The idea of using an electronic exosuit or exoskeleton in order to gain certain physical abilities is nothing new. It has been the stuff of science fiction for the past several decades. Now a group of researchers have thought of adding one more ingredient into the mix, in the hopes that it could further improve current state-of-the-art exosuits.

The researchers from Harvard University have decided to add Artificial Intelligence (A.I.) and machine learning to provide users of future exoskeletons with much-needed assistance depending on what task they are trying to perform. Although crude and rather limited in its current implementation, this kind of development is what could eventually lead to the existence of digital robot assistants like Jarvis from the movie Iron Man.

According to Ye Ding, a post-doctorate fellow at the Harvard School of Engineering and Applied Sciences, the kind of A.I. they are developing will be pivotal in the commercial viability of exosuits in the future. Speaking with the futurology news website Inverse.com, Ding mentioned the benefits of incorporating A.I. in such technologies. “Every individual needs assistance that is specifically fit for them, so this type of optimization is the right way to go in terms of wearable devices,” he explained.

“If you’re trying to get the best performance out of exosuits, you have to have something that is figuring out how you should tailor your device at an individual level.” (Related: DARPA awards $2.9 million to Harvard to develop ‘soft’ robotic exoskeleton.)

Based on that statement, one could surmise that it is possible to mimic their methods and offer similar capabilities to exoskeleton users without the need for A.I., simply by logging various points of data in all instances when exoskeletons are used, and simply submitting it all for analysis afterwards. However, the main advantage of including A.I. in the exosuits themselves is the ability to monitor usage and make relevant decisions in real time.

The researchers shared their findings in a paper that was recently published in the journal Science Robotics. In it, Ding and his colleagues explained how they took real-time measurements of certain human physiological signals, such as one’s breathing rate, in order to tweak the relevant parameters for the benefit of the exosuit users.

A machine learning algorithm was created specifically for use in exosuits by the researchers, and it is meant to provide the best level of assistance that needs to be delivered at any given moment, and to adjust the suits as necessary. The study focused specifically on a “hip extension assistive device,” but their methods could potentially be applied to much more complex systems in the future.

Ding states that incorporating A.I. and machine learning resulted in marked improvements. “There were huge improvements in the operation of the suit using this technology compared to other state-of-the-art devices and our previous research,” he said.

“We saw an improvement of metabolic reduction [reduced energy use] more than 60 percent, which are some promising results about using this type of technology in the future.”

With proper use of these A.I. and machine learning technologies, it is easy to see how the experience of using exosuits can be improved drastically. It could be just as the researchers intended, and exosuits could fit perfectly with whomever ends up using them. That way, using them could be easy as well as extremely effective for their designated purposes.

Read more about the development of exosuits and robots in Robotics.news.

Sources include:

Inverse.com

Robotics.ScienceMag.org