Agent based modeling and simulation to analyze and measure the Cooperative-Intelligent Transport Systems KPIs

Abstract

Cooperative-Intelligent Transport Systems (C-ITS) aims to facilitate cooperative, connected and automated mobility. C-ITS is currently demonstrated by projects such as C-MobILE, which is a large-scale demonstration project spanning from 2017-2020 that is currently advancing C-ITS by aiming for a fully safe and efficient road transport without casualties and serious injuries on European roads. To support this, Key Performance Indicators (KPIs) are used to evaluate and assess the impact gained by C-ITS developments. Out of many KPIs set out in the project context, we focus mainly on improvement of traffic efficiency, safety, and sustainability. We picked an unsignalized intersection, Neckerspoel Eindhoven, where cyclists and buses share a common space, which compromises on safety and efficiency. Assessing the KPIs is challenging, because of the limited infrastructure (no camera, no traffic lights), no empirical data and unpredictable behavior of road users such as cyclists and bus drivers. We adopted an agent-based modelling and simulation (ABMS) technique, which is an efficient method to simulate real-life situations and has lower cost compared to empirical and field experiments. Since Neckerspoel junction lacks traffic control infrastructure, we have modelled the intersection including various Intelligent Transport System (ITS) agents: roadside units, cameras, on-board units. This model allows us to measure the impact on cyclists and bus driver behaviors, and comparing the KPIs of different scenarios. In particular, we have studied the intersection with and without ITS agents. The results of the simulation served as a recommendation for the Municipality of Eindhoven to improve the situation at Neckerspoel junction.