Neuromuscular Rehabilitation Immediately After Stroke
The Barrett WAM arm
Researchers collaborating at the Rehabilitation Institute of Chicago (RIC) (Sandro Mussa-Ivaldi, James Patton, Michael Peshkin, and Zev Rymer of Northwestern University collaborating with Prof. Robert Kenyon at University of Illinois, Chicago) and a related but alternative approach at the University of Washington (led by Prof. Yoky Matsuoka) are using WAMs to perform highly effective stroke recovery during the critical first six months (acute stage) in the aftermath of a stroke.
After a person suffers a stroke, reaching for common objects like a cup of coffee can become difficult or impossible. As the person attempts to reach, the nervous system no longer perceives trajectory errors during reaching. Correcting immediately afterwards results in spilled cups and similar frustrations.
For decades physical and occupational therapists held a patient’s arm to help guide them to the cup (or other task goal). The moral boost from task achievement notwithstanding, improvements appeared to be short lived.
Then researchers at RIC tried a counterintuitive approach that would challenge the best therapist. The patient grabs a gimbaled handle attached to the end of a WAM™ arm and is then encouraged to attempt to reach for the cup of coffee. As the reach is initiated, the WAM™ arm measures the (initially small) error between the actual and ideal path. It then amplifies this error by generating a lateral force away from the path in proportion to the error. The first time reaching for the cup, a patient’s hand is forced grossly off path.
While this procedure might seem like a mean trick, it addresses the root problem, in which the patient’s neuromuscular system has lost the ability to sense moderate errors. By amplifying the error, the patient senses the error early along the trajectory so that the neuromuscular system can begin to rewire itself to compensate.
To avoid limiting learning only in one direction, the patient is immersed in a 3-D graphically virtual environment. Then targets (like a virtual cup of coffee) are generated at random locations around the patient, who then operates in a video-game world attempting to find and reach for each new target.
The results have been striking. Not only can this game be fun, early clinical trials suggest that the neuromuscular effect may be long lasting.
Prof. Yoky Matsuoka at the University of Washington is an expert in using the WAM™ arm and has achieved similar results by generating the errors in the visual field. The enabling tool in common between the two teams (RIC and University of Washington) is the WAM™ arm.
Affordable Assisted Living Enabled Through Mobile Dexterity
The Barrett WAM arm
We all grow older. At some point, age, injury or disease create challenges for many in accomplishing such basic tasks as preparing meals, bathing, and using the toilet. Consider for a moment how it might feel to rely on hired help to lower you to a toilet seat. The challenge to one’s independence and dignity* is enormous.
However, there is no loss of dignity if you command a robot to extend an arm to secure your balance. The robot is a sophisticated tool and, like a car, is an extension of the person.
Since it is not economically feasible to have different robotic manipulators in each room of a house, the concept of mobile manipulation is essential. Until recently, through, the notion of mobile manipulation was impractical. Although there are several mobile platforms available today, robotic arms are poorly designed for mobile manipulation. They have high moving inertia that can tip a mobile platform; generate bone-breaking forces, do badly in the face of a collision because their joints are not adequately backdrivable, and require a substantial volume of support electronics stored in a “controller cabinet” that rival the size and mass of the robot itself. Most importantly, these robots draw an order of magnitude too much power to be practical and the type of power they require is generally incompatible with the variable-DC reality of batteries that jump from deep discharge to high charging voltage once each charge cycle.
Barrett’s engineers have spent over 20 years overcoming these obstacles in the design of the WAM™ arm and its array of support technologies, like its ultra-miniature, ultra-power-efficient Puck controller (patents pending worldwide).
The fundamental cost of the WAM™ in anticipated quantities would be less than that for any conventional geared or harmonic-drive robot because of the fundamental simplicity of the cable drives and Puck servoelectronics. And the ability to finance the household purchase of a mobile manipulator is not far fetched for any cost less than ~US$50,000 by leveraging the car-financing model with monthly payments of a few hundred dollars. In many ways a household mobile-manipulator is a natural follow-on from owning a car. Furthermore, the need for living assistance comes at the same point in life when continued use of a car becomes hazardous, so there is a natural trade-up in mobility tools.
* The importance of dignity came to Barrett by way of Joseph Engelberger in the 1990s. Joseph Engelberger is the father of the robotics industry.