Presentation data set to Nov 12th
  • JM
    • Emails: to Cheryl Obremski about CAD files, from Andy Kuether and Bob Newton re:-base station-now forwarded to team account
    • Pathfinding algorithms online-will research more by Sunday
  • MM
    • SDK &. SAPI: on-campus resources, who to contact
  • JZ
    • DGPS and WAAS-product research

10/28

• CET funding
• New research sections, JZ, MM, JM, XX
• WAAS & GPS enabled receivers
  • Possibility of running a cable between two units
  • Engines & trackpads
  • Serial port over WIFI
  • Blutooth technology
  • Use TRIMBLE-brand receiver—NNEA format
  • 2 serial-port receiver?
  • Location accuracy- 3 meters deemed reasonable, will ask Tasha Friday at 4 pm
• Prototype: Connections by wire now, possibly wireless in future
• Campus paths in ArcGIS map—to be buffered out to 3m to check system requirements
• Windows CE now loaded on dual-boot computer (donated laptop)
• How does one query the ArcGIS system?
  • GIS layers-paths, roads
  • Shortest path algorithm already in place, needs ‘smart network’ to work with
  • Send latitude-longitude value in, return data: buildings, elevation, etc.
• VoiceAssist for ArcPad—made by Voice Insight Co.
  • VQL applications website examined
  • VQL connects ArcPAD with voice software (as shown in website diagram)
  • Voice Insight is a Xybernaut partner company
• In future, email website links ahead of time
• How will the path-finding system be adapted for ArcPAD?
  • Path-finding: Graphs, nodes, links and intersections
  • Must make network-aware path layer—is this possible for ArcPAD?
  • Must find paths and routes, adapt algorithms for pedestrian path-finding
• Working Presentation rehearsal on Nov. 11 th- videoconference to be arranged.

10/31

• Campus WIFI coverage and GIS map updates
• Individual Research Reports:
  • JZ, XX: Interview with Tasha, JAWS reader, user testing next Tues
  • MM: inertial navigation systems which contain INS/GPS
  • JM: base station books, contact Cheryl Obrymski
  • XX: OS loaded on laptop, RAM requirements
• Working Presentation: videoconference to be set up
• Happy Halloween!

Items for discussion

• Progress to date: Where do we stand?
• Divide the contacts list and assign contacts to each team member
• At our Sunday meeting, we will review our state when compared to the Gantt Chart (and timeline)
• Discuss a possible reassignment of team members to different areas (more on software; keep mic/phone for hardware)

Plan of action/tasks

• Continue working on Tasks listed in Chronological Task List:
• Tasks listed starting today are below:

All Work-through presentation 30-Oct-04 05-Nov-04

MM Research voice-enabled software compatibility with hardware 01-Nov-04 15-Nov-04

MM Research voice-enabled software compatibility with OS, other software 01-Nov-04 15-Nov-04

All Choose voice software to use for the system 01-Nov-04 08-Nov-04

MM, XX Investigate the capability of Naturally Speaking (voice software) 01-Nov-04 05-Nov-04

MM, XX Investigate the capability of Dragon Dictate (voice software) 01-Nov-04 05-Nov-04

MM, JZ Investigate the capability of Speech API (voice software) 01-Nov-04 15-Nov-04

MM, JZ Investigate the capability of Microsoft SDK (voice software) 01-Nov-04 15-Nov-04

MM, XX, JZ Investigate the capability of other voice software 01-Nov-04 15-Nov-04

• Prepare ideas for Working Presentation meeting
• Jie and Xiaoning have a meeting with Tasha on Tuesday to refine user interface
• Jessie will get in contact with Phys Plant for the latest CAD files, and work on updating the campus map. Other team member tasks are listed above.

top of page

Team Reach
Progress Report Week 7
Jessie McCartney, Jie Zheng, Monica Marinescu, Xiaoning Xu
Dates: November 1, 2004 to November 7, 2004

Completed actions/tasks

• Jessie
  • Researched the different route-finding software available
    • ArcNetwork with ArcInfo (desktop) – Since ArcNetwork is an extension of ArcInfo, we would need to purchase ArcInfo. ArcInfo is a fancier version of ArcGIS. It is only compatible with desktops and would have to be placed on the server due to its high processor/OS requirements.
    • ArcGIS with Map Objects (desktop) – Mentioned by Jon. Need to find out if can be used separately from ArcInfo.
    • ArcPad with Streetmap (handheld) – Has routing capability and comes with database. We select stops and choose routes. We can set impedance values, but program defaults to shortest length. Need to find out if database that it comes with can be modified.
    • ArcGIS Network Analyst (desktop or handheld?) – Coming out in the first half of 2005. Seems to be exactly what we need, but cannot depend on it due to its uncertain release date.
    • ArcLogistics Route – Has routing possibility, but could not find evaluation version online
  • Installed Map Objects on desktop and Arc Streetmap on handheld
  • Had difficulty with getting sample Map Objects and Streetmap files (which were found online) to work, because administrative permissions were needed to access them.
• Jie
  • Read relevant info in Trimble manuals on the base station located at Smith College. Station transmits data in CMR format.
  • Did not find conversion software between CMR and RTCM formats (two DGPS formats outputted by base stations).
  • DGPS receivers enabled to for CMR are considered RKT (Real Kinetic Time).
  • Contacted Trimble about an appropriate model and was recommended Lassen iQ Starter Kit
    • $695 + more for add-ons
    • Very low power requirement
    • Comes with AC/DC converter, car adaptor, serial cable, software, 2 antennae
  • According to Trimble Sales representative, WAAS is generally less effective than DGPS and mostly a marketing scheme. DGPS gives 1-2m accuracy, while WAAS only gives 3m. WAAS signal obstructed with tall buildings, as only one geosynchronous satellite sends it.
• Monica
  • Established a firmer understanding of SAPI, SDK and VQL and their relationship
  • Researched available Inertial add-on component for our system (INS) for the places where our device loses GPS signal
  • Found an INS that can serve as an add-on part
    • Rate Gyroscope GR 203 by Silicon Sensing System Japan (29mm x 29mm)
    • Outputs voltage through pins
  • Concluded that Kaulman filter will not serve as a good model since it only works for linear paths, which does not apply in our case
• Xiaoning
  • Partitioned the hard drive so that the laptop has 3 operating systems (OS): Windows 2000 for file transfer, Windows 98 (the original OS) and Windows CE 2.0
  • Installed trial version of ArcPad on laptop
  • Has found IBM Voice (voice software) which provides speech-to-text. Will investigate further.
  • Contacted Prof. Judith Cardell about helping the team program in Visual Basic (used for ArcGIS)
  • Setup Research page on Team's website for members to post their sections
• Team
  • Reviewed critique summaries of our Proposal presentation given by classmates
  • Each member prepared two items to be addressed in the Working Presentation
  • Had first video conference via the internet with Prof. Fagg in place of weekly phone conference
• Jie and Monica prepared the CET proposal in the specified online format and emailed it to Prof. Howe
• Prof. Howe submitted the CET form on the team's behalf
• Jie and Xiaoning
  • WiFi Testing
    • Walkway from between McConnell and Sabin-Reed to end of Burton first tested
    • From Burton to Bass to Neilson and finally Washburn tested
    • Two kinds of connections discovered
      • Network to Computer (access points established by Smith)
      • Computer to Computer (likely to be established by students with permission)
    • Only part not covered was the Neilson tunnel
  • Testing with Tasha on 2 routes
    • Hubbard House to Wright Hall (common route she takes)
    • Hubbard House to Bass Hall (route we suggested as we wanted to do testing in Science Center where there ample WiFi coverage
    • Original plan was to guide Tasha using Flowchart, but problems with being able to read the menu in the dark and also potential incompatibility with Tasha’s navigation style
    • Therefore, we decided to have Tasha talk us through 2 routes so that we could observe her navigation style, the landmarks she uses, potential difficulties.

