Undergraduate Research and Creative Activities Journal

Stroke recovery is a difficult process, so there are many forms of robot-assisted therapy (RAT) that seek to make it easier for patients and physical therapists. However, machines designed for this type of therapy are often expensive unitaskers that limit their therapy assistance to only one part of a limb. We have developed a series of prototypes that have the potential to aid in drawing therapy for stroke rehabilitation and assessment, as drawing can engage every muscle group in the arm and is often used as a method of estimating limb and neural pathway function. Our current focus is refining a system of surface electromyography (sEMG) and internal motion unit (IMU) sensors processed via machine learning to quantify limb function and location in order to assist the user in creating their drawings. We believe this robot has the potential to be incredibly useful to artists with unsteady hands, physical therapists, and physical therapy patients.