Inferring passenger destinations from Smart Card validation records in the Oise department, France

In a more and more critical economic context and cut of public subsidies, transport authorities need support to take smarter decisions. Automated Data Collection Systems, installed on public transport vehicles, can be used to investigate individual mobility patterns and gather demand characteristics, with several benefits, if compared to manually collected data. In fact, in most fare collection systems worldwide, only the boarding information is recorded. In order to overcome the lack of knowledge related to the users’ destinations, many researchers investigated and proposed several models that attempt to solve this issue by analyzing the smart card individual validation data. This work focuses on the development of a system based on the well-known Trip Chaining Model in order to analyze the smart card data of the most important public transport networks of the Oise French department and produce stop to stop origin-destination matrices. An analysis of the literature is reported, mainly on the destination inferring methods starting from smart card data. Then, the methodology, along with a detailed implementation description is proposed. The results of the model are then presented. Finally, some propositions for future studies are given, in order to enhance the results of the Trip Chaining Method, as well as to validate its results. The results of this thesis will indeed assist public transport operators of the Oise department in the planning of services that are more optimized and in line with the users’ needs.