Rail Lone Worker Safety Solution with Deep Learning
Cristina Rodriguez Vera, Lin Lan, Frédéric Bernaudin, Benoit Besson, Jun Hui Fu · 2022 IEEE International Conference on Pervasive Computing and Communications Workshops and other Affiliated Events (PerCom Workshops) · 2022
The European Framework Directive on Safety and Health at Work sets requirements that companies must provide adequate supervision and protection to protect workers’ health and safety. A watch-based solution has been designed and prototyped specifically for lone worker safety protection in the rail industry, characterized by a harsh working environment. The prototype integrated edge-based Artificial Intelligence accident detection, rule-based management functions for automatic & manual safety alert and for battery energy consumption, as well as worker support functions. An event-driven messaging function has been developed for transmission of detected incident data between the worker and the remote assistance team, in addition to voice communications. This event-driven message has been inspired by European Standards on Cooperative Intelligent Transport System. The developed prototype has been tested in a lab environment with promising results. Real on-site testing with lone worker is currently ongoing.