UAV Design for Fully Autonomous Man Overboard Detection

More than 85% of the people falling overboard ships lose their lives at sea. This paper enables the real-world implementation of our past theoretical works by designing and validating a UAS against the imaging and flying specifications required for search and rescue missions of people falling over-board vessels. The UAV is designed to be deployed from a vessel, autonomously search for a man overboard person, provide an inflatable device to them with a GNSS marker, report their location back to the vessel, and return. It is equipped to handle the adverse weather conditions of maritime environments. It employs a sensor suite combining RGB, near-infrared, thermal, and LiDAR sensors to detect people under different light and weather conditions at sea. A custom-designed 3-axis gimbal on board the UAV hosts all sensors and enables their smooth and accurate movement while maintaining stability for improved data capture quality. Furthermore, this paper presents a method to adjust the altitude of the UAV and the angle of the gimbal, in order to minimize power consumption while preserving the detection capabilities of the sensor system. The preliminary flights indicated that the implemented UAV fulfills all the necessary requirements for fully autonomous search and rescue operations at sea.