SLAM based adaptive navigation of AUVs for oil spill cleaning

Oil spill pollution causes serious contamination of the ocean, harms marine life, and results in severe impact to the economy and ecological balance. Several efforts have been dedicated for detection, monitoring and localization of oil spills; however, the complete cleaning of oil spills remains a challenging problem. Due to dynamic ocean currents on the surface and limitations in accuracies of the localization systems, the exact locations of oil spills are difficult to estimate. Thus, the estimated oil spill area may not match the actual shape of the spill. Furthermore, the access to GPS may not be always available and obstacles may exist as well. Therefore, it requires an adaptive oil cleaning method that detects oil in situ and allows the autonomous vehicle to adapt its cleaning trajectories to the oil shape while building the map and avoiding obstacles. This paper presents a SLAM based adaptive oil cleaning system using an autonomous vehicle in GPS-denied environments. The exact a-prior knowledge of the oil spill area is assumed to be unknown or merely partially known. The autonomous vehicle can dynamically detect oil shape and obstacles, and adapt its trajectories for complete oil cleaning. The proposed system introduces a discrete event supervisory controller that develops a concept of multi-resolution navigation to prevent the autonomous vehicle from getting trapped into a local minimum which is encountered generally in potential field based methods. The SLAM algorithms provide online estimate of the vehicle location. The efficiency of the proposed method is validated on the widely used high-fidelity Player/Stage simulator.