RAMI 2023: The Experience of the UNIFI Robotics Team with FeelHippo AUV

Robotics competitions represent an exciting and in-valuable opportunity to evaluate the performance of Autonomous Underwater Vehicles (AUVs), as they provide a useful test bed for the most advanced underwater technologies and an opportunity for a constructive assessment of the strengths and weaknesses of the participating platforms. In this context, the Department of Industrial Engineering of the University of Florence partici-pated, in July 2023, in the Robotics for Asset Maintenance and Inspection (RAMI) 2023 competition held in La Spezia (Italy) using FeelHippo AUV, a compact and lightweight AUV. Despite its diminished dimensions, the FeelHippo AUV is equipped with a comprehensive navigation system and several payload acquisition and analysis capabilities, During the sea domain trials, throughout the competition, robots were asked to carry out multiple autonomous missions. FeelHippo AUV precisely navigated through the competition arena, closely following a path, planned according to the estimated positions of the objects, evaluated from the in-water runs and autonomously performed a lawnmower path to cover the area of a precise structure to realize an acoustic mosaic of the area around the structure, using the payload sensor. Furthermore, FeelHippo AUV was used to inspect a buoy region and localize the buoy online, with an Automatic Target Recognition (ATR) strategy. The proposed approach relies on the convolutional neural network (CNN) YOLO (You Only Look Once), which ensures optimal results in terms of precision and accuracy. Finally, the main outcomes achieved by the team in the field, during the competition are reported, to demonstrate how satisfactory performance (both in terms of navigation accuracy and payload data acquisition and processing) can be achieved even with small vehicles such as FeelHippo AUV.