Project: The Use of GPS to Improve Travel Data B31
Reference: STP 14/5/14
Last update: 13/08/2003 18:44:28
The specific objectives of this research are to determine the:
acceptability of passive personal GPS data recorders;
reliability of data collected from passive personal GPS data recorders;
extent to which these devices can improve travel survey data quality;
extent to which passive personal data recorders might be used to replace conventional travel diary data;
cost-effectiveness of using such devices.
Travel data collection is an imperfect craft. One of the key problems is the unreliable nature of human beings. Respondents tend to underreport trips, simplify the reports of journeys to save time and we are particularly poor at estimating times and distances.
This project will establish the extent to which GPS technology can improve the quality and cost effectiveness of travel data collected in the UK now, and in the future.
GPS devices have already been tested with promising results in the US, Canada and the Netherlands. The aim of this project is to test their application within the context of the London Area Transport Survey (LATS).
The methodology involves equipping a sub-sample of LATS respondents with passive personal GPS data recorders and comparing the paper travel diary data with the GPS derived data. This will enable an evaluation of the benefit of GPS data, for example in terms of the reduction in missing trips. The reliability of inferred journey purposes and modes chosen from GPS data will also be investigated.
While this project focuses on the LATS as a specific application, there are much wider possibilities, for example in aiding TfL to monitor the effects of congestion charging, or local authorities to monitor the effectiveness of their Local Transport Plans. Furthermore, as the technology develops and becomes more widespread, for example it is likely to appear in cellular phones, so the importance of this approach to travel data collection will increase.
Steer Davies Gleave
28 - 32 Upper Ground, London, SE1 9PD
020 7919 8500
Atlanta, Georgia, USA, .
Cost to the Department: £49,970.00
Actual start date: 01 April 2002
Actual completion date: 11 April 2003
The Use Of GPS To Improve Travel Data
Author: Steer Davies Gleave
Source: Steer Davies Gleave
Summary of results
- It is well known that travel data collection suffers from many imperfections due to human nature. Thus, locations and times can be imprecise, while some trips may be forgotten entirely. What has been less clear is the extent of the problems and, perhaps more pertinently, what can be done to overcome them. GPS (Global Positioning System) technology appears to offer a potential solution since it can collect very accurate spatial and temporal data, tracking exactly where and when a subject is on the move. All that is required is that the subject carries with them a recording device they are prepared to hand over to a researcher for analysis.
There are potential problems to overcome though: are people willing to co-operate in what might be seen an invasion of privacy? Is the technology and equipment reliable enough? Will a GPS signal be able to be picked up continuously or will there be large gaps in the data caused by "urban canyons" or travelling on systems such as the London Underground? A number of car-based studies have been undertaken in the U.S. and these gave some confidence that the approach was worth looking at in a U.K. context, and in a personal travel
The project explored how GPS could enhance the London Area Transport Survey (LATS), a very demanding environment for the technology. Transport for London (responsible for LATS) therefore, played a key part in facilitating this project. Research International (RI), the fieldwork contractor for LATS, also conducted the fieldwork for this study.
The primary objectives set for the study were to:
test the acceptability and reliability of the equipment;
demonstrate whether GPS can identify trips missed from diary surveys;
explore how GPS might be used to improve the quality of diary data in terms of addresses, timings, and route choice.
The survey method involved returning to a sub-sample of LATS respondents, asking them to complete another two-days travel diary plus a one day recall interview, while at the same time carrying a GPS data recorder ("Geo Logger") at all times during three days. The sample size for the project was 154 respondents, each providing up to three days of data.
Although there were some comments about the size and weight of the GeoLoggers (mainly due to the battery packs) RI were able to place the required number of GeoLoggers and concerns over security and privacy were less apparent than expected. The GeoLoggers proved reliable with just a 1% equipment failure rate.
Of all the possible days for data collection, excluding equipment failure, 82% yielded worthwhile GPS data. The remaining 18% were less usable due to human error: not carrying the GeoLogger correctly, or forgetting to carry it at all. Although there were gaps in the GPS data in which the signal was lost for part of a trip, the overall data quality was good with the GeoLoggers usually able to pick up a signal from at least four satellites at a time. By comparing the diary data with the GPS data it was possible to verify the GPS result and fill in data missing from the diaries.
We were able to identify missed trips, and in fact the level of trip under-reporting was calculated at 16%. The project has demonstrated the wealth and precision of spatial data obtained by GPS, so not only are origin and destination addresses accurate, information is also available on routes taken and any stops made along the way.
Improvements in timing accuracy are also possible using GPS, with a very noticeable rounding effect evident within the diary data which do not affect GPS which can therefore detect minor day to day variations in travel times.
This study has helped to demonstrate that GPS technology has a future in travel data collection. In the short term, GPS is likely to be used in a supporting role to more traditional diary methods, for example in calculating correction factors. Longer term, the roles may switch so that GPS is the primary method with recall interviews use dto verify and supplement it. At the same time, there may be a move towards continuous data collection rather than periodic surveys.