Fault detection of tool/load grasping for telerobotics using neural networks

For the safe and reliable execution of tasks, the tool grasping conditions of the manipulator must be checked to determine whether the tool has been grasped in the desired manner in real time. Especially in the case of telerobotics, grasping errors are critical to the completion of tasks since the human operator cannot access the hazardous and remote work environment. This paper proposes a time-delayed neural network to identify the load of manipulators in real time. The developed scheme is applied to a two-link manipulator, and the simulation results show the feasibility of the approach for grasping fault detection