Pick the Right Co-Worker: Online Assessment of Cognitive Ergonomics in Human-Robot Collaborative Assembly

Marta Lagomarsino,Marta Lorenzini,Pietro Balatti,Elena De Momi,Arash Ajoudani

DOI: https://doi.org/10.1109/TCDS.2022.3182811

2022-07-08

Abstract:Human-robot collaborative assembly systems enhance the efficiency and productivity of the workplace but may increase the workers' cognitive demand. This paper proposes an online and quantitative framework to assess the cognitive workload induced by the interaction with a co-worker, either a human operator or an industrial collaborative robot with different control strategies. The approach monitors the operator's attention distribution and upper-body kinematics benefiting from the input images of a low-cost stereo camera and cutting-edge artificial intelligence algorithms (i.e. head pose estimation and skeleton tracking). Three experimental scenarios with variations in workstation features and interaction modalities were designed to test the performance of our online method against state-of-the-art offline measurements. Results proved that our vision-based cognitive load assessment has the potential to be integrated into the new generation of collaborative robotic technologies. The latter would enable human cognitive state monitoring and robot control strategy adaptation for improving human comfort, ergonomics, and trust in automation.

Robotics,Human-Computer Interaction

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to evaluate the operator's cognitive load online in the human - machine collaborative assembly system. Specifically, with the application of industrial collaborative robots on the production line, although work efficiency and physical ergonomics are improved, the cognitive demands on operators may be increased, thus affecting their health and safety. Therefore, the paper proposes a vision - based method, using low - cost RGB - D cameras and advanced artificial intelligence algorithms (such as head pose estimation and skeleton tracking) to monitor the operator's attention distribution and upper - body movements in real - time, in order to quantify the workload caused when interacting with human operators or industrial collaborative robots under different control strategies. This method aims to improve human comfort, ergonomics and trust in automation by monitoring the human cognitive state and adjusting the robot's control strategy according to this information.