Using GPS trajectories for further understanding of spatial behaviour

Abstract

Global Positioning System (GPS) technology has changed the world. We now depend on it for navigating planes, ships and motor vehicles. We use it in our everyday lives to extract information about our locations and track our movements. In order to fully harness their potential as more and more GPS data are generated, there is a need to understand the necessary steps involved in transforming the raw data collected by individuals into usable trip trajectories and origin-destination matrices.

We present a methodological framework for filtering, segmenting, classifying and semantically enriching GPS data. Following this, we calibrate spatial interaction models using GPS data collected from a survey conducted in the Kingdom of Fife and present frameworks de- signed for modelling commuting flows between local labour markets and weather-specific origin- constrained models for retail analysis. There is an on-going challenge to describe, analyse and visualise the actual and potential extent of human spatial behaviour. This thesis explores this by using existing methods of creating activity space measures to identify their value and highlight their respective limitations. In particular, we identify the “third places” in one’s mobility patterns and design a new activity space measure called the “activity triangle” based on this sociological concept (Oldenburg, 1989).

This research shows that GPS movement data can be used for calibrating spatial interaction models and defining activity spaces. We are at the dawn of an era in which vast new sources of mobility data will be available that will allow spatial behaviour to be modelled by the time of day, the weather and many other variable conditions. This study shows that this is not only possible but provides a route map to go from raw GPS data to calibrated spatial interaction models.

Date of Award	30 Nov 2016
Original language	English
Awarding Institution	University of St Andrews
Supervisor	Stewart Fotheringham (Supervisor) & Urska Demsar (Supervisor)