Decision Tree Based Incident Detection for Distributed Progressive Signal System in an Organic Traffic Control System

Today's world-wide traffic congestion significantly contributes to wastage of time and fuel in addition to environmental stress, such as air pollution due to CO2 emissions. To address these concerns, traffic management systems are continually developed and improved to become intelligent and adaptable. For instance, self-organizing strategies like the Organic Traffic Control (OTC) system present benefits such as enhanced efficiency, resilience, and scalability. Beyond the de-centralised and traffic-responsive manipulation of traffic signals, synchronised adjustment of traffic lights through Progressive Signal Systems ("Green Waves") plays an important role. As previous work we presented an approach for creating decentralised PSS, taking recognised incidents into account to steer the traffic flows. In this paper we further propose a decision-tree based incident detection mechanism.