Building a realistic environment for multiagent mobility simulations

Transport systems are increasingly complex and are made of more and more connected entities. It becomes critical to develop micro-simulation tools to understand the new transport systems dynamics. However, the data for building mobility simulation are quite hard to get, and simulations on new areas are not easy to set up. In this paper, we propose methods for building a realistic data environment for multimodal mobility simulators. We also propose a method to integrate travel patterns (patterns of travelers' origins and destinations). The methods presented in this paper can be used when dealing with new areas for which we have few and incomplete data.