Abstract: For decades experts in the field of transportation have used traffic and transport models to describe and predict mobility patterns and traffic conditions. Most of these models mainly use 'static' data of regions as input and the models are validated based on questionnaires and average workday traffic counts. In the past years, large open data sources have allowed for a paradigm shift towards more data-driven modelling. Large real-time loop-detector, floating device and camera data sets enable us to use the data as a starting point to develop real-time self-learning models. These data-driven models give more improved insights for strategic measures and tactical and operational traffic management. In this presentation traffic and transport models are briefly introduced and examples are given of data driven models for the road network and for the canals of Amsterdam.