Modeling Perceived Safety for E-Scooter Users in an Urban Road Network: A Case Study in Central Athens, Greece

Electric scooters (e-scooters) have become popular in many urban areas around the world due to their contribution in the reduction of environmental impacts and the convenience to purchase and use them. Due to their innovative role within the transport system, a simulation software for modeling their dual behavior is not yet available. This research aims to develop an agent-based model (ABM) for e-scooters by incorporating users' perceived safety. Perceived safety differs per urban road and individual; thus it is modeled for three transport modes that may perform first/mile trips: car, walking, and e-scooter. The resulting model is integrated in MATSim platform, and it is used to evaluate different travel scenarios within the central Athens, Greece. Results show that the perceived safety impacts differently the evaluated transport mode users. In small networks, perceived safety does not affect routing behavior of car drivers compared to e-scooter users, who change routes when experiencing safety discontinuities.