Items discussed

• Will not consider using JAWS after demo by Tasha to Jie and Xiaoning (2nd Interview)
  • JAWS is a screen reader which converts all text on screens to voice
  • For our project, converting speech input to commands will be more important, need something with more functionality
• On-site (saved on laptop) vs. off-site storage (GIS server) of route-finding software
  • Software for 2 applications
    • A) Shortest distance when user is off-path
    • B) A-to-B route-generation
  • B) is most important for our purposes at this point, and should be stored in Server due to processing required and size.
  • A) should be stored in laptop to save time and would not required much processing power or hard drive space.
• Inertial GPS is not essential for our project
  • Considering the high cost, very new technology of INS in general, we decided that it is not necessary for our one-year project. INS GPS is usually used in aircraft. If we can include it as an add-on component, time and budget- permitting, then that would be wonderful, but it is not as essential as other parts like just the GPS receiver, for example.
  • When the user is close to a building she doesn’t need great location accuracy
  • The system should guide the user to a point nearby the entrance, not to the door
  • Monica is investigating the compatibility and pricing of INS add-on parts. Jie will ask about INS-enabled GPS receivers at Garmin. Trimble already confirmed that they have no INS-enabled GPS receiver that fits the team's criteria of price and DGPS-capable.
• WiFi connection for project should be set to log on to strongest signal until the signal is lost
  • There is a time delay to switch signals
  • Perhaps we can use API to search for a strong signal
  • Linux software has this capability
  • Lower connection bars mean more background noise and a greater possibility of losing the signal
• Addition of Landmarks
  • Need to establish area of testing and focus on adding landmarks into the GIS database within that area of test
  • We must choose the landmarks to include in new GIS layers
    • Grasslines
    • Fences
    • Walkway edges
  • Two ways to add more info
    • Use Ortho maps
      • MassGIS, which has not been updated for about a year so inaccurate
      • Sharon O. at Phys Plant knows a guy who takes close aerial photos for fun. If we could obtain photos, we could make them ortho using software
    • Field testing using receivers and tracklogger
• Ideal test area would be where Tasha visits regularly and with general high traffic: Hubbard House, College Hall, Art Library, JM Green Hall, Campus Center, Chapin House, Wright Auditorium, Neilson Library, Washburn House and Seelye Hall
• Team planned out Working Presentation (WP) and divided up parts
  • Interactive System Demo
    • W/ team member (Monica acts as blind person and Xiaoning acts as machine) (3 min)
      • Gives class overview of what our device does
      • Allows class to see potential problems with system
      • Monica will take a pre-planned route across the class from Start to End passing 2 “landmarks.” She will purposely go off-path so we can show alert function.
      • Xiaoning must walk by Monica closely while giving commands so her voice does not indicate location and simulates worn or handheld device
    • W/ class volunteer (Find volunteer to be blind person and Jie acts as machine) (6 min)
      • The volunteer will be guided with instructions down a different route from the first demo
      • Blind-folded and turned in circles several times before beginning.
      • From Start to End, she will pass 2 “landmarks.”
  • Flowchart Discussion (Jie) (10 min)
    • Discussion on how to improve upon current one designed by team.
  • Technical Overview (Jessie)
    • Board drawing of the basic components of system with arrows: device, GPS, DGPS base station, server with GIS database and route-finding software and user (to show voice interface) (5 min).
    • Show and tell of the actual laptop. Basic explanation of hardware and software that needed to be added (5 min).
  • User-testing (Monica) (10 min)
    • How to go about testing users to evaluate system?
  • Communication (Xiaoning) (10 min)
    • How to make communication better between Phys Plant, Public Safety, Noho DPW and us?

Items for discussion

• Whether to purchase a WAAS-enabled receiver. If affordable and already a part of a unit, it would add a level of integrity to signal. But this is not as essential as DGPS, because DGPS is more accurate.
• Whether to use an add-on INS component. Depends on price, size and compability once specific products are identified.
• Depending on response from CET and whether they can implement new access points this year, we may have to modify ideal test area.

Plan of action/tasks

• Jessie
  • Call Physical Plant about the CAD files
  • Update those files
  • Need to find out if Arc Streetmap database that it comes with can be modified.
  • Can Map Objects be used independently from ArcInfo?
  • Prepare the Technical Overview part of WP
• Monica
  • Email or call Silicon Sensing System Japan to find spec sheets for INS part
  • Prepare team demo of interactive menu and user testing sections of WP
  • Decide whether handout is necessary, will prepare if necessary
  • Ask UMass Comp. Sci students and professors for more info on SAPI, SDK, VQL
• Jie
  • Buy phone card
  • Email MassGIS to find out how often ortho photos of Smith are updated and whether they can take a current photo
  • Call Trimble
    • Ask about CMR and RTCM conversion
  • Call Garmin
    • Ask about CMR and RTCM conversion
    • INS GPS products?
  • Prepare volunteer demo of interactive menu and Flowchart discussion of WP
• Xiaoning
  • Get in contact with these professors regarding path-finding software.
  • Look for WiFi adaptors on Ebay and will purchase one
  • Email to make appointment to test Dragon Dictate and Naturally Speaking with Monica
  • Find ArcPad serial number
  • Prepare team demo of interactive menu and communications sections of WP
  • Decide whether handout is necessary, will prepare if necessary
  • Email ITS to find out more about WiFi logon
• Team will do a practice Working Presentation for Prof. Fagg and Jon Caris this Thursday for 30 minutes during weekly meeting time using video conferencing. A meeting will continue afterwards.
• Team will set dates for writing protocol for human subjects testing to be submitted to the IRB board. Need at least two weeks notice for approval.
• Team will prepare questions to ask Prof. Fagg’s friend at the U. of Minnesota who runs the disability services program and can provide feedback on technical problems.

top of page

Team Reach
Progress Report Week 8
Jessie McCartney, Jie Zheng, Monica Marinescu, Xiaoning Xu
Dates: November 08, 2004 to November 14, 2004

Completed actions/tasks

• Working Presentation 1
• Contacted MassGis and found out the next time current orthophotos will be available is in Spring => need to record landmarks “by hand”

Items discussed

• By the end of this semester, we should have a list with all components to be purchased, and be sure they are compatible; we can purchase part of them before and part immediately after Winter break
• March (during Spring Break) NCIIA conference in San Diego
• Prof Howe will see if there is any paperwork we need to do
• Check online at www.nciia.org (9 th annual conference)
• “Tasha talked us through two routes” means

• Said what landmarks she was reaching while walking
• Pointed out problem areas (bifurcations of the path in front of Neilson Library)
• Xiaoning contacted Prof. Pau Atela about path finding algorithms

Items for discussion

• Did we meet our goals with the WP?

• Help with constructing our flowchart; get input on possible improvements
• Provide out classmates with a better understanding of what the progress involves in terms of technical background and functionality of final product
• Get them brainstorm about other ways to reach possible users from the community and get more people to test our prototype
• Check task list and how we are compared to that
• Use feedback from WP1 to update task list

Plan of action/tasks

• Monica and Xiaoning test Naturally Speaking and Dragon Dictate software on campus
• Xiaoning contact UMass Amherst about the association for the visually impaired; buy wireless adaptor card
• Monica – find professor in the Com Sci department at UMass that has working knowledge on SDK, SAPI; get in contact with the Japan company producing the only separate inertial system we’ve found, to see if it would be compatible with the Trimble GPS receiver
• See Prof. Pau Atela on Tuesday at 4:00pm as a team
• Record landmarks across Central Campus using the Trimble receiver on Monday

top of page

Team Reach
Progress Report Week 9 and 10
Jessie McCartney, Jie Zheng, Monica Marinescu, Xiaoning Xu
Dates: November 15, 2004 M to November 29, 2004 M

Purchases/Plan of Purchases

Finalized items to be purchased by the first week after Thanksgiving (by December 3, 2004 F). For upgrades to the laptop, Xiaoning will check with Technical Services for donation first before purchasing. The estimated prices include tax and shipping unless indicated otherwise.

• 802.11b/g Wireless Adaptor (~$25)
• USB and Serial Expansion Hubs (~$15)
• 128 MB DIMM RAM (~$30)
• Trimble Lassen IQ Starter kit, which includes power adaptor, data software, connection cables and external and internal antennas. (~$750)

Items in question

• Mobile internet assess ($19.99 additional to existing T-mobile plans, approximately $25 with tax and regulation fees per month)
• Wireless headset with USB blue tooth sound output adaptor (~$150)
• External battery for the laptop (~$100)

Completed actions/tasks

• Working Presentation feedback review. We have read all the feedbacks and critiques for WP1. The summary of the feedbacks and critiques are available on our website under “Deliverables.”
• Adopt and incorporate suggestions from WP1. Based on the suggestions, we modified our flowchart. The “program a new route” is now a separate category, without verbally outputting the option every time the user is asked to choose a route.
• Other suggestions, such as distance reported in steps and meters, angles reported in degrees and clocks and encryption for the program code will be considered later if we have time, since they are add-ons to perfect an working prototype. We are more worried about geting the unit to work right now.
• Second Meeting with Professor Atela. During this meeting, we did not get information on existing path-finding software, but Professor Atela suggested us to contact the geology department.
• We received more instruction on conducting the error testing for our GPS unit. Method 1. We will choose a path with known “true” coordinates stored in the GIS database. Then, we will walk that path with the GPS unit to create a track log. Then, we will compare the points stored in the track log to the “true” coordinate in the GIS database at different times to determine the accuracy.
• Method 2. When we have more than one GPS units of the same model, we can put them at the same location and compared the difference that each reported.
• Update Landmarks. Using the Trimble GPS unit, Jie and Monica charted the critical landmarks, including grass lines and ground cements on Central campus.
• Jessie transferred these locations to the ArcMap and juxtaposed them on the Smith Campus Orthomap. The charted landmarks from the GPS overlapped well with the Orthomap with reasonable error. We expect the GPS data to be more accurate.
• GIS Day. During GIS day, we have talked to Don Clark; the guest speaker specialized in GPS, and obtained his contact information.
• Another group offered to give us their path-finding analyst software (currently applicable for providing driving directions), we will contact them for more information when necessary.
• WP1 Video uploaded on the website. The compressed video file is 80.4 MB in size, stored in the format of rmvb. To view the video, one will need Real Player 10.
• An update email was sent at the beginning of the break to Professor Fagg since we would not be contacting him for a while due to Thanksgiving.

