Characterization of Multiple Sensors to Operate a Bionic Joint for Rehabilitation Applications

To facilitate the rehabilitation exercises for patients to regain their mobility, an Assistive Bionic Joint (ABJ) device was developed at SJSU. The ABJ device operates on various sensors, including electromyography (EMG) sensors, inertial measurement unit (IMU) sensors, and rotary encoders. EMG sensors are crucial to the success of the ABJ, enabling precise tracking of muscle activity to initiate a human walking sequence. This was executed via a control system, consisting of an Enfield LS-V15S proportional directional valve, a D1 bidirectional proportional PWM valve driver, and an Omega PX309 transducer. The ABJ will respond more precisely to EMG data initiation through further tuning. After extensive trials, it was determined that the sensors placed on the Biceps femoris, the Gastrocnemius lateralis, and the Tibialis anterior provided the muscle data most precisely. In this work, attempts are made to optimize the signals from the EMG sensors as they are placed along the leg muscle. This was achieved by varying the locations of the EMG triplets along the leg that produce electrical signals as a function of muscle tension. After an optimized signal was obtained, a filtering system was employed through the Arduino microcontroller to eliminate noise from the signal, which was subsequently used to trigger the main controller. The EMG sensors were placed on a human leg, whereas the ABJ was mounted on a mannequin leg that moved via a control system activated by the EMG signal. In addition, IMU sensors and rotary encoders mounted on the mannequin leg were tested and optimized. By integrating these sensors with the fluidic muscle system of ABJ, the prototype can provide users with effective knee rehabilitation.