Key points are not available for this paper at this time.
Positioning of underwater robots in congested and enclosed spaces remains unsolved for field operations. Well existing field ready systems are generally more suited to use in to use in large, open marine environments. In enclosed and congested environments, which are common in industrial settings, existing systems suffer from a mixture of issues, including: poor coverage, reliance on added infrastructure and the need for feature rich environments. Accurate and readily deployable positioning is a prerequisite for performing repeatable autonomous missions and therefore, until now, there has been a technological bottleneck in such environments. The Collaborative Aquatic Positioning system presented in this paper uses a mixture of collaborative robotics and sensor fusion to solve the problem. The proposed positioning system is deployed in a large water tank and repeatable autonomous missions are performed using the system's position measurement for real-time feedback. Experimental results show that the system can achieve a Euclidean distance Root Mean Square Error (RMSE) of 70 mm while operating in real-time. The system enables almost complete coverage of the body of water in large pools without requiring fixed infrastructure, lengthy calibration, or feature rich environments. The Collaborative Aquatic Positioning system builds upon recent advances in mobile robot sensing and a recently developed leader follower control system to provide a step-change in positioning capability for real-world, high-precision autonomous underwater navigation.
Cheng et al. (Sat,) studied this question.