Harnessing Tech for Humanity

January 7, 2021
person holding a metal exoskeleton
Associate Professor of Electrical and Computer Engineering Domen Novak holding his team’s exoskeleton prototype.

From exoskeletons to artificial intelligence and virtual reality, UW researchers are bringing the future home to improve lives.

 

By Micaela Myers

 

We’ve all seen futuristic inventions in science fiction movies—making lives easier in ways we couldn’t imagine. Now, that future is not so far away. Researchers at the University of Wyoming are working on a wide range of formerly futuristic ideas. For example, Associate Professor of Electrical and Computer Engineering Domen Novak, along with students and colleagues, are fine-tuning an exoskeleton that can help those with back problems by providing support. They’re also examining ways artificial intelligence, machine learning and affective computing could help with everything from driver alertness to video game development. In addition, Novak is using virtual reality to improve at-home physical therapy.

 

Exoskeletons

Right now, millions of Americans are experiencing back pain. Novak together with kinesiology Associate Professor Boyi Dai and their respective teams are studying the use of exoskeletons—mechanical spines a person wears. Paired with sensors and artificial intelligence, the exoskeleton can use machine learning to detect when a person is doing certain activities such as bending down to pick something up and apply the appropriate support. 

“We started on this project a couple years back,” Novak says. “We showed it could affect the human body. It changes posture, and it changes muscle load. It does that differently depending on what you’re doing.”

Now, with National Science Foundation (NSF) funding, they’re refining optimal configurations for each situation. In Dai’s lab, subjects are tracked and measured with electrodes and cameras.

They’re also comparing the exoskeleton to an exosuit—a softer version made of fabric and bands.

In the future, exoskeletons could be used for rehabilitation, for people with long-term injuries or for those needing occupational support such as luggage handlers.

 

Affective Computing

Affective computing uses computers to recognize how humans feel subjectively—something Novak has been working on for the past 12 years.

“We put sensors on people, and we try to detect if they are stressed out, bored, distracted or feeling great,” Novak explains.

The sensors are similar to those on a lie detector test and measure things including sweat, pupil dilation, heart rate and brain waves. Results are run through machine learning to try to detect patterns based on previous examples.

Novak first used affective computing in rehabilitation and robotics. For example, a robot used for physical therapy could note if the patient was stressed out so that exercises could be made simpler or, if a patient was bored, more challenging.

Then, they moved on to driving studies with civil engineering Associate Professor Mohamed Ahmed, who serves as director of the driving simulator. Using the simulator, they analyzed if people are distracted or tired in order to create warning systems.

Novak also studies affective computing in video games.

Funded by the NSF, the research focused on improving machine learning overall so that it can better detect the mental state of the user. A new project began Oct. 1 in partnership with Associate Professor of Psychology Joshua Clapp. The new project focuses on how subjects feel when the machine misjudges their state of mind.

Novak says that there are many possible applications for affective computing from the things they’ve looked at, as well as to other areas such as education.

 

Virtual Reality

Physical therapy can be boring—repeating the same exercises over and over. Novak and students are using virtual reality to make it more fun.

A screen makes the workout into a game that can be played between the patient and a physical therapy robot, between two patients or between a patient and a loved one.

“People have started putting them in commercial rehabilitation systems,” Novak says of the technology.

They’ve also looked at applications for cognitive training, such as memory games on tablets. These applications are easy to deploy and for people to use at home.

The possibilities for artificial intelligence, machine learning, affective computing and virtual reality are endless, with many UW researchers leading the way.

 

 

Contact Us

Institutional Communications
Bureau of Mines Building, Room 137
Laramie, WY 82071
Phone: (307) 766-2929
Email: cbaldwin@uwyo.edu


Find us on Facebook (Link opens a new window) Find us on Twitter (Link opens a new window)