Intelligent self-evacuation path planning for fire emergencies in underground coal mines

Abstract

In the case of fire emergencies in underground mines, the mine workers undergo significant psychological and physical stress in their battle with time to self-evacuate safely. This can impart their ability to correctly assess the fire-induced hazards in the vicinity of their location and therefore to choose the safest action or the safest escape route. At the same time, the workers do not have any way to know the state of the mine tunnels beyond the immediate vicinity that their senses can provide information about potential hazards. Moreover, the highly dynamic state of a mine, especially under a fire emergency, can render previously safe routes extremely dangerous in the blink of an eye. This study proposes a framework and presents proof of concept for a real-time smart evacuation route-planning approach based on graph theory. In the effort to assist mine workers to safely reach the surface or a refuge chamber, a smart system could provide invaluable acquisition of mine-wide situational awareness to the workers. An IoT of sensors, such as gas concentration, temperature, smoke, oxygen, and air speed sensors, combined with a real-time path planning algorithm could be a powerful tool to such situations. A mine can be represented by a topological map and every location can be assigned a real-time updated value that quantifies the fire-induced hazard based on data collected by a mine-wide IoT. This combinatory risk considers parameters such as concentrations of toxic gases, oxygen levels, heat, and visibility. Safety and health exposure limits as defined from the various regulatory entities are combined with simulated IoT data to calculate the combined risk. The optimized escape routes could significantly assist mine workers to reach a safe location.