Using AI to Improve Safety at Grade Crossings

C. McGlynn, H. Zhang
Rowan University,
United States

Keywords: railroad, grade crossing, safety, sensors, AI


Railroad grade crossings pose a unique, life-threatening hazard to motorists and pedestrians alike. Approximately every three hours, a person and/or vehicle is struck by a train in the United States and this has remained largely unchanged over the last sixty years despite unprecedented technological advancement in other realms. For example, warning bells, train horns, flashing lights and motion activated gate technology have remained largely unchanged over this time frame while automobile technology has leapt forward. The Sonic Shield aims to solve this via a portable device that tracks the resonant signature of a moving train within defined parameters including distance, direction and speed. The novelty of this technology will be the proprietary algorithm that processes sensor data in real-time to identify a potential collision with enough warning time for drivers to stop. The intellectual merit of this technology include: (a) Passive sensor technology that does not require installation at rail crossings to function, (b) Ability to be widely distributed at minimal investment while offering cross platform functionality, (c) Additional margin of safety provided to users, especially at night or in otherwise low visibility situations, and (d) High potential to create new knowledge regarding the optimal design and range for vehicular mounted sensors.