Robots have been trained by researchers at the University of Cincinnati to work independently, yet able to collaborate with each other.
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’ve made it a bit more difficult for ourselves. We want to complete the task with as little communication as possible between the robots,” said lead author Andrew Barth.
Neither robot controlled 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 adapt individual solutions to learn from past results to optimize performance over time.
“Ultimately, we want to expand this to 10 or more robots working together on a project,” Barth said. “For example, if you want to build a giant habitat in space, you need a lot of robots working together. But if you depended on a communication network and it goes down, then your whole project is done.”
If robots can work independently, losing a robot will matter less. The others can compensate to complete the mission, Barth said.
Robots were tasked with carrying the rod around two obstacles and through a narrow door. They successfully completed the task 95 percent of the time in simulations.
The robotic work partners were also 93 percent successful in a completely new scenario, with two new unknown 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, you’ve made your system more robust against that failure and made it easier for large groups to work together,” explains Barth.
In the long term, it is hoped that the AI will enable multiple robots to work together to perform difficult tasks, such as moving furniture. The system is scalable, so any number of robots can work together on a task.
“There are many applications. Wherever you have jobs that will be done by multiple people in the future, you can have multiple robots do it,” said researcher Ou Ma. “Currently, most robots work alone. But in the future we will need multiple robots working together, just like humans do now.”