Exoskeletons–wearable, motorized machines that can assist a person’s movements–have largely been confined to movies or military use, but recent advances might soon bring the devices to the homes of people with paralysis.
So far, exoskeletons have been used to augment the strength of soldiers or to help hospitalized stroke patients relearn how to walk. Now researchers at the University of California, Berkeley, have demonstrated an exoskeleton that is portable and lets paraplegics walk in a relatively natural gait with minimal training. That could be an improvement for people with spinal-cord injuries who spend a lot of time in wheelchairs, which can cause sores or bone deterioration.
Existing medical exoskeletons for patients who have lost function in their lower extremities have either not been equipped with power sources or have been designed for tethered use in rehabilitation facilities, to correct and condition a patient’s gait.
In contrast, the Berkeley exoskeleton combines "the freedom of not being tethered with a natural gait," says Katherine Strausser, PhD candidate and one of the lead researchers of the Berkeley project. Last week at the 2010 ASME Dynamic System and Control Conference in Cambridge, Massachusetts, Strausser presented experimental results from four paraplegics who used the exoskeleton.
Other mobile exoskeletons–like those developed by companies such as Rex Bionics or Cyberdene–don’t try to emulate a natural gait, Strausser says. Because walking is a dynamic motion that is essentially falling forward, Strausser says, many designs opt for a shuffle instead of a natural gait, because "it’s safer and a lot easier." However, emulating a natural gait mimics the efficiency of natural walking and doesn’t strain the hips, Strausser says.
The Berkeley device, which houses a computer and battery pack, straps onto a user’s back like a backpack and can run six to eight hours on one charge. Pumps drive hydraulic fluid to move the hip and knees at the same time, so that the hip swings through a step as one knee bends. The device plans walking trajectories based on data (about limb angles, knee flexing, and toe clearance) gathered from people’s natural gaits. Pressure sensors in each heel and foot make sure both feet aren’t leaving the ground at the same time.