Bottom-following for remotely operated vehicles

This paper addresses the problem of designing high-precision bottom-followers for remotely operated vehicles. The proposed control architecture, able to uncouple the robot's kinematics (guidance) and dynamics (velocity control), allows the system to handle unmodeled, i.e. not measured or estimated, kinematics interactions between the robot and the operating environment. In order to increase the bottom-followers' reliability, the paper discusses possible techniques for modeling and handling the environmental and measurement uncertainty in the estimate of the local interactions between the vehicle and the operating environment, i.e. altitude and bottom slope.