Real-time probability distribution function of waiting time
In light of the newer developments in transportation systems, the Dienstabteilung Verkehr (DAV), the Traffic Service Department of the City of Zurich is interested in upgrading and preparing its traffic-management systems for the V2X era. Based on different projects presented at ITS conferences, DAV has come up with a research question, which they have asked ETH Zurich to address. The traffic control in the City of Zurich is highly versatile, allowing for many different, non-repetitive control sequences. While this allows to prioritize the public transport of the city and to adequately cover the demand created by cars, it makes it difficult to determine or even predict the different variables relevant for traffic. The aim of this research project is to determine the expected starting time of the next green phase for all intersection movements (or signal groups). This calculation is performed for all intersections of the City of Zurich and in real time. We propose to use an algorithd based on artificial intelligence (AI) that receives real-time and historic data from stationary-traffic loop Detectors (LD) and from signal settings as inputs. The output of the algorithm corresponds to the real-time probability distribution function of the waiting time for any signal phase of the network.