Towards an Autonomous UAV-based System to Assist Search and Rescue Operations in Man Overboard Incidents
Valeria Alexandra Feraru, Rasmus Eckholdt Andersen, Evangelos Boukas
Abstract
Man overboard (MOB) incidents in open sea travel are a risk for the shipping industry. In this work we aim to design an autonomous system in charge of planning and executing a search for a person in water (PIW), victim of MOB. The proposed solution consists of an autonomous UAV that initiates a SAR mission when it receives a MOB signal. The UAV is located in an autonomous docking mechanism onboard the ship. The probabilistic leeway model is implemented to find the location of the PIW and a Deep Learning object detection algorithm (Faster R-CNN) is employed to locate the PIW. To the best of our knowledge, this is the first time that a search system on-board a vessel -capable of live planning and execution- is proposed. Our results indicate that employing our solution would increase the success probability of finding a PIW and decrease the total search time.