Using modern ICT for semi-automatic data collection of activity-travel patterns in the spatio-temporal context of ubiquitous urban environments: Experiences with tracing and data imputation

The activity-based approach views the city as a collection of individual activities, actions, reaction and interactions in the spatiotemporal context of a large-scale environment. For understanding the choice behaviour of individuals and describing urban place in terms what is going on data on daily activity patterns of individuals should be collected. Advanced tracking technologies could facilitate and improve empirical research on human special behaviour. They provide researchers a far more accurate database than traditional means of data collection of a much higher degree of temporal and spatial resolution. Thus the purpose of this paper is twofold: first it aims to report an overview and experiences on data collection with advanced tracking technologies; secondly, it presents the theoretical background of the activity based approach and illustrates how resulting data obtained with GPS technologies can be used to represent the human activities in a spatiotemporal context according to the concepts of the activity-based approach. In addition, this paper presents a system TraceAnnotator for imputing different facets of activity-travel patterns from GPS tracers by using Bayesian belief networks. Activities and movement/travel patterns are represented using a GIS environment at the aggregated and individual levels.