Preventing and Mitigating Repetitive Strain Injury with an Intuitive Hand Recognition Interface for Multi-Robot Teleoperation

• Lucas Alexandre Zick UDESC
• Dieisson Martinelli UTFPR
• André Schneider de Oliveira UTFPR
• Vivian Cremer Kalempa UDESC

Abstract

As robotics increasingly integrates into technical and industrial fields, seamless and accessible human-robot interactions become crucial. Current teleoperation models pose significant challenges for operators with conditions such as Repetitive Strain Injury (RSI), leading to increased risks of musculoskeletal injuries and potentially decreased productivity. This article analyzes the benefits of an innovative palm recognition-based interface in developing more assistive robotic teleoperation. By utilizing the MediaPipe Python library for gesture recognition and the TurtleBot 3 robot for navigation, the system minimizes physical strain through simple hand movements and autonomous navigation. Testing in simulated environments using Gazebo and RViz, coordinated by the Robot Operating System (ROS), demonstrated, compared to a traditional teleoperation method, a 6% reduction in total time, a 100% reduction in the incidence of pain and discomfort among operators, and an increase in free time by about 95%, highlighting the potential of the solution to achieve more ergonomic human-robot interactions.

Keywords: Pain, Navigation, Robot kinematics, Prevention and mitigation, Human-robot interaction, Gesture recognition, Robots, Strain, Context modeling, Testing, autonomous navigation, human-robot interaction, multi-robot systems, gesture-assisted teleoperation, ergonomic support