Experimental Evaluation of a User-Centered Unilateral Telemanipulation System for Training and Control in Robotic Industrial Scenarios

This paper presents a unilateral teleoperation system for robotic manipulators, designed to enable intuitive and effective control in complex industrial tasks. The system features a haptic interface that allows users to operate in a hybrid control mode - switching fluidly between direct control (position to position mapping) and rate control (position to velocity mapping) - with force feedback providing intuitive cues on activation and intensity of the rate control command. A Graphical User-Machine Interface (GUMI) is developed to offer camera views and real-time 3D visualization of the input from the haptic device according to its workspace. Furthermore, interactive commands are implemented for robot orientation regulation, gripper closing/opening and system state monitoring (e.g.: mapping activation, mapping deactivation, orientation regulation mode, homing procedure). This ensure intuitiveness and accessibility in a user-centered sense, also for operators without advanced robotics expertise. The proposed framework also provides a simulation environment based on MuJoCo, used for training through simplified tasks, and a real-world setup with the physical robot, ensuring users can practice and familiarize with the system before operating in real scenarios. We provide a detailed description and validation of the system, reporting on an experiment which reproduced a robotic wiring task in a laboratory setting. Results show robust system behavior and demonstrate the effectiveness of the hybrid control strategy, highlighting its potential for industrial applications such as remote operation and programming by demonstration in high-skill, weakly-structured environments.