Toward Upper-Body Posture Monitoring for Upper-Limb Rehabilitation Robot

This paper proposes an improved upper-body monitoring method using an end-effector rehabilitation robot and an RGB-D sensor. To enhance the monitoring method, we conducted tests in different robotic environments and utilized the Jacobian to detect sensor inaccuracies. The robotic system was replaced with more commonly used structure and sustainable development. Additionally, we proposed a method for detecting sensor inaccuracies and confirmed it by predicting joint angular velocity using the Jacobian.The proposed method showed improved root mean squared error compared to the elbow joint of the sensor, even in different robotic environments.