Summary: A new study details how artificial intelligence and computer simulations train robotic exoskeletons to help users conserve energy while walking, running and climbing stairs. This method eliminates the need for lengthy human experiments and can be applied to a variety of assistive devices.
This breakthrough offers significant potential to help those with mobility problems and improve accessibility in everyday life. The researchers found that participants used up to 24.3% less energy on the exoskeleton.
Key facts:
- Artificial intelligence and simulations train exoskeletons without human experimentation.
- Exoskeletons helped users save up to 24.3% energy in movement tests.
- The method can be applied to various assistive devices, including prosthetics.
Source: New Jersey Institute of Technology
A team of researchers has demonstrated a new method that uses artificial intelligence and computer simulations to train robotic exoskeletons that can help users conserve energy while walking, running and climbing stairs.
Described in a study published in NatureThis new method rapidly develops exoskeleton controllers to aid locomotion without relying on lengthy human experiments.
In addition, the method can be applied to a wide variety of assistive devices beyond the hip exoskeleton demonstrated in this research.
“It may also apply to knee or ankle exoskeletons or other multi-joint exoskeletons,” said Xianlian Zhou, associate professor and director of NJIT’s BioDynamics Lab.
Additionally, it can be similarly applied to above-the-knee or below-the-knee prostheses, providing immediate benefits for millions of healthy and mobility-impaired individuals, he said.
“Our approach represents a significant advance in wearable robotics, as our exoskeleton controller is developed entirely through artificial intelligence-driven simulations,” explains Zhou. “Additionally, this driver seamlessly transitions to hardware without the need for additional human testing, so it’s experiment-free.”
The breakthrough holds promise for helping individuals with mobility issues, including the elderly or stroke patients, without requiring their presence in a laboratory or clinical setting for extensive testing. Ultimately, it paves the way for restoring mobility and improving accessibility for everyday life at home or in the community.
“This work proposes and demonstrates a new method that uses physical information and data-driven reinforcement learning to control wearable robots to directly benefit humans,” says Hao Su, corresponding author of the paper and an associate professor of mechanics. and aerospace engineering from North Carolina State University.
Exoskeletons have the potential to improve human locomotion performance for a wide range of users, from injury rehabilitation to permanently assisting people with disabilities. However, lengthy human tests and control laws have limited its widespread adoption.
The researchers focused on improving the autonomous control of embedded AI systems – which are systems where an AI program is integrated into physical technology.
This work focused on teaching robotic exoskeletons to assist able-bodied people with various movements and extends previous research on learning-based empowerment of lower limb rehabilitation exoskeletons, also a collaborative effort between Zhou, Su and several others.
“Previous successes in reinforcement learning have tended to focus mainly on simulations and board games, our method provides the basis for a turnkey solution in the development of controllers for wearable robots,” says Shuzhen Luo, assistant professor at Embry-Riddle Aeronautical University and first author of both works. Luo previously worked as a postdoc in Zhou and Su’s labs.
Normally, users must spend hours “training” the exoskeleton so the technology knows how much force is needed — and when to apply that force — to help users walk, run or climb stairs.
The new method allows users to immediately use exoskeletons because the closed-loop simulation includes both the exoskeleton controller and physical models of musculoskeletal dynamics, human-robot interactions, and muscle responses, thereby generating efficient and realistic data and iteratively learning a better control strategy in the simulation. .
The unit is pre-programmed to be ready to use immediately, and it is also possible to update the driver in hardware if the researchers make improvements in the lab through extended simulations. Future prospects for this project include the development of individualized, tailor-made controllers to assist users in various activities of daily living.
“This work basically makes science fiction come true – it allows people to burn less energy doing different tasks,” says Su.
For example, when tested on human subjects, the researchers found that study participants used 24.3% less metabolic energy when walking in the robotic exoskeleton compared to walking without the exoskeleton. Participants used 13.1% less energy when running in the exoskeleton and 15.4% less energy when climbing stairs.
While this study focused on the scientists’ work with healthy people, the new method aims to help people with mobility impairments using assistive devices.
“Our framework can offer a generalizable and scalable strategy for the rapid development and widespread adoption of various assistive robots for both able-bodied and mobility-impaired individuals,” says Su.
“We are in the early stages of testing the performance of the new method on robotic exoskeletons used by older adults and people with neurological conditions such as cerebral palsy. And we are also interested in how this method could be used to improve the performance of robotic prosthetic devices.”
Funding: This research was supported by the National Science Foundation under awards 1944655 and 2026622; National Institute on Disability, Independent Living, and Rehabilitation Research, under award DRRP 90DPGE0019; Swiss Community Life Administration Research Fellowship Program; and the National Institutes of Health under number 1R01EB035404.
About this news from AI research and neurotech
Author: Derek Raymond
Source: New Jersey Institute of Technology
Contact: Deric Raymond – New Jersey Institute of Technology
Picture: Image is credited to the New Jersey Institute of Technology
Original Research: Closed access.
“Experiment-free exoskeleton assistance through simulation learning” by Xianlian Zhou et al. Nature
Abstract
Experiment-free exoskeleton assistance through learning in simulation
Exoskeletons have enormous potential to improve human locomotion performance. However, their development and widespread dissemination are limited by the requirement for lengthy human tests and hand-crafted control laws. Here we show an experiment-free method to learn a versatile control policy in simulation.
Our learning-in-simulation framework uses dynamics-aware locomotor and exoskeleton models and data-driven reinforcement learning to bridge the gap between simulation and reality without human experimentation.
The learned controller is deployed on the hip’s own exoskeleton, which automatically generates assistance in various activities with a reduced metabolic rate of 24.3%, 13.1% and 15.4% when walking, running and climbing stairs.
Our framework can offer a generalizable and scalable strategy for rapid development and widespread adoption of various assistive robots for both able-bodied and people with reduced mobility.