Items discussed

• Path-finding Software. Two kinds of algorithms are necessary, one for finding the entire route and one for getting back on the route when strayed off. For finding the entire route, we will maximize the use of sidewalks, where pavements and grass lines are, and minimize street crossing. We are likely to find existing software for this part of path-finding necessity.
• For getting back on the route, we will need to do some programming, using GIS database, GPS track and directional data from the digital compass.
• DGPS correction transmission The DGPS correction signals will be downloaded from the server to the laptop through internet using WiFi or the cell modem.
• Then these corrections will be send to the GPS unit through the unit’s input serial port. Either on the GPS unit or the laptop, the time of the correction signals will be screened to filter out the outdated ones. The exact definition of “outdated” in unit of second will be determined.
• After incorporating the corrections, the GPS track log (the path the user walked) from the GPS unit will be transferred to the laptop for further processing.

Items for discussion

• Test for GPS accuracy with different correction signals. We will need to set up times to test for the GPS accuracy with different correction signals and different combinations of the signals. These correction signals include DGPS signals from the Smith base station, the Beacon DGPS signals available in MA and the WASS correction signals.
• Cellular Modem vs. WiFi. CET funding for the expansion of WiFi assess points was not approved, we will need to reconsider the option of using a cellular modem.
• Wireless Headset vs. Wired Headset. We just find out that the Wireless headsets around $60 are too bulky to use. The more compacted ones cost about $150. We will need to discuss whether to buy the more expensive wireless headset or to use a wired one, which is much smaller in size than the size of wireless headset priced around $60.

Plan of action/tasks

• Email to request a tour for the base station (Jie, by Nov 28, 2004 Sun)
• Email to request for RTCM file setup (Jie, by Nov 29, 2004 M)
• Read Midyear report guidelines by next meeting with Professor Howe (all, by November 29, 2004 M)
• Purchased Wireless headset (Jie, by November 29, 2004 M)
• Purchased Wireless adaptor and Ram (Xiaoning, by November 29, 2004 M)
• Find out the price and speed for cellular internet connection (Jie, XX, by November 29, 2004 M)
• Visit ITS for possible equipment donations (XX, 3:10pm, Nov 29, 2004 M)
• Contact Geology Department for suggestions of path-finding software (JM, by Nov 29, 2004 M)
• Email to obtain the IP address for DGPS correction (Jie, by Nov 20, 2004 M)
• Visit Neilson Library for Dragon Dictate and Naturally Speaking voice software (MM, XX, afternoon, November 30, 2004 T)
• Watch DVD for WP1 (all, by December 3, 2004 F)
• Identify doors in GIS database, using Cad files, only one door is necessary per building (JM, December 5, 2004 Sun)
• Test current path finding software (JM, XX, by December 5, 2004 Sun)
• GPS unit error testing without any correction signals (Monica and Jessie, December 6, 2004 M)
• Digital Compass Research (Jessie, XX, by December 6, 2004 M)
• Incorporate the use of a digital compass (JM, XX, by December 3, 2004 F)
• Mid-year Report (all, December 13, 2004 M)
  • Revise the Budget to accommodate changes (XX, December 5, 2004 Sun)
  • Revise Timeline and Tasklist for the Fall semester (all, December 5, 2004 Sun)
  • ReviseTimeline and Tasklist for the Spring semester (all, December 10, 2004 F)
• Mid-year presentation (all, December 10, 2004 F)

top of page

Team Reach
Progress Report Week 11
Jessie McCartney, Jie Zheng, Monica Marinescu, Xiaoning Xu
Dates: November 29 to December 5, 2004

Summary of Work done this week (working chronologically backwards):

The Mid-year Presentation has been outlined, the sections assigned to team members, and work on it will continue Monday through Thursday of Week 12. The Mid-Year Report has been drafted and turned in. On Friday we reviewed the peer evaluations in order to focus on elements that the team needs to work on (individually and as a whole). During the liaison meeting, we discussed with Professor Fagg the requirements for integrating a digital compass into our prototype.

Jie and Xiaoning visited the voice lab on Thursday to familiarize themselves with the technology there. A demonstration of the Terrasync software and ‘BOB’ DGPS unit that the geology department uses in ArcGIS field applications was scheduled for Tuesday, December 7 th at 1 p.m.; Monica will be the main group member in attendance. Jessie and Monica conducted uncorrected GPS signal error testing using a Trimble receiver; the data points they collected were later analyzed for accuracy. At approximately the same time, a group of geology students conducted field-testing in order to compare the accuracy of a WAAS signal and a DGPS corrected signal. They found the difference between the mapped locations using the two different signals to be very slight. The ESRI representative contacted the team liaison (Jon Caris) to report that the process of getting a beta copy of Network Analyst was still ongoing. 

Completed Actions/Tasks:

• Peer evaluation review
• Working Presentation recording viewing
• Voice Lab visit and software review
• GPS signal accuracy testing (see above)
• Preliminary review of digital compass options
• Outside communication: ESRI representative contacted us about Network Analyst, Terrasync and “BOB” demonstration set up with Bob Newton
• Mid year report drafted and handed in for review/comments
• Mid year Presentation PowerPoint outlined, speaking sections assigned, handouts completed—rehearsals and further development this week

Topics of Discussion (From meeting minutes):

• Excel purchase sheet updated:The purchase list will now be updated along with the team budget binder for every purchase made.
• Pathfinding:Further investigation of MapObjects is not recommended for pathfinding applications as it uses a developer library system which would take too long to learn.
• WiFi adapters are sold out at ITS:we are looking into the purchase of a wireless adapter (B/G), and OS compatibility The expansion ports for the laptop should include at a minimum a serial port, USB port, and a docking port
• Error testing:SAL testing indicates WAAS signal and uncorrected GPS signal have comparable accuracies. The tracklog taken by MM &JM is under analysis. The theory is that the data points taken at certain times (recorded on a sheet of paper using a digital watch for measurement) should match up with the data points time-stamped by the satellites in the tracklog.Professor Fagg suggests that we should also research possible ‘drift’ in the signal due to misestimation of where the satellites are, and also that in order to truly test the accuracy of the GPS signal readings need to be taken at different times and dates to experience different periods of satellite coverage.
• Outside Communication:The ESRI representative has been in contact, he is still waiting to hear back from the developers about our copy of Network Analyst. Andy Kuether has written back about outputting an RTCM signal from the base station, but it is still unclear whether or not this is feasible or necessary. The team needs to investigate further. Bob Newton was contacted to set up a demonstration with the DGPS ‘BOB’ unit and Terrasync pathfinding software.
• J-term programming classes are available:The team should brush up their skills in C, C++, and VisualBasic in order to be of greater use in programming applications and custom-designing system interfaces.
• DGPS vs. WAAS: The receiver ordered is not WAAS-enabled, and the team is considering whether or not switching the output of the base station signal is really necessary if a WAAS signal would be just as accurate. The team is also considering ordering more than one receiver unit, if the company is amenable to our testing the units and returning the one which does not suit our needs.
• Laptop testing:We need to consider how to test the correction signal on our laptop, and how to get the ‘cold’ correction data off of the base station website for testing.
• Scheduling suggestions:The team should focus its energy right now on testing equipment and software, specifically voice and pathfinding applications.
• Voice lab visit/Voice interface software:There is a diagram online at the Team Reach website showing a visual diagram of a Screen Reader and comment recognizer. The team should consider that they need an entire software package, not just bits and pieces. Prof. Fagg suggests that there are Linux screen reading programs readily available. XX suggests that it might be possible to delete everything else from the computer workspace in order to make the screen reader easier to use. The Dragon Dictate program requires the user to input a list of commands; there is a menuing interface and another interface for a set of fixed phrases or sentences. The team should look into the SDK software developers kit also. There are two possible programming levels; goto, goto, goto, and goto X, where X is a list of commands. Also worth investigating is VIA, an IBM software competitor.
• Digital compasses:XX proposes to test a system whereby two receivers in parallel are used to calculate the direction the user is facing based on the signals received. Prof. Fagg and Xiaoning suggest that the team should perform some research into the possibility of magnetic interference on campus. Prof. Fagg has also sent a list of suggested sources for digital compass sensors, focusing on a cheap ($20) sensor unit for which the team would need to design an interface module using Pick microcontrollers. The sensor works via an I 2C interface, which is a synchronous serial protocol and a high speed signal. XX suggests that there may be ready-made adapters available. It might be possible to read analog data directly from the sensor, in which case some data conversion programming might be required of the team.
• Mid-year Report and Presentation:The bulk of the meeting on Sunday and the upcoming meetings was/will be spent in discussion, proofreading, and rehearsal for these.
• Necessity of the server:The team now thinks it may be possible to load all of the necessary software components onto a laptop, negating the need for server-based applications.

