The researchers see a place for robotic assistance for people who can't dress themselves, whether due to injury, age or other ailments.
The robot is a PR2 from Willow Garage and it's using neural network technology to learn its new skill. The PR2 taught itself over the course of a day by analyzing over 11,000 simulations of a robot putting a gown on a human arm.
Not all the simulations had ideal results. Some demonstrated a perfect dressing scenario where the gown went on easily while others involved "dangerous forces to the arm when the cloth would catch on the person's hand or elbow." The robot used these doomsday-dressing scenarios to help it learn what not to do.
"Doing thousands of trials on a human would have been dangerous, let alone impossibly tedious. But in just one day, using simulations, the robot learned what a person may physically feel while getting dressed," says Zackory Erickson, a Georgia Tech Ph.D. student and one of the team's lead researchers.
The PR2 relies on its sense of touch and force, rather than vision, to put the gown on a person. It adjusts its movements as it goes along.
The robot won't be making its debut as a personal valet anytime soon. It takes the PR2 about 10 seconds to slide the gown onto one arm and it hasn't yet advanced to fully dressing a person. But it's off to a good, gentle start.
CNET Magazine: Check out a sample of the stories in CNET's newsstand edition.
'Hello, humans': Google's Duplex could make Assistant the most lifelike AI yet.