A novel compliant ankle rehabilitation robotic system based on a 6-DOF robot arm

To address the limitations of conventional ankle rehabilitation robots in accuracy, safety, and quantitative assessment, this study proposes a novel robot arm-based system integrating compliance control. We developed a novel robotic system using a serial 6-degree-of-freedom (6-DOF) robot arm. By defining the ankle joint center (AJC) as the midpoint between the medial and lateral malleolus, a coordinate transformation method dynamically coincides the robotic rotational center (RRC) with the AJC, validated through a novel alignment accuracy metric. A real-time gravity compensation algorithm based on a 6-axis force/torque (F/T) sensor was designed to quantify ankle joint torques. The measured torques were mapped to Modified Ashworth Scale (MAS) criteria to enable quantitative muscle tension assessment, improving objectivity over manual evaluations. Three compliance rehabilitation training modes, namely patient-passive training, patient-active training, and resistance training, are designed based on the admittance control strategy, which also support simultaneous multi-axis compliance training. The experimental results on five subjects show that the designed training modes performed well in terms of compliance and trajectory tracking. The alignment accuracy for all three training modes was less than 1.59 mm, as validated by a motion capture system. The system successfully recorded and classified torque profiles indicative of different levels of muscle tension. The integration of high-precision sensing and adaptive control provides a standardized measurement platform for ankle rehabilitation, promoting the translation of robotic systems into clinical practice.