Items to be discussed:

• Mid-year presentation—individual sections review
• Rehearsals, rehearsals, rehearsals
• Who to invite to Mid-year presentation and rehearsals
• Formal invitations for Jon Caris, Prof. Fagg, and Tasha
• Necessity of the server
• Choosing voice interface software
• Choosing a digital compass application

Plan of action/tasks for upcoming week:

• Mid-year report and presentation (ALL)—this will take up the bulk of the time in the upcoming week
• Agenda for Monday:
  • Review of progress for the week
  • Progress Report comments
  • Discuss Mid-year presentation (do handouts need to be made for entire audience?)
• Review Individual tasks on updated Chronological Task List/Gantt chart
• Outline individual tasks for the final days of Fall semester and winter break
• Purchasing and review of necessary software/hardware

top of page

Team Reach
Progress Report Week 12

January 25 to January 30, 2005

Recorder: Xiaoning

Follow-ups on Equipments

·         Trimble Lassen iQ came in incomplete package because the kit was not fully developed. Therefore, Jie negotiated a refund with Mike, the sales person. Currently she is still working on getting the refund. Due to time constraints, we may have to purchase a new GPS unit from Garmin before we actually get the refund.

·         Garmin GPS unit does not use DGPS correction Signals. Instead, it uses WAASS. Therefore, we will need to worry about setting up an internet connection to obtain the correction signal

·         Network Analyst. To date, we have not received the software yet. Jie and Monica had emailed the company to obtain updated information.

Completed actions/tasks

• GIS database output. Monica and Jie had looked at the output file from GIS database and will investigate into outputting the database continuously and automatically. Although we have not obtained the path-finding software yet, we will need to ensure that we can feed the relevant GIS database out to programs that will process them.
• Contacting Network Analyst. Monica and Jie had email the company developing the Network Analyst to request for the software.
• The laptop intended to emulate PDA is formatted and all Operating Systems are erased to provide for a clean start.
• Windows CE 5.0 Platform Builder. Xiaoning had downloaded the Platform Builder to her computer and the Speech SDK Library and started experimenting with them. The Platform Builder lets the user select the features to include in a Windows CE Operating System. During this stage, programs can be written and included into to the OS. After finalizing the features, a Operating System Image can be build using the build function. Then the OS image can be downloaded to the targeted machine, in this case, our outdated laptop using USB link, Serial link or Ethernet Cable. An emulator is also available to emulate the OS running on the targeted device on the developing computer (in this case, XX’s computer).
• SDK Library. The SDK library is downloaded. The library is used to program voice interfaces.
• SAPI. SAPI is the platform that bridges SDK to program. SAPI is included in the Windows CE 5.0 Platform Builder. The help guide for using SAPI is downloaded. The guide provides instructions and samples for programming using the SDK library, which can be loaded into the Platform Builder.

Items discussed

• Purchasing GPS unit from Garmin. Due to time constraints, we really need to have the GPS unit soon. Our team had discussed and we all agree to purchase the Garmin unit before actually get the refunds from Trimble, if it the refunding process with Trimble takes longer than a week from now.
• Internet Connection. Since we’ll be using WAASS, our unit will not require internet connection to obtain DGPS correction signals. As a result, we do not need to subscribe to any mobile internet services.

Items for discussion

• Requesting for a powerful PC devoted solely to run development tools. Because XX’s computer loads many other programs simultaneously, not all processing power can be devoted to the Platform Builder. This is why our team wants to request for a PC, desktop or laptop, devoted solely for OS and application development, so that we can remove all unnecessary programs that consumes the processing power. Preferably the PC can be located in a restricted lab, so that we can feel comfortable to leave equipments there.
• Requesting for another outdated laptop. Due to the nature of Windows CE, an embedded system, it is more convenient for us if we can use a separate laptop to run it rather than have it coexist with Windows 2000, because as we update the download the operating system from the Platform Build, we risk to rewrite the boot sector each time, which can cause booting problems. We remember that Bill (from ITS) had said that he can easily find another laptop for us if we need it.

Plan of action/tasks

• Finalize the meeting schedule with Professor Howe and the Liasons (all, by January 31, 2005 M)
• Contact Liaisons and setup the final meeting schedule with them (all, by February 2, 2005 T)
• Request for a PC (desktop or laptop) with high processing power solely devoted to Operating System, Voice Interface and application development, using developer kits provided by Windows (all, by Feb 1, 2005 T)
• Obtain a copy of Windows 2000 from available resources (All, by February 2, 2004 W)
• Request for Visual C/C++, Visual Studio and Visual Basic to be installed on the developing PC (JM and XX, by Feb 1, 2005 T)
• Prepare a simple test version of Windows CE build with Platform Builder on the Targeted Laptop (the old one obtained from ITS (JM & XX, by Feb 4, 2005 F)
• Prepare a simple test version of speech application, with the function of recognizing speech and screen reading in the Windows CE OS using Platform Builder (JM & XX, by Feb 11, 2005 F)
• Install Windows 2000 and on the laptop (JM and XX, by Feb 9, 2005 W)
• The team will all look at the Platform Builder and get familiar with using it, since features can be add through it (all, by Feb 6, 2005 Sun)

top of page

Team Reach
Progress Report Week 13
Jessie McCartney, Jie Zheng, Monica Marinescu, Xiaoning Xu
Dates: February 8 through February 15
Recorder: Jessie McCartney

Summary of Work done this week (working chronologically backwards):

        The team has been contacted about the upcoming NCIIA conference; the project is considered still too much in the development stage, so the team will not be presenting at the conference.  We will still be in attendance, though.

        The laptop from Dell has been purchased and has already arrived in the mail. The team has decided that the laptop will be used to test the new beta version software of ArcGIS and Network Analyst, while the slightly older software programs will be tested on a computer in the Engineering and Science Sandbox lab, for which a temporary administrative account has been set up so that team members can install programs such as SDK.  Xiaoning has put together a C++ program using SDK, which runs a speech recognition function for the speech-to-text portion of our device.

        Working Presentation 2 was done on Friday; the main feedback from the class was that the audience liked the visuals and the clear speaking style.

        The Network Analyst software beta version is en route; meanwhile, Monica and Jie worked on familiarizing themselves with the current functions of the Network Analyst application in ArcView.  The GPS receiver has not yet been installed on a computer due to the fact that we were using the wrong kind of flash adaptor; the correct adaptor has been purchased and should arrive this week.

        The team has been in contact with ITS and Julie Thompson about obtaining a copy of Visual Studio for the laptop; Smith College has a site license for the program, and Julie Thompson is going to aid in the installation.  We have also been in contact with the people at ESRI about the campus geodatabase; one of them (Patrick Demer) has offered to look over the campus database and give us advice on how to proceed in modifying it to work with Network Analyst.

Completed Actions/Tasks:

• Working Presentation 2
• Laptop purchase
• SDK text-to-speech application
• Flash adapter purchase
• SDK installation on Sandbox computer
• Network Analyst familiarization and command structure analysis
• Outside communication: ESRI representative contacted us about Network Analyst and critiquing the campus geodatabase, Julie Thompson has arranged to help us install Visual Studio

Topics of Discussion (From meeting minutes):

• Network Analyst: The 'old' ArcGIS programs and Network Analyst will be run on the computer in the Sandbox, while the beta copy version will be installed on the new laptop.  The geodatabase needs to be updated before relocation to the laptop.  We will focus first on the central area of campus: the intersection in front of Neilson will be the first pathfinding test.  Monica plans to use string codes in ArcView to interface easily with the voice software
• ArcGIS: The team needs to create a layer containing campus building doorways.  The campus pathways layer needs to have lines merged.  The 'passing by' function could be represented by comments in a field in the attribute table of the pathways layer.
• SDK 5.1: XX has compiled and run text recognition program that uses ASCII to UNICODE conversion.  We will need text file with Network Analyst instructions for real test.
•  Increment goals: Network Analyst (ArcView) set start and end point on campus path, find least 'costly' route.  Read out a set of instructions in SDK application for Network Analyst. 
• Program Integration: Professor Fagg recommends that we put more work into software interaction and integration--looking at how these programs will work together
• SQL server: Jie is working on establishing a SQL server for a database
• Working Presentation: The presentation went well overall; despite snafus, the team is satisfied with their performance

Items to be discussed:

• Flight/travel plans for NCIIA conference, general plans for conference
• Visual Studio installation—who is in charge
• Key requests:
  • Ask Jon Caris about keys to Bass 105--JM
  • Ask Prof. Howe or Voss about keys to Circuits lab to store laptop
• Installing ArcGIS 9.1 with Network Analyst software on to laptop
• Line code commands for Network Analyst, speech readout using SDK
• Team members need to create a list of tasks and deadlines

Plan of action/tasks for upcoming week:

• JM & XX: Speech to text applications, libraries limitation, limiting vocabulary on engine level, Find a way to move output directions into input for text-to-speech file,  Use speech control to open Network Analyst; test code to open program, run software
• JZ: output to Network Analyst, establish SQL server, enquire into receiver antenna replacement, work with receiver when installed (with JM)
• MM: output directions, become knowledgeable about Network Analyst, look at the source code for ArcView, get line commands to JM & XX, and locate where 'Directions' file is stored in memory

top of page

Team Reach
Progress Report 14
Jessie McCartney, Jie Zheng, Monica Marinescu, Xiaoning Xu
Dates: February 15, 2005 to February 21, 2005
Recorder: Jie Zheng

