KY Module 3: Geocoding and GPS

1
KY Module 3 Geocoding and GPS
2
What is the single hardest information to get
from a travel survey?

• Location, location, location
• What is the address of YOUR grocery store?
• Easiest to report are home and work location.
  Even school addresses are often difficult for
  respondents to report.

3
Difficulties in location reporting

• Interviewer and processing difficulties
• CATI interviewers not familiar with local
  landmarks
• Multiple matches to GIS basefile (Main Street,
  Washington Street).

• Respondent difficulties
• Street suffix (St, Ave, Blvd)
• Directionals (N, S, E, W, NE, SW, etc)
• Difficult street names
• CATI interviewers not familiar with local
  landmarks

4
Why do we need this geographic detail?

• Attempt to code origins and destinations to
  latitude longitude.
• Previously code to TAZs, but TAZs keep getting
  smaller, as computer technology allows for larger
  O/D matrices.
• By coding to lat/long, TAZs can be modified in
  the future and the data can be summarized for new
  TAZ definition.
• Want to link detailed geographic info to model,
  e.g. slope, sidewalks, distance to bus stop.

5
Better location information

• How to make it easier on respondents to provide
  location information?

6
Can web-response play a role?

• Complete it on their time, not when the phone
  rings.
• Sample bias, but high income hhlds and young
  males are also part of non-respondents, and this
  method is useful for reaching these hhlds.

7
Geocoding on Internet

• Interactive Yellow Pages look-up
• Who knows the address of their grocery store or
  barber shop/beauty salon?
• Point and click on a map to identify a location
• Type in an address or intersection and have the
  GIS interface locate it on the map

8
GPS may be more accurate and complete, by relying
of passive data collection.

• GPS equipment is small and getting less expensive
  everyday.
• A new Trimble GPS receiver is the size of a
  pencil eraser! But, you still need an antenna.

9
GPS in Household Travel SurveysStudies in the
U.S.

• 1995 Lexington, KY proof of concept (100
  vehicles)
• 2001 CalTrans 3 counties in CA
• 2001 SCAG (Ben Pierce)
• 2001 Ohio DOT (Ben Pierce)
• 2002 Smartraq Atlanta
• TYPICALLY 200-500 HOUSEHOLDS, with up to 3
  vehicles equipped per hhld.

10
(No Transcript)
11
Example from Geostats

• Comparison of self-reported trips (phone
  retrieval) to GPS recorded data
• Estimated 25-30 of trips are unreported, but
  still working to determine impact on VMT, cold
  starts, trips by purpose

12
Example from Geostats

• GPS generated maps as a recallmechanism.
  People may be able to recall what they did a week
  ago if provide a map with numbering sequence and
  time-stamp. Use of internet to pass Map to the
  respondent.
• Test in Louisiana, continued work in Australia
  (Peter Stopher)

13
Other approaches

• Equipping sample of taxis with vehicle navigation
  service in Japan
• Tracking cell phone users in Germany and in Japan
  (Dokomo)
• Activity Motivation and re-scheduling Sean
  Doherty

14
GPS in truck surveys

• Very difficult to get good response rates from
  commercial vehicle surveys
• One GPS survey in California (CARB and FHWA),
  conducted by Battelle
• Only fleets were included, not independent
  truckers who make up 50 of the CA fleet.

15
GPS in truck surveys

• Large fleets may already have GPS for fleet
  management, but not likely to share data with
  public sector, cuz it is considered proprietary
  information.

16
GPS for travel speed studies

• Ability to cover large number of miles
• Only need one person per vehicle (old fashion
  method with stop watches needed 2 people)
• College students, working part-time, very
  comfortable with using palm-tops and laptops for
  data entry and data uploading.

17
Sample of digital video image, georeferenced to
GPS data
18
(No Transcript)
19
GPS for travel speed studies

• Travel Speed along links
• Travel speed compared to posted speed
• Travel speed by functional class
• Examine recurring congestion
• Examine queuing length (defined when speed falls
  below 5 mph or less for 5 seconds)

20
Show 2 .jpg from Steves CD