Ozgur Ege Aydogan

Role: Undergraduate Research Assistant and Team Leader

In my undergraduate research, as the team leader of a team of 3 undergraduate students, I developed the adjustable mechanical structure and electronic hardware of a 4-DOF grounded robotic exoskeleton for shoulder and elbow rehabilitation. Based on the average height of adults, I designed the exoskeleton, which can be easily used by people with a height of 150 - 200 cm, with an adjustable length of 200 - 300 mm for the forearm and 250 - 400 mm for the upper arm.  I designed the mechanical structure to provide three active shoulder movements and one passive elbow movement: shoulder vertical flexion-extension, shoulder horizontal flexion-extension, shoulder abduction/adduction, and elbow flexion-extension.

I developed a low-cost EMG sensor and integrated it into the control system to measure muscular activation for feedback and instantaneous muscle activation control for the physiotherapist during the therapy. I implemented sophisticated signal processing algorithms to extract meaningful insights from the collected EMG data. More precisely, I utilized time-domain analysis to identify patterns of muscular activation and frequency-domain analysis to discern nuances in muscle activity. I used a full-wave rectifier, an instrumentation amplifier, a high-pass filter, and a low-pass filter are used for the rectification, amplification, and filtering steps. Furthermore, I developed a user-friendly GUI, where patient information can be entered and exercise types and movement selections can be made. 

The robot was developed to perform 4 different therapeutic exercises: passive, active-assistive, isometric, and isotonic. While the PID control method was used for passive exercise, I developed a position and force-based impedance control system to ensure efficient human-robot interaction in active-assist, isometric modes, and isotonic exercise. As a result, a rehabilitation robot that is cost-effective, appropriate for home usage, can perform 4 different exercises and has a user-friendly human-machine interface was developed.