Rice University scientists have been awarded a National Science Foundation grant to develop a novel framework for human-robot collaboration.
Lydia E. Kavraki, the Noah Harding Professor of Computer Science and professor of bioengineering and Moshe Y. Vardi, the Karen Ostrum George Distinguished Service Professor in Computational Engineering and director of the Ken Kennedy Institute for Information Technology, have received a three-year, $749,291 grant. Their research aspires to formally model scenarios where robots and humans can have varying roles.
The proposal is titled ‚ÄúRobotic Collaboration through Scalable Reactive Synthesis.‚ÄĚ It will focus on developing scalable methodologies to help robots adapt to human behavior without changes to underlying software or hardware.¬†
The¬†National Robotics Initiative funded the proposal. The initiative supports fundamental research in the U.S. that will accelerate the development and use of collaborative robots who work beside or cooperatively with people.¬†
‚ÄúWe are interested in creating technologies that serve humans. By formally modeling a robot-human team, we can reason about reliability, safety, and correctness. Earlier work in this domain has largely ignored the preferences of humans,‚ÄĚ Kavraki said.
Kavraki and Vardi will use scenarios drawn from assembly tasks to mimic manufacturing settings where robots and humans work together. In it, the actions of robots can improve the quality and safety of the humans at work. The project is a critical step towards making robots collaborate with humans, while permitting them to remain in control.¬†
‚ÄúRobotics will revolutionize the way work is performed,‚ÄĚ Vardi said. ‚ÄúOur research will establish how robots can adapt to the actions of humans, while still contributing to an end goal as efficiently as possible,‚ÄĚ he said.¬†
The focus of the research is task-level collaboration, as opposed to physical interaction with humans. The framework takes as input a task specification defined in a novel formal language, a language suitable for robotics problems. It produces a policy for robots to assist humans regardless of which subtask or order for the subtask the human agent chooses.¬†
The NSF has awarded five new grants to the Computer Science Department since June 1, 2018. The grants will provide $4 million in research funding for the next four years.
Cintia Listenbee is a Communications and Marketing Specialist in Computer Science¬†