|
Listen to this article  |
The NERVE Center’s robotic arm can perform a variety of tasks. | Source: UMass Lowell, Brooke Coupal
Routine human tasks, such as picking up a water bottle, remain a huge challenge for robots. Robotics researchers and developers have been working for decades to improve robot manipulation skills, but there are broader issues that slow their work and make innovation difficult.
According to Adam Norton, vice provost for academic affairs at the University of Massachusetts Lowell, the problem is “a systemic issue.” New England Robotics Validation and Experimentation (NERVE) Center. For a robot to complete a simple task, many parts must be considered, from the gripper used to grab the object to the software that controls the robot. Researchers who want to develop one aspect of the manipulation process also have to figure out how to build all the other parts of the robot before testing their development.
There is research to alleviate these problems, but the work varies greatly from the techniques and methods researchers use to the challenges they face. – Makes it difficult for others to integrate it into their own research.
“This really hinders researchers who have to do all this work before they can actually implement something that contributes to the field,” Norton said. “It creates a huge barrier to entry.”
This is Norton and Holly Yanco As the Miner Chair of the College of Computer and Information Sciences at Kennedy School of Science and a director of NERVE, he and a team of researchers are working to develop what he calls a standardized “ecosystem” to improve robotic manipulation.
What would a standardized ecosystem look like?
Adam Norton (left) and Holly Yanco, who lead the COMPARE project. | Source: UMass Lowell, Brooke Coupal
Researchers say the ecosystem will be a community-driven repository of guidelines, activities, and working groups that will make it easy to compare and implement open-source or publicly available research.
“By building a new, community-driven, open-source ecosystem for robotics manipulation research and benchmarking, the field will move more quickly toward solutions for robotic perception and grasp,” Yanco said.
The program is called: Collaborative Open Source Manipulation Performance Evaluation for Robotic Enhancement (COMPARE). COMPARE is supported by a National Science Foundation (NSF) grant of approximately $1.5 million.
“Everyone working in this space approaches it in a very different way,” Yanco said. “The goal of the COMPARE ecosystem is to create greater cohesion and compatibility across research efforts.”
Yanco and her team plan to work with the robotics community to develop research standards that will establish cohesion among roboticists. The team believes that following the same standards will allow researchers to compare their own work with research shared in the COMPARE system, giving them the opportunity to implement others’ research into their own efforts.
COMPARE builds on previous research conducted by Yanco and Norton.
Funded by a previous NSF grant worth nearly $300,000, Yanco and Norton conducted workshops with robotic manipulation researchers that provided an opportunity to learn what researchers want from standardized systems.
Since that workshop, Yanco and Norton have formed an advisory board that brings together people from industry, government agencies, and academic research institutions to provide ongoing feedback. The team is also working with robotics experts. Rutgers University, Yale University, University of South Florida, Worcester Polytechnic Institute. All of these institutions have a variety of areas of expertise that can contribute to the system.
“We hope that when we bring people in, they will go into industry or academia and use this standard as part of their research,” Yanko said. “This could have a huge impact on robotics.”
As the guidelines develop, more workshops will be held at the NERVE Center and other robotics facilities to help researchers apply the standards to their practices.
