Contact-based object inspection with mobile manipulators at near-optimal base locations

Abstract

This paper presents a control and motion planning algorithm for a mobile vehicle-manipulator system such that the mobile vehicle and the manipulator mounted on it work in harmony to inspect unknown objects. Forward Dynamic Control method is used for the manipulator to accomplish a stable interaction with the environment and constrained particle swarm optimization is applied so that the vehicle can be localized at the estimated points maximizing the dexterity of the manipulator. Quartic splines are implemented to generate a smooth path for the vehicle in between the optimal locations. The proposed architecture is validated via an experimental setup consisting of a robotic arm with a force sensor at its end-effector mounted on a parallel manipulator. These experiments emulate an underwater vehicle-manipulator system, where the mobile base is subject to disturbances due to the physical interaction of the end-effector with the environment, typically a pipe. The advantage of the proposed approach is that it allows continuous and smooth movement of the base in harmony with the robotic manipulator while executing a task on a large surface (larger than the manipulator workspace can cover from a fixed position) and maintains a high level of dexterity index for the manipulator.