The Robotic Parallel Bars (RPB) device is a low-cost, high data acquisition platform that therapists can use to assist ambulatory rehabilitation and perform gait studies. The RPB’s support bars track the amount of force the patient places on the device. The RPB is also self-propelled, meaning that the patient does not have to physically move the support device as they would with a traditional walker or rollator. The RPB offers more flexibility than stationary parallel rehabilitation bars because the device follows the movement of the patient wherever they may walk and turn. In this study, improvements were made to the RPB’s controller program to reduce potentially hazardous and erroneous behavior of the device. A Proportional and Derivative (PD) controller replaced the previous state space controller. The PD controller was implemented and tested by using a webcam to track the patient and the RPB at three different patient walking speeds. Experimental results showed that the PD controller decreases the linear error between the RPB and the patient by an average of 4.2 cm as the patient walks forward. Improving the control response of the device allows it to be more effective in tracking the human user’s movements, which improves its ability to record data, and act as a functional support for the human user.
