A Modified Social Force Model (SFM) for Pedestrian Behavior in the Presence of Autonomous Vehicles (AVs)

The social force model (SFM) is a widely used method to predict pedestrian behavior. This paper develops a modified social force model to understand pedestrians' behavior at a signalized crosswalk with only autonomous vehicles (AVs) on the road. Previous research indicated that pedestrians are likely to feel less safe around driverless "ghost vehicles" like AVs, and a repulsive behavior of the pedestrians is observed toward AVs. Hence, a new repulsive force is incorporated into the traditional SFM to account for this phenomenon, and the modified SFM's principle used to simulate pedestrian behavior. Simulation of pedestrian behavior is performed using the open-source crowd simulation software, VADERE. Simulation results show that pedestrians' walking behavior becomes chaotic indicating that pedestrians may require more time to cross the road in this scenario. The paper discusses the results of these simulation experiments and their implications on transportation engineering.