Intelligent seafloor survey by cooperative multiple AUVs based on real-time analysis of a bathymetry map

In this paper, we propose a method for cooperative surveys of the seafloor using multiple autonomous underwater vehicles (AUVs) where they detect areas of interest within the survey area in real-time based on a bathymetry map and then details are observed by them. The AUV firstly makes rough observations by mapping using an acoustic sensor and detects areas of interest based on a gradient descent and clustering approach. Second, the set of waypoints for detailed observations and the optimal landmark position are generated. Tests were conducted in a controlled tank environment during which a bathymetry map was obtained using an AUV called “Tri-Dog 1.” Based on this data, the capability of the proposed method for detection of areas of interest and generation of final waypoints for detailed observation were evaluated. It was also verified that survey efficiency can be improved compared to conventional survey methods.