Nonlinearity Compensation in a Multi-DoF Shoulder Sensing Exosuit for Real-Time Teleoperation

The compliant nature of soft wearable robots makes them ideal for complex\nmultiple degrees of freedom (DoF) joints, but also introduce additional\nstructural nonlinearities. Intuitive control of these wearable robots requires\nrobust sensing to overcome the inherent nonlinearities. This paper presents a\njoint kinematics estimator for a bio-inspired multi-DoF shoulder exosuit\ncapable of compensating the encountered nonlinearities. To overcome the\nnonlinearities and hysteresis inherent to the soft and compliant nature of the\nsuit, we developed a deep learning-based method to map the sensor data to the\njoint space. The experimental results show that the new learning-based\nframework outperforms recent state-of-the-art methods by a large margin while\nachieving 12ms inference time using only a GPU-based edge-computing device. The\neffectiveness of our combined exosuit and learning framework is demonstrated\nthrough real-time teleoperation with a simulated NAO humanoid robot.\n