Architecture of a Public Transport Supervision System Using Hybridization Models Based on Real and Predictive Data

Abstract

Management of a multimodal transport network is a challenging task and operators regularly have to deal with different types of disturbances that affect the quality of service of the transport system. To support them in their decision-making, we present the architecture of a multimodal supervision system that implements monitoring, prediction of affluence, disturbances, impacts and evaluation of functionalities focusing on one or multiples lines of the network. Our system uses the real-time data of transport operators to compute several key performance indicators (KPI) in order to monitor the network status and to detect disturbances. The prediction function is used to predict passengers’ attendance at stations, incident duration and the evolution of the whole network, in particular, phenomena emerging from the interconnections linked to the affected line. Finally, we present a use case of the supervision system applied to a real-time control of a bus line belonging to a very dense and busy segment of the multimodal transport network in Ile-de-France area.