In the last years, several major industrial disasters happened around the world, such as the one of the mining dam collapse in Brumadinho in 2019 and the explosion of Beirut Port in 2020. This study focuses on the use of drones to identify survivors and victims in an area impacted by a major industrial disaster. Heuristics and an ALNS are proposed to solve this problem in a large scale. Preliminary computational experiments shows the efficacy of these methods.