This exoskeleton uses machine learning to put a personalized spring in your step

This exoskeleton uses machine learning to put a personalized spring in your step

Exoskeletons have been largely confined to the realm of fiction, appearing in science fiction or superhero movies to make characters stronger, bigger, or more destructive (in James Cameron’s Avatar, the somewhat terrifying AMP suit serves as a “human operator’s amplifier”, but looks more like a humanoid war machine with a real human inside). In terms of real-world uses, exoskeletons have been tested or developed in industries such as automotive, airline, military, and healthcare; these are primarily intended to help people lift heavy objects and materials.

A new exoskeleton serves a different purpose: to help people walk. Developed by engineers at the Stanford Biomechatronics Laboratory, the device is described in an article published this week in Nature. In a nutshell, it’s a motorized boot that gives wearers a forward thrust with every step they take. What sets it apart, however, is that its function is tailored to each person who uses it rather than being standard for different heights, weights and walking speeds.

“This exoskeleton personalizes assistance as people walk normally in the real world,” Steve Collins, an associate professor of mechanical engineering who directs Stanford’s Biomechatronics Lab, said in a press release. “And it resulted in outstanding improvements in walking speed and energy saving.”

The personalization is made possible by a machine learning algorithm, which the team trained using emulators, i.e. machines that collected data on the movement and energy expenditure of volunteers who were connected to it. The volunteers walked at varying speeds in imaginary scenarios, such as trying to catch a bus or walking through a park.

The algorithm made connections between these scenarios and people’s energy expenditure, applying the connections to learn in real time how to help wearers walk in a way that actually benefits them. When a new person puts on their boot, the algorithm tests a different assist pattern each time they walk, measuring how their movements change in response. The learning curve is short, but on average the algorithm was able to effectively adapt to new users in just an hour.

The exoskeleton works by applying torque to the ankle, replacing some of the function of the wearer’s calf muscle. When users take a step, just before their toes leave the ground, the device helps them push off. It worked pretty well; on average, people walked 9% faster than usual while expending 17% less energy. In head-to-head comparisons on a treadmill, the exoskeleton provided about twice the effort reduction of similar devices.

Reducing the effort required to walk isn’t usually something most of us should be aiming for; if anything, Americans need the opposite. But the team that developed the exoskeleton sees it being used to help people with limited mobility, including the elderly or disabled.

“I think over the next decade we will see these ideas for personalized assistance and effective wearable exoskeletons help many people overcome mobility issues or maintain their ability to live active, independent and meaningful lives. “, said Patrick Slade, study author and bioengineering researcher. in a press release.

Since the exoskeleton is currently in the prototype stage, it won’t be reaching a wider user base anytime soon. Also, so far it’s only been tested on healthy adults in their mid-twenties, so further testing should be done and adjustments made for people who really need help with walk.

The team also plans to design iterations that help improve wearers’ balance and even reduce joint pain. They are optimistic about the potential of their device. “I really think this technology is going to help a lot of people,” Collins said.

Image Credit: Stanford University/Kurt Hickman

Similar Posts

Leave a Reply

Your email address will not be published.