NSF Award to Launch Study of How Older Adults Interact With Robots

With the number of older adults in the U.S. population rising and straining the systems in place to take care of them, Matthew Gombolay sees a solution — robots.

Gombolay received a National Science Foundation (NSF) CAREER Award for research that could make assistive robots the standard of care for older adults. The award is the most prestigious the NSF offers to early-career faculty. 

“When people age, they deserve to age with dignity and not just be locked away,” said Gombolay, an assistant professor in Georgia Tech’s School of Interactive Computing. “If you don’t have enough resources or access to home nurses or adult children who have extra time to take care of you, what’s going to happen?”

Gombolay will receive nearly $600,000 to collect the largest data set of its kind on how older adults interact and communicate with assistive robots. Gombolay will then use that data to create algorithms that can be deployed in assistive robots and understand the needs of older adults.

“I think robots offer a middle ground,” Gombolay said. “If we can get robots to help people enough with activities of daily living, cleanliness in the home, meal preparation, monitoring the well-being of individuals, it could be enough for someone to be able to live without someone else coming into their home and taking away their right to privacy.”

Currently, a robot could do little to care for older adults, Gombolay said. The robot would be expensive, an engineer would need to customize the program for each user, and it would require periodic maintenance. 

“We must remember that the people who will teach these robots to perform these tasks won’t be roboticists,” Gombolay said. “It will be the patient or a family member, so we’re developing the algorithms and data sets we need to enable robots to learn through interaction and watch user demonstrations of assistive tasks they want the robot to perform.”

Because the needs of older adults vary from person to person, Gombolay said it could take two to three years to write these algorithms before they’re ready for deployment.

“People are going to have different goals and values and preferences, and that’s going to make it hard for these robots to learn. Algorithms don’t like that. We also must take into account safety features and make it interactive so we can understand what the robot thinks the user is teaching it.”

Gombolay said he also wants to use the grant to help grow the next generation of roboticists. He’ll be working with educators from local high schools to integrate robotics into the classrooms of underserved communities. 

Gombolay credited his Georgia Tech colleagues— Associate Professor Sonia Chernova and Professor Mark Riedl — as his mentors for the NSF CAREER Award application process.

“It’s very affirming that I am working on an important problem,” he said. “It’s good to know we’re on the right track to make a positive impact on a problem in our country.”

For more on Gombolay’s research, visit the CORE Robotics lab website.

Top photo by Terence Rushin/College of Computing.