Scientists have showcased a fresh approach using AI and computer simulations to teach robotic exoskeletons to assist users in conserving energy during various activities such as walking, running, and climbing stairs. The team at North Carolina State University are behind this innovative method, which aims to autonomously improve energy efficiency for the users. The researchers are introducing a machine-learning framework that connects the gap between traditional control approaches and human-in-the-loop optimization for these robotic devices.Hao Su, the corresponding author of a paper on the work that will appear in the journal Nature on June 12, stated that “simulation and reality to autonomously control wearable robots to improve mobility and health of humans.” Su, who is an associate professor of mechanical and aerospace engineering at North Carolina State University, also mentioned that “Exoskeletons have enormous potential to improve human locomotive performance.” However, their development and broad dissemination are limited by the requirement for lengthy human tests and handcrafted control laws. The key idea here is that the embodied AI in a portable exoskeleton is learn.Su explains, “We are learning how to assist people in walking, running, or climbing using a computer simulation, without the need for any physical experiments.” The researchers specifically concentrated on enhancing the autonomous control of embodied AI systems, which are AI programs integrated into physical robot technology. Their focus was on teaching robotic exoskeletons how to support able-bodied individuals in various movements. Typically, users have to spend hours training the exoskeleton so that it can determine the necessary force and when to apply it to help with walking, running, or climbing stairs. However, the new method allows users to utilize the exoskeleton without extensive training.
“This work is essentially turning science fiction into reality — enabling people to use less energy while performing various tasks,” Su says.
“We have created a method for training and controlling wearable robots to directly benefit humans,” says Shuzhen Luo, the primary author of the paper and a former postdoctoral researcher at NC State. Luo is currently an assistant professor at Embry-Riddle Aeronautical University.
For instance, in experiments with human subjects, the scientists discovered that study participants used 24.3% less metabolic energy when walking in the robotic exoskeleton than w
Without the exoskeleton, participants used 13.1% more energy when running and 15.4% more energy when climbing stairs. This means that the robotic exoskeleton reduced energy expenditure significantly.
It’s important to highlight that these energy savings are based on a comparison between the exoskeleton and a user not wearing one. This provides an accurate measure of the exoskeleton’s energy-saving capabilities.
Although the study primarily focused on able-bodied individuals, the findings can also be applied to robotic exoskeletons designed to assist those with mobility impairments.
The framework developed by the researchers may have widespread and adaptable applications for robotic exoskeleton technology.The plan is to quickly develop and widely implement various assistive robots for both able-bodied and mobility-impaired individuals,” says Su.
“We are currently testing the new method’s effectiveness in robotic exoskeletons used by older adults and individuals with neurological conditions like cerebral palsy. We also want to see how the method could enhance the performance of robotic prosthetic devices for amputees.”
This study was supported by the National Science Foundation under awards 1944655 and 2026622, as well as the National Institute on Disability, IndependThe research was funded by the National Institute on Disability, Independent Living, and Rehabilitation Research, and the National Institutes of Health. Shuzhen Luo and Hao Su have intellectual property related to the controller, and Su is also involved with Picasso Intelligence, LLC, a company that develops exoskeletons.
