Pedestrians' safety using projected time-to-collision to electric scooters.

Safety concern among electric scooter riders drives them onto sidewalks, endangering pedestrians and making them uncomfortable. Regulators' solutions are inconsistent and conflicting worldwide. Widely accepted pedestrian safety metrics may lead to converging solutions. Adapting the time-to-collision from car traffic safety, we define projected time-to-collision and experimentally study pedestrians' objective and subjective safety. We design isolated and crowd experiments using e-scooter-to-pedestrian interactions to assess the impact of various factors on objective safety. In addition, we conducted a pedestrian survey to relate the subjective safety and the metric. We report a strong correlation between subjective safety and the projected time-to-collision when agents face each other and no relation when the e-scooter overtakes a pedestrian. As a near-miss metric correlated with pedestrian comfort, projected time-to-collision is implementable in policy-making, urban architecture, and e-scooter design to enhance pedestrian safety.