  Completed actions/tasks

• Administrative tasks
  • Rotations of Facilitator and Recorder positions on Tuesdays
  • The 2 keys for SAL and the cabinet where Go!PAL is stored will be placed in our Locker in the Engineering Building
  • The 3rd Working Presentation will be given by Xiaoning and Jie, hopefully on the requested date of Monday, March 7th
  • Visual Studio software obtained from Julie Thompson
  • ArcGIS 9.1 beta software received from ESRI
• Campus geodatabase and Network Analyst (Monica)
  • Updated the pathways layer in the central campus region so that all the lines are connected and intersections are present. Experiencing problems in ArcMap, where some lines cannot be selected or modified, which seems to be bug in the program.
  • Used Network Analyst to successfully generate a path on modified parts of this layer, although the directions were not accurate (number of turns and direction of turns were incorrect).
• Voice interface (Xiaoning, Jessie)
  • Trials of text-to-speech conversion that confirmed functioning
  • Online search of speech-to-text conversion code in Visual Basic and Visual C++ and corresponding manuals
• MSAccess (Jessie, Jie)
  • Completed 4 tutorials on using Queries in MSAccess.
  • Constructed a small database of campus buildings specified by latitude and longitude coordinates (that were obtained from tree points data) and successfully designed a query to find a building within the database using inputted X and Y coordinates (defined an arbitrary tolerance level in units of seconds).
  • Found via the internet a module in Visual Basic to calculate distance using the Great Circle Distance, which we will use in calculating the threshold radius level around any location.
• Teletype GPS Receiver (Jie, Jessie)
  • Installed the CompactFlash PCMCIA adaptor card but the GPS receiver .inf driver file that we downloaded online still cannot be identified by our laptop, even though the online directions were followed closely.
  • Contacted tech support without receiving helpful advice
  • Contacted customer service and received confirmation that our damaged parts can be mailed to them for repair or replacement.
• Go! PAL (Xiaoning, Jie)
  • Installed virus and spyware software as well as Visual Studio, SDK
  • Tried to install ArcGIS 9.1 beta by following the ESRI instructions to generate a Loopback to imitate a network with one computer but could not do so since the key generator was missing.
  
  

Items discussed

• For the NCIIA conference, Go!PAL should be able to perform the following
  • Successful identification of the user’s current location on the Smith campus
  • Successful generation of a path and corresponding directions on the Smith campus based upon specified start and end locations
  • Since we will not be able to complete the above demonstrations in San Francisco, we could save files and make a short video of these processes while we are at Smith right before the conference
• We are using 2 computers for our project
  • Go! PAL – ArcGIS 9.1 beta with the new Network Analyst, SDK, geodatabase, Visual Studio
  • SAL #4 – Old ArcGIS software, SDK, geodatabase, Visual Studio
• We will mail back the Teletype GPS receiver and external antennae for repair or replacements
• Possibilities on linking GPS output in NMEA data format to the geodatabase
  • ArcGIS software that directly accepts NMEA data (reads data stream, converts it to a layer and finds position)
  • ArcObjects developer kit
  • ArcIMS, GPS toolbar
  • Use Visual C++ to open bi-directional file
  • No sample code provided by Teletype online for converting the NMEA data but a sample of converted data is shown
• Technical challenges of SDK SAPI 5.1
  • How to test the accuracy of the text-to-speech conversion? Prof. Fagg will read our code and provide suggestions
  • Difficult to decipher the meaning of speech-to-text sample code in Visual Basic and Visual C++ found online. The code will be sent to Prof. Fagg for feedback and suggestions on how to simplify (i.e. remove visual elements).
  • No tech support for free

Plan of actions/tasks

• Tasks to be completed by March 15 th, the day before we leave for the NCIIA conference
• ** Install Beta ArcGIS ** (Xiaoning) ASAP
• ** Install driver for Teletype GPS Receiver ** (Jie) ASAP
• Jessie
  • Program text-to-speech conversion, conduct testing and modify as necessary
  • Figure out dynamic input of GPS coordinates to GIS software
  • Conduct building entrance point collection using handheld receiver
• Jie
  • Figure out dynamic input of GPS coordinates to GIS software
  • Generate correct directions using the new Network Analyst with the Smith database
  • Decide on a naming procedure for Campus paths and then proceed to assign all paths in central campus with names
  • Choose an appropriate method to store buildings as points (i.e. points along the perimeter spaced out 1m) and do conversions so this data is stored in database
• Monica
  • Finish modifying the pathways layers
  • Generate correct directions using the new Network Analyst with the Smith database
  • Prepare a new layer of building entrances and conduct point collection using handheld receiver
  • Decide on a naming procedure for Campus paths and then proceed to assign all paths in central campus with names
  • Choose an appropriate method to store buildings as points (i.e. points along the perimeter spaced out 1m) and do conversions so this data is stored in database
• Xiaoning
  • Text-to-speech implementation using the older Network Analyst-generated directions based upon the San Francisco map from the tutorial
  • Speech-to-text testing and improvements
  • Integration of voice interface with GIS software

top of page

Team Reach
Progress Report Week 15
Jessie McCartney, Jie Zheng, Monica Marinescu, Xiaoning Xu
Dates: February 22, 2005 to February 28, 2005
Recorder: Monica Marinescu

Completed actions/tasks

• Flight tickets for NCIIA conference
• Created Task web that shows which tasks are stand-alone, and which ones depend on the completion of other tasks (attached)
• GPS antennae was sent back through mail
• GPS receiver tested with ArcPad on laptop; works fine; accuracy might not be great
• 80% of the campus paths connected; the layer has been tested in Network Analyst, and the shortest path function works well
• Campus database and orthomaps are copied onto the laptop
• Querying the database was successful when using coordinates of the points in the trees layer; random tolerance level used
• Accuracy of recognition for speech to text software slightly improved
• On laptop, currently installed and running: ArcView 3.3, old version of Network Analyst, ArcPad
• On laptop, currently installed, not running: ArcMap, ArcCatalog (the license allows us to only use them while connected to the network)

Items discussed

• GPS Unit
  • imported track log data onto the SAL computers was projected in the NAD 83 MA Mainland System and superimposed on the campus map; accuracy is not great; looks like we’ve walked as far as the fountain on the other side of Burton Hall and on the roof of McConnel; real path: front of McConnel
  • only after we verify the readings are accurate, can we send the receiver back to the distributor, so that we can use it with the antennae
  • Jie and Xiaoning could not check the accuracy of the data, since the GPS receiver does not get readings below freezing point
• Laptop
  • Jon might have a functional license for the Network Analyst Beta version; will have to be tested on the laptop
• Speech-to-text recognition
  • Professor Fagg suggested using the SAPI tutorials
  • Problem: figure out what all the inputs of the functions used there mean
  • The accuracy can be improved by limiting the vocabulary through the grammar interface (a good way to improve speed and performance, i.e. limit focus)
  • Problem: there doesn’t seem to be a tutorial for this
  • "Coffee" tutorial seems to link the dependencies (functions and variables called for) we couldn't find
  • The module is very visual; we'll have to the out the visual part
  • We have found one program working without visual interface, but we couldn't find the file where we'd specify the grammar
  • Check at SDK/samples/utilities -> grammar compiler (specify a very small grammar)
  • We should try run the grammar compiler (tutorial extremely abbreviated), since we might have to convert files before they can run (.xml extension to .cfg)
• Network Analyst
  • Even if we cannot look at the modules, make a small subset of paths and test what directions it outputs for different simple cases
  • Decide afterwards how the different paths should be divided into different objects (sections); (ex: Insert "nodes" at windings in order to get the right directions (for all turns)?)
• Database
  • A tolerance level for each point for database querying can be included by using the correlation between the two coordinate systems and change from degrees to meters
  • Prof. Howe suggested, the “reasonable tolerance" could be taken as the shortest distance between two buildings on campus
  • There is a way to have all actions be recorded in VB and use the generated code as a reference
  • Jessie and Monica could not make the new entrances layer because they did not get readings when close to an entrance; a Laser Range Finder will have to be used (calculated difference between GPS reading and place at which it is pointed)
• WP 3
  • Topics will be discussed on the Tuesday meeting with Professor Howe
  • Jie and Xiaoning will be presenting Monday, March 7th; Monica and Jessie will give advice, feedback, and help along the rehearsals
  • A list of ideas provided by Jie and Xiaoning is attached

Plan of action/tasks

• Try run ArcGIS –Beta version using the new license
• Push to get a license for individual use of ArcGIS software, since we need mobility!
• Continue improving the accuracy of the speech-to-text recognition function
• Next meeting in SAL Tuesday night 7pm to work on WP3
• Jessie pick up the Laser Range Finder from Tony Caldanaro (Geology Department, 1st Floor of Burton Hall) on Tuesday
• arrange meeting with Jon to learn how to use it
• Jessie and Monica record the building entrances layer when weather allows and after learning how to use the Laser Range Finder
• Check accuracy of GPS readings when temperature allows
• Xiaoning continue working on speech-to-text software following Prof Fagg’s suggestions

Find conversion function from one coordinate system to another for tolerance level and querying the database

top of page

Team Reach
Progress Report Week 16
Jessie McCartney, Jie Zheng, Monica Marinescu, Xiaoning Xu
Dates: March 1, 2005 to March 11, 2005
Recorder: Xiaoning Xu

Follow-ups on Equipment

• Booster Antenna. We received the exchanged booster antenna. The lowest operating temperature range is -40 degrees Celsius.

