Transcript
Research recently published in Science Robotics represents a big step toward real-world implementation of robotic exoskeletons designed to help humans with walking and doing physically demanding work. Such technology has appeared in sci-fi stories for decades. You may remember Ellen Ripley using a “Power Loader” in Alien and George McFly wearing a mobile platform in Back to the Future, Part II.
As this new research documents, engineers are already working on real-life robotic assistance technology that could protect workers from painful injuries and help stroke patients regain their mobility. So far, such devices have required extensive calibration and context-specific tuning, which has kept them largely limited to research labs. However, mechanical engineers at Georgia Tech may be on the verge of changing that, allowing exoskeleton technology to be widely deployed in homes, workplaces, and more.
The team of researchers developed a universal approach to controlling robotic exoskeletons that requires no training, no calibration, and no adjustments to complicated algorithms. Instead, users can simply don the “exo-system” and go. The “exo-system” uses a kind of artificial intelligence called deep learning to autonomously adjust how the exoskeleton provides assistance. And they’ve shown it works seamlessly to support walking, standing, and climbing stairs or ramps.
The goal was not simply to provide control across different activities, but to create a single unified system. Users of this system don’t have to press buttons to switch between modes or have a classifier algorithm that tries to predict that you’re climbing stairs or walking. Previous work in this area has focused on one activity at a time.
On the other hand, the Georgia Tech team focused on what’s happening with muscles and joints, which meant the specific activity didn’t matter. They quit trying to categorize human movement into discrete modes such as “level ground walking” or “climbing stairs.” Instead, they based their controller on the user’s underlying physiology. They assumed that what the body is doing at any point in time was telling them everything they needed to know about the environment.
Then they used machine learning as the translator between what the sensors measured on the exoskeleton and what torques the muscles were generating. What’s so great about this approach is that there’s no need for subject-specific tuning or changing parameters to make it work. It adjusts to each person’s internal dynamics without any tuning or heuristic adjustments. That leads to a lot less work in the field. The control system in this study was used with partial-assist devices.
These exoskeletons support movement rather than completely replacing the effort. Importantly, this study is “application agnostic.”
As such, it offers the first bridge to real-world viability for a range of exoskeleton devices. Imagine how robotic assistance could benefit soldiers, airline baggage handlers, or any workers doing physically demanding jobs where musculoskeletal injury risk is high.
|
