Last decades have been marked by deep socio-economic transformations, an uneven evolution of transport demand in main urban areas and the emergence of new and more sustainable modes of transportation (carpooling, self-services bicycles). These changes have strongly impacted the interaction between service supply and demand in the transport industry. In this context, passive data as Wi-Fi and Bluetooth become a key source of information to understand individual mobility behaviors and ensure the sustainable development of transport infrastructures. In this paper, we present a framework that uses disruptive technology to collect passive data in buses, continuously and at a lower cost than traditional mobility surveys. Previous research, conducted over a more limited spatial and temporal framework, uses filtering methods, which do not allow the results to be replicated. This study uses artificial intelligence to sort transmitted signals, get transit ridership and build Origin–Destination matrices. Its originality consists in providing a concrete, automatic and replicable method to transport operators. The comparison of the results with other data sources confirms the relevance of the presented algorithms in demand forecasting. Therefore, our findings provide interesting insights for data-driven decision making and service quality management in urban public transport.