Underwater Vehicle System for Drowning Intervention

The rescue of an individual underwater in swimming pools, water parks and in lakes have become an increasingly critical need in the current world. Traditionally, drowning individuals have been rescued by the human divers; however, the need for safe and costeffective rescuing has fostered the development of unmanned underwater vehicle (UUVs) which are capable of performing the underwater surveillance and rescue operations. This research presents the design and development of an unmanned underwater vehicle for operating in situations of search and rescue. The significant challenges addressed include reliable communication in the underwater environment, efficient manoeuvring, and an image processing for human detection. The designed UUV has a single thruster configuration consisting of a cylindrical projection serving as thruster housings and inbuilt buoyancy chambers, arranged to provide multidimensional manoeuvrability and stability underwater. The UUV features an integrated victim retention system, which is operated along with a pool float deployment system through servo motor actuation. A 3D CAD Model of the UUV was subjected to Finite Element Analysis to verify that the design allows for the structural integrity of the UUV in 10-meter-deep waters. For the machine to detect human shapes in underwater environment, we trained YOLOv8 object detection algorithms on a custom dataset of 2400 underwater images. It was tested, and underwater humans were consistently detected with success. The proposed system improves on its accuracy and also shows considerable decreases in the time delay and amount of energy required by comparison with remotely operated underwater vehicles and conventional rescue systems. The contributions of this research to advance UUV technology and possible applications towards public safety, disaster response, search operations, and opened space for the development of future AI enhanced aquatic emergency systems.