Completed actions/tasks

• GPS unit testing. We tested our GPS unit twice, once without the booster antenna and once with it. The tracklogs we obtained from both did not match well to the orthomap. However, the overall location of the tracklogs seem ed to contain the testing path we walked.
• WP3 Completed. Jie and Xiaoning completed the last working presentation on Monday. The overall comments from the class are positive, noting the clear organization and good explanation skills.

Items discussed

• Voice Interface. Professor Fagg sent us some sample grammar files, showing us the structure and language to specify grammar using XML language. We looked at them to try to understand how to write the grammar files, but we still need to find out how to load these files into our program.
• Building entrances. We will need to use a range finder to gather the points for entrances, because our GPS unit looses signal near buildings. When using the ranger finder, one person will hold a reflective glass at the door way, another person will shoot a laser beam using the range finder connected to the GPS unit to the glass piece, at a distance away from the entrance. The range finder detects the distance and altitude and sends a correction distance to the GPS position, which will give us the position of the building enrances.

Items for discussion

• Network Analyst. Network analyst is an object component used with the older version ArcView 3.3 or the newer version ArcInfo 9.1 beta. Each software has its own problems

We currently know how to use Network Analyst with ArcView 3.3 but not with ArcInfo 9.1. For ArcView 3.3, we cannot directly read GPS feedings into the software. Everytime we initiate our GPS receiver, the software crashes. For ArcInfo 9.1 beta, we can receive GPS data directly, but we must be connected to the SAL network through internet to use the software.

Currently, we are using ArcPad to gather our GPS tracklogs, and then transfer it to the software with Network Analyst. To use Network Analyst in real time, we must obtain GPS readings in real time.

Plan of action/tasks

• Meeting with Professor Howe ( March 11, 2005 F, all)
• Make an index card sized team business card and bring to Professor Howe ( March 11, 2005 F, XX)
• Trip to San Diego. Jessie, Monica and Xiaoning will leave together. Jie will fly our from Boston. (March 16-20 including transportation time, all)

Team Reach
Progress Report Week 17
Jessie McCartney, Jie Zheng, Monica Marinescu, Xiaoning Xu
Dates: 3/12/05 to 3/26/05
Recorders: Jessie McCartney/Jie Zheng

Completed actions/tasks

• Team attended the NCIIA Conference ( Dynamic Learning: Changing Models for Changing Times) in San Diego from Thursday, 3/17 th to Saturday, 3/19 th.
• Team found the conference to be an enjoyable and valuable learning and networking experience. We met other college students and learned about the e-team projects. We talked with various academic professors and administrators from across the country in different disciplines about programs at their colleges. We also attended various lectures and workshops on very interesting topics such as networking, angel investing and natural capitalism.
• Other conference observations:
  • Most team projects displayed had been developed for more than two years.
  • Many teams were working together with startup companies or established companies within the industry under which their product falls.
  • Many teams had not considered subject testing as being part of their project, and are leaving this to the companies that would manufacture or market it
  • Many teams had members and advisors from different industries and departments other than engineering (psychologists, biologists, business)
• Xiaoning and Jessie recruiters several beta testers for Go!PAL from Amherst at the conference
• The website address we used on our business cards is still password-protected, so we have contacted ITS to help us remove that extra security measure and make it accessible to all.
• Modifications made to the geodatabase
  • Monica added a building entrances layer manually based upon an ortho map and the campus buildings layer
  • This new layer contains the additional, non-automatically formed fields, names and comments. The names field correspond to nearby buildings, or of the form, “McConnell Entrance Facing Burton,” when more entrances were recorded for the same building. The comments field can be used for any additional comments that the user might want to check. Currently, some of the objects contain the comment “Limited hours” for the doors that close at a certain time.
• Tracking Analyst: dynamic log of GPS receiver data explored
  • Tracking Analyst, an extension in ArcView, allows for the feeding of GPS positioning data real time. After installing the software, Jie and Xiaoning created a local tracking connection and a corresponding theme to test it.
  • The time they spent outside was very short due to heavy snow, and thus two points were collected. The accuracy of these points has to be verified on the orthophotos.
• The team was experiencing difficulties with projections not being compatible, thus being unable to load any image of the campus as a background when going out with Go!PAL and taking data. Jon went through the process of specifying the projection setting in ArcPad with Jie for both existing ortho photos or geodatabase layers and also the tracklog file of the real time GPS data collection. After making the necessary modifications, she tried testing the system several times but was unable to receive a response, which is probably related to the limited receiver operating temperature.

Items discussed

• Team has discovered the software that will help link our different applications: Avenue, the programming language that comes with ArcView, can be used as the integration and automation software for ArcView (whose important extensions are Network Analyst for path-finding and Tracking Analyst for real time GPS data collection) with other GIS applications.
• We decided that it is not necessary to use the GPS API in order to get the coordinates for the database query. The coordinates are already in the tracklog readings=> we only work with them within the GIS software
• Network Analyst/Avenue developments
  • We have previously found no way of automatically outputting the directions through speech, since we do not have access to the script behind network analyst; the directions are automatically outputted in a pop-up window on the screen
  • Using the programming language that comes with ArcView (Avenue), we should be able to establish communication between network analyst and the voice interface
  • Using modules in Avenue, we might be able to output the directions from network analyst into a text file to be read, or feed them directly to the voice software (try to find in the little script visible the section that opens the pop-up window; according to Jon, the “little script” might be enough to provide us with some information on the GIS route object)
  • We need a reference to the route created by network analyst so that we can continuously compare the current position with the route and alert the user in case she gets off-route. Problem: hard to get the reference if we do not have access to the network analyst script
  • Alternative: at regular intervals check where the user is and create new set of directions keeping the same destination. Only output through speech the first set of directions. Constantly check the current position compared with the path she is supposed to be on as in the “directions” text file
  • For getting the user back on track we will use Network Analyst, since we assume she will stay on a path, even if on the wrong one, and not go on grass.
• Team decided that Jessie and Xiaoning will give Working Presentation 4 next Friday
• Jie addressed the fact that the team needs to work on the project more frequently and collaboratively and also at a set schedule so that we can move forward. We have worked a lot independently this semester on the project and are at a stage now where the best approach is to spend chunks of time trying things out and working together to achieve progress. Therefore, we decided on a new working schedule starting next week to work in SAL on weekday evenings from 6:30pm to 9pm:
  • Monday Jessie, Monica
  • Wednesday Jessie, Jie, Monica, Xiaoning (sometimes)
  • Friday Jie, Xiaoning

Items for discussion

• WP4 Ideas will be discussed at our meeting with Professor Howe and on at our meeting on Sunday.
• We must decide whether we should average GPS readings over time in order to avoid false off-path alerts. The tracklog, once beyond the first couple of minutes, settles to a consistent reading, and we need to check in the technical specifications of our receiver the time of first fix = time required until it gets the first reading as well as do a few basic tests with our receiver and antennae together to determine this figure. If we decide this time is sufficiently short, we will not need to average the readings.
• Whether Avenue in itself is sufficient to meet our integration and automation requirements for the system. We will be able to assess this more accurately after we familiarize ourselves more with this programming language and test it out.
• Whether user-testing is still something we want to keep on our agenda for tasks this semester or whether it is more feasible and smarter to focus on our technical components.

Plan of action/tasks

• Big picture goal for the next few weeks: create one functional cycle of the entire process. Get the current position, record the destination (Speech-to-text, find the route, output directions, read the first set of directions, check if the user is on-route, after a while find a new shortest route between the current point and the destination. The key to achieving this level of functionality is through Avenue.
• Team will meet Monday, Wednesday and Friday evenings starting next week to get familiar with Avenue and being testing it out. It will be used for to link the following:
  • Network Analyst Directions -> Voice Interface
  • Network Analyst Directions - > Compare with user’s actual position in the case she gets off-route
  • GPS Receiver - > Geodatabase
• Jie and Monica will create a “tree” in a text format for categories of words to be recognized by the speech to text software. This will include building and house names, where for each location, several names will be included in the library (i.e. Neilson, Neilson Library). The purpose of this tree is that it will be used to limit the library of words to be recognized
• Weather-dependent, the testing of the GPS receiver accuracy needs to be completed. The device should be held at one constant position for a span of five minutes so that the fluctuation in positioning data can be assessed.
• Jessie and Xiaoning will give Working Presentation 4 next Friday

top of page

Team Reach
Progress Report Week 18
Jessie McCartney, Jie Zheng, Monica Marinescu, Xiaoning Xu
Dates: 3/26/05 to 3/29/05
Recorder: Jie Zheng

Completed actions/tasks

• We decided on two sets of images for WP4 that we should be able to incorporate in our final report and presentation.
• First part of WP4: Design process web presented from the team/designer perspective. The order of technical components that fly in and out would correspond sequentially to our design process.
  • Handheld PDA -> laptop -> various technology components fly in and out with every click (i.e. GPS receiver, DGPS, WAAS, etc.) -> laptop with GPS receiver, geodatabase, GIS software (including path-finding)
• Second part of WP4: Operation of system presented from the user perspective. Begin with an animation with audio showing how our device works. Divide the animation into 7 clips, and show one at a time, and then after each, follow up and elaborate in more detail about all the system components working for that step to be carried out. We will use "what's going on in the system" diagram/flowcharts.
  • Turn on/location recognition
  • Destination location
  • Directions generation
  • Output of first part of directions
  • Guiding directions (user stays on path)
  • New directions (user strays off path)
  • Destination reached!

