GPS tracking data of individuals as substitute for travel diaries. Experience from a pilot project in Berlin Germany.

Abstract

The usage of mobile electronic devices to determine human travel behavior became very popular particularly during the last decade. A number of different projects based on new electronic technologies - stationary and mobile - have been conducted. Some of the most promising technologies are GPS for tracking purposes and electronic questionnaires on personal digital devices (PDAs) for activity based approaches. The results discussed in this paper are based on a project which is currently carried out in Berlin and Brandenburg in Germany from December 2004 until mai 2005. During this period ~400 adolescent (age 16 to 20) participate on a one week, activity based mobility study. The participants where asked to fill in an activity based electronic questionnaire on PDAs. In addition they are equipped with a GPS logging device. The project is carried out in waves of about 30-40 (max. 50) participants at the same time. The paper will focus on the quality, completeness and explanatory power of the reported GPS data in different “every day life” situations. As discussed in the literature, using GPS tracking technology does not mean automatically good or complete data regarding the track, origin/destination as well as the mode of a trip. One major goal applying GPS tracking technology in transportation research is the replacement of stated user response in traditional paper pencil or electronic questionnaire with automatic GPS tracking data. Up to now even the GPS signal itself is sometimes nonexistend. There are a number of situations where GPS data needs to be complemented by other information sources. The most common problem based on GPS technology is the lack of tracking data during start times and in “urban” tunnels. It can also be shown that tracking people in public transport with many trip chains on different modes will cause problems. This is especially crucial in societies with a high share of public transport. The combined analysis of GPS data and electronic questionnaire data will show if and how trip information can be derived just by analyzing the GPS data. This has been shown in the literature for example by assigning certain trip modes to GPS data (e.g. car use and walking) via speed analysis. The complete replacement of the travel diary by GPS tracking has not been proven yet. Even the determination of modes other than car driving just by analyzing GPS data is causing problems.