Exploring the Potentials of Automatically Collected GPS Data for Travel Behaviours Analysis

Schoenfelder, S., Axhausen, K.W., Antille, N. and Bierlaire, M. (2002) Exploring the potentials of automatically collected GPS data for travel behaviour analysis A Swedish data source, Arbeitsberichte Verkehr- und Raumplanung, 124, IVT, ETH, Zuerich.

This paper presents an approach to gain longitudinal travel behavior data by means of GPS. The purpose of this work is to fill the absence of mobility survey that last longer than one week. Beside the problem inherent to longitudinal surveys (limited pool of respondents, fatigue effects), the authors identify the potential technical drawbacks as: transmission problems, warm-up times before getting a fix, cost of post-processing of the GPS data. They also highlight the importance off taking the user into account:

The level of user interaction is believed to be an important issue for the development of future survey design incorporating GPS data collection elements.

A clustering technique was used to identify unique origins and destinations of travel. Their approach sets a tolerance distance or zone in which different final positions are per definition considered as only one destination.

Relation to my thesis: Similar to Elimination of the Travel Diary: An Experiment to Derive Trip Purpose from GPS Travel Data, this study suggests an approach to analyze people’s mobility based on sometimes missing sensed data, GIS issues such as digital land-use(data not including private road and parking spaces. The inaccuracy of available GPS data requires additional processing work. Indeed, the data post-processing must take these uncertain information to predict travel behaviors.