A Rapid Control Prototyping Approach for the Design of a Trunk SEA-Actuated Rehabilitative Exoskeleton

This work was funded by the Istituto Nazionale per l’Assicurazione contro gli Infortuni sul Lavoro (INAIL) under grant agreements "PR23-RR-P2 – ClinicEXO. Robotic-assisted rehabilitation has emerged as a promising approach in the field of physiotherapy. In particular, the use of exoskeletons is spreading, as they offer full support to the patient's body, relieving the physical burden on the physiotherapist, while increasing the intensity of the exercises. Additionally, it collects useful data on the patient's condition and his improvements. Currently, in the literature, is difficult to find trunk rehabilitation exoskeletons, although trunk rehabilitation is crucial in posture maintenance, balance, and mobility, which are key elements of many Activities of Day Living (ADLs). Indeed, impairments of the trunk can severely affect the patient’s quality of life, limiting his independence and increasing the risk of psychological issues, such as depression. The thesis focuses on the development of a trunk rehabilitation exoskeleton (TRUNKY) aimed at treating stroke, incomplete lumbar-thoracic spinal cord injury, and Pusher syndrome. The work covers joint testing, electrical setup, machine assembly, and control. Starting from a CAD design and leveraging existing electronics and control systems of previous exoskeletons, a Rapid Control Prototyping (RCP) approach with Hardware in the Loop (HIL), has been used to speed up the process. All the control algorithms have been developed on Simulink, exploiting the Series Elastic Actuators (SEAs) joints, to measure the external torque applied to every joint and obtain a more compliant mechanism that could better adapt to the patient’s needs. Which is very important in the rehabilitation field, as this helps to reduce the possibility of further injury and to follow all the steps of the rehabilitation, from beginning to end. In order to follow the full process of rehabilitation with this exoskeleton, three control algorithms have been developed: position, impedance, and resistance. The position control is used during early rehabilitation, to passively guide the user along the exercise trajectory. Impedance control provides assistance as needed, allowing the subject to actively control the motion within certain limits. While Resistance Control is useful in the last stage of recovery, to strengthen the muscle, by opposing a torque to the user's movements. Furthermore, all the control algorithms' parameters are tunable in real time, to allow doctors to adjust them according to the clinical needs. Additionally, a Graphical User Interface has been developed to ease the control of the exoskeleton, allowing users to easily manage the execution of exercises and ensuring its accessibility. After having completed the first part of testing, we built the actual exoskeleton to test and validate its functionality. In order to rigidly connect the user to the end effector of the device and ensure proper execution of the exercises, a supportive corset was designed. Then using Electromyography (EMG) sensors placed on the muscle of interest for the trunk, we measured muscle activity in different exercises with and without the exoskeleton, to try to understand the level of efficacy of the exoskeleton by comparing them. In conclusion, we successfully developed a functional prototype of a trunk rehabilitation exoskeleton, that can be used to treat different neurological patients.