Development of Adaptive Traffic Control System – Dynamic Intersection Signal Control Optimisation (DISCO)

This project will extend the developed DISCO prototype for general traffic scenarios, speed up optimisation by parallelisation, AI-based engine, and machine learning, scale up applications to network-wide junctions by decentralisation algorithms and cloud computing, and establish a software-inthe-loop connection with a micro-simulation software for validation. The project will also link the DISCO software platform to an actual traffic signal controller used in Hong Kong for validation, and establish linkage between DISCO and a cloud sensor database, in which traffic data will be imported and used in DISCO for model calibration and optimal signal plan calculation.