Walk This Way
This robotic exoskeleton uses a Raspberry Pi to help stroke survivors relearn how to walk as they go about their normal, daily routines.
Strokes occur when blood flow to a part of the brain is interrupted or reduced, depriving brain tissue of oxygen and nutrients. This medical emergency can lead to serious and sometimes permanent neurological damage, affecting a number of bodily functions. Millions of people worldwide experience a stroke each year, so effective interventions are sorely needed to improve the lives of these individuals.
One of the most common challenges faced by stroke survivors is difficulty walking. It is estimated that around 80 percent of stroke survivors experience some degree of mobility impairment, which can severely impact their independence and quality of life. Many of these individuals develop walking asymmetries, where one step is longer or shorter than the other, leading to an uneven gait.
To address these asymmetries and help stroke survivors regain functional mobility, rehabilitation often involves specialized interventions such as split-belt treadmills. These treadmills consist of two side-by-side belts, one for each foot, which can move at different speeds. By gradually adjusting the speeds of these belts to encourage symmetrical stepping patterns, patients can relearn to walk with a more balanced gait.
However, while split-belt treadmill training can be effective in a controlled rehabilitation setting, the challenge lies in transferring these learned skills to real-world situations. A novel assistive device developed at the University of Massachusetts Amherst may help to make this transition much easier. The device created by the researchers takes the form of a portable robotic exoskeleton that can provide gait corrections to the wearer as they walk under normal conditions. The hope is that these corrections can be gradually reduced until the exoskeleton is no longer needed.
With the goal of reteaching lost skills, the exoskeleton does not simply give a boost to the lagging leg or slow the leg taking a longer step. Rather, it works to exaggerate the asymmetrical gait that has developed. This encourages the stroke survivor to modify the way that they walk to compensate. Accordingly, they relearn a more normal, symmetrical walking pattern by working against the even more abnormal pattern induced by the robot.
The hip-worn exoskeleton is a custom robot made by the Human Robot Systems Laboratory at the University of Massachusetts Amherst. The device weighs in at a bit over six pounds, and is strapped around the waist like an oversized belt. Extensions reach down to the thighs, where they are attached to enable the system to apply torque to the hip joints. Actuation is supplied by brushless DC motors, with encoders and other sensors to provide feedback. This data is fed into a Raspberry Pi 4 single board computer that orchestrates the exoskleleton’s actions.
Throughout the course of this study, the power source and the computer were located offboard. Before such a system could be practically integrated into a user’s daily activities, these functions would need to be incorporated into the exoskeleton itself. That should be a simple matter on the computing side, given the diminutive size of the Raspberry Pi, however, supplying sufficient battery power for the robot, while maintaining comfort and wearability, could be a challenge.
To date, the device has only been tested on a small cohort of unimpaired individuals. These experiments showed that the exoskeleton can provide similar forces on the legs as a split-belt treadmill, but with the obvious benefit of allowing the users to use the device in the course of their daily activities. But to prove that this translates into real-world results for actual stroke survivors will require further studies, which the team intend to carry out in the future.