Active Motor Control for an Upper Extremity Exoskeleton

An active control system for a five degree of freedom (5-DOF) upper limb robotic exoskeleton was developed to fill in the need for an accessible and cost-effective rehabilitation system. Since physical therapy sessions are labor intensive, it aims to be a tool in augmenting the capacity of local rehabilitation centers in catering more patients. Many different control systems were already done on similar devices but the challenges remain in adapting to the limits of the mechanical design of the exoskeleton it is implemented on and availability of local electronic components. Arduino Microcontroller was used as the embedded platform in the implementation of the control system. Using Myoware electromyography (EMG) sensor, signals were measure from biceps and triceps for elbow flexion and extension; deltoids and teres major for shoulder abduction and adduction. The rectified signal was further smoothened thru running average. Experimental results showed that baseline EMG and activation level of the target muscle groups were different from each other, thus different threshold levels were established. An adaptive algorithm using pulse width modulation (PWM) was also implemented in varying the supplied power depending on the applied load on the arm. Stall conditions via the current sensor was monitored and used as positive feedback. The system was successful in interpreting intent to move and translating it to motor movement.