Items discussed

• Division of tasks regarding Avenue
  • Jon sent us an ArcView Avenue Help File, which we will consult
  • Tutorials available through the files provided in the manual
  • Difficult to divide tasks now, since we still need to get a better sense of the capability of Avenue through getting familiar with it. Everyone will go through tutorials and learn about the different functions relevant to her divided tasks.
  • Jessie and Monica: GPS <-> Geodatabase (Buildings, Houses, Route)
    • Once a user has embarked on a route, what does her “current” position need to be compared with so that the device decides whether she is on path or off path
    • This presents problems since it seems as if Avenue is still not the most accessible from a programmer’s point-of-view.
    • Xiaoning suggests that a function might exist which compares the two layers of track log data and route data
  • Xiaoning and Jie: Arcview <-> Voice Interface
    • Speech recognition: User’s voice command (i.e. Neilson Library) needs to be translated into text and then feed into the destination location of the Network Analyst route-finding function
    • Creation of the library of text to be recognized
    • Speech output: Automation of obtaining the text in the directions and translating the beginning portion of this from text to speech
  • This division may not be balanced, since the bulk of the work lies in programming the on-route, off-route functions, which falls under the GPS <-> Geodatabase part. We can reassess and divide tasks again in a week with clearer understandings of Avenue
• Team Reflections
  • Monica: Thanks to Jessie and Xiaoning for giving WP4.
  • Jessie: Will not present in the final presentation (maybe Intro slide), but will participate in the Q&A session.
  • Xiaoning: It’s good that we are looking towards the end but anticipating some frustrations.
  • Jie: Waiting for good weather to test our device (snow and now rain). Realizing that we will be balancing technical progress with final deliverables soon. We should all communication more frequently to teammates so that when we share decisions with our liaisons, everyone is on the same page. We have experienced times when members seem to make decisions single-handedly that do not necessary reflect group opinion. Team will meet for 15 minutes before technical liaison meetings from now on to gather thoughts.

Plan of action/tasks

• Jessie and Xiaoning will meet Tuesday afternoon following our meeting with Professor Howe to finalize the sections/slides for WP4 and perhaps figure out tasks that Monica and Jie can help out with
• Team will meet this Wednesday from 8 – 8:40pm to do a run-through of this presentation
• Jessie and Xiaoning will give Working Presentation 4 on Friday
• Xiaoning and Jie will meet Friday evening to work on integrating Arcview <-> voice interface using Avenue
• Team will meet Sunday as a session to experiment with Avenue

top of page

Team Reach
Progress Report Week 19
Jessie McCartney, Jie Zheng, Monica Marinescu, Xiaoning Xu
Dates: 3/29/05 to 4/5/05
Recorder: Jie Zheng

Completed actions/tasks

• Directions generation by Xiaoning and Jon. Using an Avenue script in Arcview, they outputted the directions for a calculated route to speech. With a pre-specified layer of two points (the current and final destinations), the shortest route between the two is calculated and text directions are generated automatically and appear on the screen. They are then stored, saved in a text file and read aloud by the voice software.
• Monica determined that we could not activate the function in Network Analyst that would take in locations by their address. Instead, it would only work if clicked on points on the map. This is problematic, since the destination location will be identified in an address form (i.e. Building name). She experimented with the San Francisco map, since route generation by the "address function" worked for WP2. Monica and Jon determined that some connections have to be added to the San Francisco shape file in order for it to be used to do address-based route generation. She made necessary changes, and the file is now working.
• Monica made progress in understanding the directions generation -> voice interface script
  • Everything inside readdirections.txt must be emptied before the script is run. Otherwise, it does not overwrite/write anything new in the text file, and the previous directions are ouputted.
  • Success with running the script with more than two stops
  • Problem: For the same route, directions obtained with Network Analyst manually have units, while those generated by the script have all "zero" distances
  • When a new shape file is added to an open project, the user must first open its attributes table to be able to run the script
• Testing by Xiaoning and Jie with Tracking Analyst and ArcPad. The goal was to collect and log data at real-time with Tracking Analyst so that it may be compared with coordinate information of objects in the geodatabase. After spending more than an hour with the testing unit, only two points visible in Arcview was collected. Viewing this data magnified reveals multiple data points. This needs to be analyzed in further detail to determine whether this data represents locations on campus. Perhaps an issue of projections. They attempted to collect data on receiver accuracy using ArcPad but were unable to save the track log, even though their positions were visible and map projections were set correctly.
• The voice library text file of campus buildings in our testing area wascompleted by Jie. The names are organized in alphabetical order with ";" as dividers. The major possible ways one could refer to a building were included in this file (i.e. bass hall, bass). This list organized as it is will faciliate more more accurate and convenient speech-to-text conversion. The first letter of a voice command will be identified, and a search in this text file will be conducted by that letter.
• All this progress with directions generation has been made with the a map layer of streets for San Francisco provided in an ArcView tutorial, which is more developed than our campus pathways layer. We need to store pathway names and possibly addresses in the table format. Jessie is working on naming the campus pathways. She has added all existing names to our pathways layer and will be developing a scheme to name the rest. Another consideration is that if we want to do route generation based upon addresses, we will need to add 4 numbers to each pathway (left side of one end, right side of the same end, left side of the other end, right side of the other end). This may not be necessary given that we are seeking routes based upon building names, not addresses.

Items discussed

• For further development of directions generation, the next step is to form an initial layer of two points automatically by address, instead of having to do it by hand and feeding it to a script function
• Confusion over our Sunday meeting due to Daylight Savings time and lack of communication led to our team not holding a meeting. Addition to Team Charter: In the case that a member does not show up to a team meeting, the present members must try to contact her to find out why she's not present and/or remind her of the meeting. The reasons for this include that it is very difficult to schedule meetings that work for our schedules, and if we can try our best to ensure that everyone is present, then it's better for everyone. Sometimes members accidently forget a meeting or confuse the meeting time, and simply need to be contacted to show up.
• Clarification of the scope of our testing area, which helps us determine the amount of work in naming campus pathways and voice library text files. We decided to include the Central Campus area as well as Bass and McConnell in the Science Quad.
• WP4 Comments
  • Start slide
    • use looping and make sure it flows, use text that is understandable without explanations, too much detail, make Go!PAL larger, need greater balance (empty on top, bottom crowded)
  • Design Process Web
    • only show components that we are using now. Perhaps include short text as each component comes up to explain it, use straight arrows
  • Image of Go!PAL user and Animation
    • correct spelling of cemet"e"ry, directions do not tell the user to turn, a little busy, label device components on Go!PAL , make the person bigger
  • Flowchart and maps
    • show simplified case first and then future expansions, don't use partial images, look off-balanced to the left, make sure text is readable horizontally, too busy, add titles for slides (building entrances), doesn't always make sense to use both flowchart and campus map on one slide, use labels for the maps, flowcharts hard to follow (use more spacing and arrows), make everything easy to see (color, size considerations), option for user to specify and save preferred option?, no skewing of images, better quality images (no pixellation)

Plan of action/tasks

• Monica and Jessie will work on preparing the poster. A draft will be completed the latest by the Tuesday meeting with Professor Howe prior to the Grinspoon Banquet.
• Avenue Script
  • Find the command that would automatically erase everything inside readdirections.txt and include it at the beginning of the script
  • While for three points the directions output file, readdirections.txt, is identical to manually generated Network Analyst directions file (except for the lengths), the voice software only reads the directions until the first stop. Need to understand why it stops.
  • Figure out how to do searches on fields other than the ID of each object (which is what is currently being used, i.e. building name. Because the object structure is very straightforward, so this should be very feasible.
  • Figure out the units, zero distances problem
  • Script finds the shortest route between wright hall entrance and neilson library entrance through loading the building entrances theme, but it does not write anything in the readdirections file. Resolve this problem.
• Compose separate text files containing building entrances and campus pathways. These campus layers should each have their own corresponding text file.
• We will continue documenting any progress we make in group or individual sessions with short email summaries to the rest of the team.

top of page

Team Reach
Progress Report Week 20
Jessie McCartney, Jie Zheng, Monica Marinescu, Xiaoning Xu
Dates: 4/5/05 to 4/11/05
Recorder: Monica Marinescu

Completed actions/tasks

Thursday Grinspoon Dinner

• Poster put together by Jessie, Jie and Monica, sent to Professor Howe
• Animation to be run in a loop on Go!PAL put together by Xiaoning, sent to Prof Howe and liaisons
• Order of slides in the animation: Title slide with team photo with Go!PAL; Photo of a person wearing it; Animation of how the interaction between the user and device is; Flowchart of the interaction; Web of design process with Go!PAL in the middle and components around

Final Report

• New parts of the Design Process section (that were not included in the Midyear Report, i.e. Voice Software detailed, Arc View with Network Analyst, GPS Receiver, Avenue; and the geodatabase in ArcMap) were divided between team members and outlines were put together for each of these parts.

