Independent robots trained to cooperate with each other using AI technique
Image credit: University of Cincinnati
Robots have been trained to work independently, yet capable of cooperating with each, by researchers at the University of Cincinnati.
The team trained a model that would allow any number of robots to work together to move a long rod around two obstacles and through a narrow door in computer simulations.
“We made it a little more difficult on ourselves. We want to accomplish the task with as little communication as possible among the robots,” lead author Andrew Barth said.
Neither robot directed the other and they did not share their strategy in advance to complete the task.
They used an AI technique called genetic fuzzy logic: an intelligent control technique that mimics human reasoning by replacing a simple binary classification (yes, no) with degrees of right or wrong. These genetic algorithms modify individual solutions to learn from past results to optimise performance over time.
“Ultimately, we want to expand this to 10 or more robots working cooperatively on a project,” Barth said. “If you want to build a gigantic habitat in space, say, you’ll need a lot of robots working together. But if you were relying on a communications network and it goes down, then your whole project is done.”
If robots can work independently, losing one makes less of a difference. The others can compensate to complete the mission, Barth said.
Robots were given the task of carrying the rod around two obstacles and through a narrow door. They successfully completed the task 95 per cent of the time in simulations.
The robot work partners were also 93 per cent successful in a completely new scenario, with two new unfamiliar obstacles and a target door in a different location. They had nearly equal success without retraining, even when researchers changed other factors such as the size of the rod.
“If you can train robots to work semi-independently with as little information as possible, then you made your system more robust to that failure and made it easier for large groups to collaborate,” Barth explained.
In the long-term, it is hoped the AI could allow multiple robots to be able to cooperate to perform difficult tasks like moving furniture. The system is scalable, so any number of robots could co-operate on a task.
“There are a host of applications. Any place you have jobs that multiple people are doing in the future, you could have multiple robots doing,” said researcher Ou Ma. “Currently, most robots work alone. But in the future we’ll need multiple robots working together just like people do now.”
Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.