Adaptive Coverage Path Planning of Marine Vehicles with Multi-Sensor

Recent advancements in autonomous underwater vehicles (AUV) have made ocean exploration a key future frontier. This study addresses the complex challenge of adaptive coverage path planning with multi-sensors (ACPPMS) in marine environments, aiming to optimize AUV's area coverage under strict time and energy constraints. We propose a novel approach that integrates considerations of energy consumption, information acquisition, and spatial coverage, assessing the intricate trade-offs between extending area coverage and enhancing data collection. Our methodology extends traditional adaptive information path planning by incorporating the decision-making dilemma of balancing coverage with information gain, which we tackle through a tree-based sequential decision-making framework, specifically a partially observable Markov decision process. Utilizing an online planning solution, we validate our approach through rigorous simulation in a search and rescue scenario. The results affirm ACPPMS's adaptability and its superior performance in environments with medium to low reward densities, although it falls short in highly rewarding settings compared to conventional full-coverage strategies. Notably, employing multiple sensors not only elevates path efficiency but also contributes to significant energy savings, showcasing the practical benefits of our multi-sensor integration strategy in adaptive maritime exploration.