GPS receiver and Tracking Analyst

• Connector pin for the booster antennae broke again, and an attempt was made to fix it in the machine shop; we do not know whether a connection is now possible, and it is not a long term solution
• E-mail sent to Teletype to request that a receiver + antennae be sent before we mail ours back (we cannot afford to sent be left without a GPS receiver now); we expect an answer by today (they estimated 2 business day for a reply)
• GPS readings with quality 0 (the GPS quality is rated either 0 or 1) are now filtered out , as are all points obtained with less than 3 satellites and all points lying within t he building polygon shapes on the map (In the tabular format, one can view for each point how many satellites were present under the "numsats" field. One of the reasons for inaccuracies far larger than 3m is that fewer than 4 satellites were present. Another possibility is that Go!PAL's operator's arm was blocking the GPS receiver)
• Tracking Analyst is indeed working for real-time GPS data logging. We verified this by viewing the logging data against a campus map. (In order to load a campus map, it must have the WPS projection (default GPS coordinates), not Mass Plane 83) Jon helped us convert a campus pathways shapefile from Mass Plane 83 projection to the WPS projection using ArcView Projection Utility. We loaded this pathways theme into the project and used it as a reference for our tracking data. At the original zoom level of the screen, it always appears as if only one big point is generated, but if one zooms in a lot, she'll see that that one point consists of many, many small ones. 

Voice Recognition

• The voice recognition software is based on the coffee0 example
• The program recognizes commands in the form "go to the counter" as being constructed on three levels:

Go to/Enter ==> the (optional) ==> Shop/Counter/Store

• Preliminary code includes only the speech engine initialization and engine shut down. The initialization includes
• open the port to listen for sound,
• create a voice event,
• load the grammar. The grammar was loaded properly from the resource file. Using a dummy variable for the grammar loading. The program compiles well but not built well. The building errors were LNK error 1220 and LNK error 2001.

Avenue Script

• T he start and end points are now read from the pop up message box that asks the user to input the building names
• T he theme layer used here is the building entrances layer. Directions were successfully generated between "Wright Hall" and "Burton Hall." The xy coordinates of the points matched their names.

Geodatabase

• All paths in our area of interest have been named
• All building entrances in our area of interest have been named
• Path layer has “length” field with the right units and values

Items discussed

Assignment 7 - 10 considerations

• Improve the quality of life for the visually impaired (as in less anxiety related to every day activities, increase confidence and independence)
• Open up meeting and activity possibilities for the visually impaired/groups of visually impaired
• Put mobility trainers out of job
• Environmentally friendly, since further improvement of the device and expansion of its use does not involve other components
• When marketed, make sure that components and packaging are environmentally friendly
• Good, since it uses resources already available (satellites signals, existent software)
• Should the end product be sold or given away to the people who need it
• Who is reliable/responsible for the problems arising when the device does not function properly
• Do we have the moral obligation to find a group of students that would bring it further
• What should the end product be (a software package vs. hardware piece)

Avenue Script

• We could not find the reason why the text output file where the directions are automatically saved overwrites only sometimes
• We added an extra “close file” command to keep things cleaner, and hope this will solve the problem
•  

Voice recognition

• The C++ script gives no error when compiled, but does not build properly (problems with “resource files?”)
• Professor Fagg looks over this

Plan of action/tasks

Grinspoon Presentation

• Tu meeting: discuss how to dress for the conference
• Jessie takes care of the chocolates for out stand
• On Thursday, bring along the remaining business cards, headphones + microphone, speakers

GIS Application

• Monica and Jessie will put together the application for the GIS Competition and send it to Jon by Wednesday morning
• Jie sends a paragraph on Tracking Analyst to be incorporated by Tuesday night
• Tuesday after meeting with Prof. Howe – team meeting to discuss details for the dinner on Thursday

Final Report

• Expand on the existing outlines starting on Sunday
• When explaining the choices made, remind each time how the user’s needs are kept in mind, and they are the constant, while “how” they are met is variable, depending on external factors (ex: when adding names of features in the building entrances, paths layers; choosing the area on campus to focus on; choosing components that could be easily carried; the way grammar is recognized)
• Contact Merrilyn Lewis for final presentation

Voice Recognition

• Rather than fixing the error, Xiaoning will start on the coffee2 or coffee3 tutorial examples, and leave, for now, the visual component there; the newer versions of coffee recognize multiple commands, including "go to the shop" and "go to the office"
• Try to add more such commands and associate actions for each recognized command (ex: store the recognized building name to a text file, so that we it can be used by Network Analyst)
• Find whether there is a method to export the building names directly to the Avenue script.
• Decide on a condition for when the directions reading should end (? Maybe after the directions are given for the next 15 meters? Or a temporal interval?)

Avenue Script

• L oad the building names directly from the voice recognition program (using command recognition) , instead of typing them up in pop up windows
• Make the Script consider the last GPS reading as XY coordinates and transform that into the “current position” that network analyst uses for calculating the route.

Geodatabase

• Jon will show us how to change the nodes in the campus paths layer and recalculate the distances so that network analyst can use those; hopefully this will solve the units problem (t he program report 567 regardless of the reporting unit chosen – in, mm, cm, m, mi or yards)
• Find what network analyst uses as default “cost field” and “cost field unit”, and adapt our attributes tables to that, or change the default setting by commands in the script

GPS receiver

• T ry average GPS readings to improve accuracy
• L oad the averaged point into our script automatically as initial start location and find the nearest building to this point
• After this, we can use the GPS fixes with either the Network Analyst or Tracking Analyst to check whether a user is on/off route
• Install the single user license version of ArcGIS 9.0 onto Go!PAL so that we can modify files in ArcMap and ArcCatalog on our C-drive (Jon has given us the CD’s). The beta version works only when connected to the network

top of page

Team Reach

Progress Report Week 21

Jessie McCartney, Jie Zheng, Monica Marinescu, Xiaoning Xu

Dates:04/07/2005 to 4/18/2005

Recorder: Xiaoning Xu

Completed actions/tasks

·         Voice Reocognition. Three building names were successfully recognized through the voice recognition software. Testing the program with Professor Fagg and Jon, the accuracy seemed to be high with male voices.

·         Teletype GPS unit.

o       We received the exchange for Teletype GPS unit in mail last Friday. The connection between the new external antenna and the GPS receiver was considerably firmer.

o       Monica and Jie downloaded the “trk_coord.ave” script form ESRI. This script automatically obtains the x and y coordinates from the tracking theme. However, it only works with point and polygons, but our tracking them contains much more information than that. We need to first convert our tracking them to zpoints, then to points. To achieve automation, we will need a script that converts the tracking theme to shapefile automatically.

·         Application for Robert R. Churchill Memorial Prize for Student Scholarship in GIS competition. The submission of the application was confirmed by Diana Sinton from Middlebury College. Jon wrote our faculty recommendation for this prize.

·         GIS maplayers.

o       All building entrances names were copied from the old “label” field to the new “name” field, because network analyst reads building names from the “name” field by default.

o       A new “meters” field was created and the respective units were copied from the “length field.” We can now output directions in meters.

o       Monica and Jessie labeled the street and pathways for central campus. For paths without official names, they are named according to the nearby buildings, such as “NeilsonWright Driveway.”

Items discussed

• Preliminary final presentation format
  • Title / Introduction             
  • Overview                                 4 minutes, Jessie
  • Video showing Go! PAL    
  • Background / Motivation
  • Objectives                                4 minutes, Monica
  • Constraints
  • Design decisions                       8 minutes, Monica, Jie
  • Results / Deliverables
  • Ethics                                       7 minutes, Xiaoning
  • Future Implementation
  • Conclusion                               1 minute, Jessie
• Final Celebration. We will like to  invite Jon and Professor Howe for a meal. We will set the schedule at later meetings.
• Professor Fagg will not be able to visit us.

Plan of action/tasks

• Final report parts. (all by Thursday evening 7 p.m.
  • Design / Decision making process. (all for corresponding parts)
  • Final results /deliverables with user manual (XX)
  • Ethics (Jessie)
  • Future implementation (Monica)
  • Read over and update the remaining parts (Jie)
  • Final report editing starts at 7 p.m. Thursday with Jie and Jessie. Monica will edit from 10 – 1 a.m. Xiaoning will join Monica at 11 p.m. and possibly participate partly during the first session.
• Email Merrilyn Lewis to setup practice presentation schedule (Jie)
• Presentation practice times
  • All presentation slides will be emailed to Monica by noon on Wednesday.
  • April 20, 2005 W, 5-6 team practice.
  • April 25, 2005 Monday with Merrilyn Lewis if possible
• Check out video equipments to make the final presentation video (Jie)
• Email Merrilyn Lewis to setup practice presentation schedule (Jie)

top of page