██████╗ ██╗   ██╗██████╗  ██████╗ ██████╗ ██╗     
██╔══██╗╚██╗ ██╔╝██╔══██╗██╔═══██╗██╔══██╗██║     
██████╔╝ ╚████╔╝ ██████╔╝██║   ██║██████╔╝██║     
██╔═══╝   ╚██╔╝  ██╔══██╗██║   ██║██╔══██╗██║     
██║        ██║   ██║  ██║╚██████╔╝██║  ██║███████╗
╚═╝        ╚═╝   ╚═╝  ╚═╝ ╚═════╝ ╚═╝  ╚═╝╚══════╝

PyroRL is a new reinforcement learning environment built for the simulation of wildfire evacuation.

Motivation

A major effect of climate change today is the increased frequency and intensity of wildfires. This reality has led to increased research in wildfire response, particularly with reinforcement learning (RL). While much effort has centered on modeling wildfire spread or surveillance, wildfire evacuation has received less attention. We present PyroRL, a new RL environment for wildfire evacuation. The environment, which builds upon the Gymnasium API standard, simulates evacuating populated areas through paths from a grid world containing wildfires. This work can serve as a basis for new strategies for wildfire evacuation.