Publication

Abstract : Coral reefs are the biggest underwater ecosystem that provides habitat for millions of aquatic species. Recent studies show that corals face extinction due to various natural and man-made imbalances in the climate. The importance of coral reef inspection and related studies is relevant in overcoming these effects of nature. Coral reefs are considered the most biodiverse ecosystem in the world. They cover only 0.1% of the earth’s surface but possess a variety of living species. They are home to 4000 species of reef fishes, 840 species of corals, and over 1 million species of other animals. The proposed work addresses the challenges faced in coral reef monitoring and inspection by the entire system design of an autonomous underwater vehicle (AUV). The designed AUV is capable of oceanic inspection, monitoring, and detecting the abnormalities in corals using Computer Vision and Machine Learning (CV/ML). AUV consists of a Single Board Computer (SBC) that runs ML models to detect coral bleaching. SBC guides the control system of AUV to navigate based on autonomous decisions. The surface beacons (SB) which are attached to AUV collect information from AUV and deploy to the surface upon SBC command. SB can detect the sea surface, upon reaching the surface of the sea, it activates the GPS and transmits the collected logs to the base station with its position information. Developed AUV has the capability of detecting healthy and bleached corals and geotagging the location for post-inspection.