Researchers at North Carolina State University have unveiled a pioneering approach that harnesses the power of artificial intelligence (AI) and computer simulations to train autonomous robotic exoskeletons to assist users optimize their energy expenditure while walking, running, and navigating stairs.
“This study introduces a pioneering machine-learning framework that seamlessly connects simulations with reality, enabling autonomous control of wearable robots to boost individuals’ mobility and well-being,” notes Hao Su, co-author of a forthcoming paper in the journal.
According to Su, an affiliate professor of mechanical and aerospace engineering at North Carolina State University, exoskeletons hold immense promise for amplifying human mobility performance. Despite their potential to improve and disseminate widely, their progress is hindered by the need for extended human oversight and customized policy development.
“According to Su, the key takeaway is that embodied AI in a portable exoskeleton is learning how to assist people walk, run, and climb through a computer simulation, without needing any actual experiments.”
Researchers focused specifically on enhancing autonomous control of embodied AI systems – those where AI programming is seamlessly integrated into physical robotics technology. Researchers focused on developing robotic exoskeletons capable of assisting able-bodied individuals in performing a range of activities. Typically, users must dedicate considerable time to “calibrating” an exoskeleton, allowing it to learn their unique energy requirements – including when to deploy that energy – in order to effectively aid walking, running, or navigating stairs. This innovative methodology empowers customers to seamlessly utilize the exoskeletons from the onset.
“This groundbreaking research is bringing science fiction to life, enabling people to conserve energy and accomplish multiple tasks with ease,” says Su.
According to Shuzhen Luo, lead author and former postdoctoral researcher at North Carolina State University, their team has created a method to prepare and manage wearable robots that directly benefit individuals. Luo is currently an assistant professor at Embry-Riddle Aeronautical University.
Researchers found that individuals using a robotic exoskeleton expended 25.7% less energy while walking compared to walking without the device, suggesting potential benefits for those requiring mobility assistance. Studies revealed a notable reduction in energy expenditure among members, with a 13.1% decrease when operating within the exoskeleton’s confines, and a 15.4% decline when navigating stairs.
“It’s crucial to recognize that these vitality reductions benchmark the robotic exoskeleton’s performance against a non-exoskeleton-wearing individual, notes Su.” “Indeed, it serves as a definitive indicator of the considerable vitality that the exoskeleton conserves.”
While the original study focused on able-bodied participants, its innovative approach can also be applied to robotic exoskeletons designed to assist people with mobility impairments, expanding its potential impact.
“Our framework offers a scalable and generalizable method for accelerating the development and global uptake of assistive robots, benefiting both able-bodied and mobility-impaired individuals.”
We’re currently conducting initial tests to assess the efficacy of our innovative approach in enhancing robotic exoskeletons’ performance among older adults and individuals with neurologically-based conditions, such as cerebral palsy. We enthusiastically endorse investigating how this strategy could improve the effectiveness of robotic prosthetic devices for individuals with amputations.
The analysis was completed with assistance from the National Science Foundation under awards 1944655 and 2026622, the National Institute on Disability, Independent Living, and Rehabilitation Research under grant 90DPGE0019, the Switzer Research Fellowship SFGE22000372, and the National Institutes of Health under award 1R01EB035404.
Shuzhen Luo and Hao Su share inventorship rights related to the intellectual property tied to the controller described in this research. Su is also a co-founder of and holds a significant financial stake in Picasso Intelligence, LLC, a company that specializes in the development of innovative exoskeleton technology.
