Communications of the ACM
Helping Robots Follow A New Path
The team focused on reservoir computing because of its ability to instill memory that gives a system learning ability.
Credit: Erika Gronek / Arizona State University
Ying-Cheng Lai, a professor of electrical engineering at Arizona State University, and a team of doctoral students and two collaborators from the U.S. Army DEVCOM Army Research Laboratory have developed a new method for programming robots' movements.
The team used a type of machine learning called reservoir computing to program a robot to move two arms on a 2D plane in a computer simulation. The robot can change trajectory between predefined paths with only partial knowledge of the surrounding environment.
"The innovative aspect of this approach lies in its capacity to operate effectively with only partial observation of the state of the system," Lai says.
Robots' movements are traditionally programmed using a machine learning method known as linear quadratic tracking, which can be time-consuming and finicky. Lai says the new approach aims to unlock potential applications in autonomous drones and ships, laser cutting tools, and more.
The team describes its work in the journal Nature Communications.
From Arizona State University
View Full Article
No entries found