Design Engineering

Researchers develop algorithm to personalize wearable exosuits


Automation algorithm machine learning wearable devices

Human-in-the-loop optimization uses real-time measurements to improve function of soft, wearable robots.

wearable robot harvard

arvard researchers have developed an efficient machine learning algorithm that can quickly tailor personalized control strategies for soft, wearable exosuits, significantly improving the performance of the device. Image courtesy of Ye Ding/Harvard SEAS.

Whether you bob, saunter or even speed through a crowd, every person moves and walks slightly different. And when it comes to wearable devices like exoskeletons or soft robotics, the user and device need to be in sync.

When it comes to tailoring the robot’s parameters for an individual user, the process can be extremely time-consuming and mostly inefficient.

Researchers from the Harvard John A. Paulson School of Engineering and Applied and Sciences (SEAS) and the Wyss Institute for Biologically Inspired Engineering have tackled this problem.

The team developed a machine learning algorithm that can quickly tailor personalized control strategies for soft, wearable exosuits.


“This new method is an effective and fast way to optimize control parameter settings for assistive wearable devices,” said Ye Ding, a postdoctoral fellow at SEAS and co-first author of the research. Ding adds that this method helps improve metabolic performance for the wearers of a hip extension assistive device. As someone walks, they are constantly adapt their movements to save energy — known as metabolic cost.

“Before, if you had three different users walking with assistive devices, you would need three different assistance strategies,” said Myunghee Kim, a postdoctoral research fellow at SEAS and co-first author of the paper. “Finding the right control parameters for each wearer used to be a difficult, step-by-step process because not only do all humans walk a little differently but the experiments required to manually tune parameters are complicated and time-consuming”

The researchers, led by Conor Walsh, the John L. Loeb Associate Professor of Engineering and Applied Sciences, and Scott Kuindersma, Assistant Professor of Engineering and Computer Science at SEAS, worked on an algorithm that identifies the best control parameters for each individual user.

In order to adjust control parameters for wearable devices, the team followed a method called human-in-the-loop optimization, using real-time measurements of human physiological signals, such as breathing rate. As the algorithm honed in on the best parameters, it directed the exosuit on when and where to deliver its assistive force to improve hip extension.

The combination of the algorithm and suit reduced metabolic cost by 17.4 percent compared to walking without the device.

“Optimization and learning algorithms will have a big impact on future wearable robotic devices designed to assist a range of behaviors,” said Kuindersma. “These results show that optimizing even very simple controllers can provide a significant, individualized benefit to users while walking.”

Walsh adds that for wearable robots like soft exosuits to be effective, the need to deliver the right assistance at the right time is key to establish synergy with the user. With online optimization algorithms, it only takes twenty minutes for the systems to learn how to automatically adjust to best control parameters.

The team plans to apply the optimization to assists multiple joints at the same time.

“In this paper, we demonstrated a high reduction in metabolic cost by just optimizing hip extension,” said Ding. “This goes to show what you can do with a great brain and great hardware.”

The research is described in Science Robotics.


Stories continue below

Print this page

Related Stories