Effect of Neural Network at Server on Robot Position Control Using Force Information

This paper investigates the effect of employing a neural network at a server for Quality of Service (QoS) control on cooperative work between two remote robot systems with force feedback through experiment. In the work, two robot arms cooperate to grasp and carry an object together. As QoS control, we address the robot position control using force information, which we previously proposed. This involves transmitting force information from the robot systems to the server to compute the optimal position adjustment value for the robots using a neural network located at the server based on the force information. The value is then applied to the robot systems. Additionally, the influence of the network delay between the robot systems and the server on the work is investigated. Experimental results demonstrate that using the server is superior to no control (i.e., without QoS control) for network delays of up to approximately 50 milliseconds.