Jiaoyang Li, assistant professor at the Robotics Institute in Carnegie Mellon University’s School of Computer Science, has earned the National Science Foundation (NSF) Faculty Early Career Development (CAREER) award. The CAREER Program grants some of NSF’s most prestigious awards to early-career faculty who show exemplary dedication to the mission of their institution and act as academic role models.
Li’s project, titled “Environment Optimization for Large-Scale Multi-Agent Coordination: Models, Methods, and Applications” focuses on rethinking how environments are designed for robots as they become a more prominent part of the workforce. Two main objectives guide her work: establishing a more robust understanding of the significance of environment optimization for multi-agent coordination and developing new environment designs that are tailored for large-scale multi-agent coordination.
“In a time where we may have hundreds to thousands of robots working in places like warehouses, we need to ask critical questions about the environment design,” said Li. “For example, how do we decide where we put the shelves, where we put the workstations or where we put the aisles? And how should the robot interact with each of these design elements?”
Li not only focuses on physical environment planning, but also dedicates ample time to virtual environment optimization. When controlling robots, virtual models help them to understand and navigate spaces. By introducing new virtual rules, such as traffic rules that guide the robots in certain directions, it makes it easier for robots to coordinate with one another.
Currently, the most successful large-scale multi-robot coordination is found in warehouse automation, where leading companies operate thousands of robots within a single facility. However, the underlying coordination algorithms are often centralized, meaning a central computer instructs them, and the robots simply follow its commands. This centralized approach limits the applicability of such algorithms in broader, more dynamic scenarios, such as coordinating large fleets of autonomous cars or drones in future smart cities.
“Our motivation is to move toward a decentralized system where robots can make their own decisions intelligently without sacrificing global efficiency,” said Li. “We believe that virtual environment optimization offers a promising path forward. It allows us to optimize the environment at a global level, while enabling robots to plan independently by following a shared set of rules.”
Li and her team have already produced some exciting results in their multi-agent coordination work. In a paper titled “Multi-Robot Coordination and Layout Design for Automated Warehousing,” she and her Ph.D. student Yulun Zhang present an approach for automated warehouses in which they extend existing automatic scenario generation methods to optimize warehouse environments. They found that their resulting optimized layouts performed better than traditional human-designed layouts in terms of throughput and CPU time.
Li and her lab, the Artificial Intelligence for Robot Coordination at Scale lab (ARCS) have also documented more notable results in physical warehouse layout design, virtual environment optimization, and fostering multi-agent collaboration.
The NSF CAREER award places equal importance on research and community engagement. As part of her education outreach efforts, Li is teaching a course at the Robotics Institute called “Multi-Robot Planning and Coordination,” which will focus on how to coordinate and control multiple robots and will heavily feature elements of her supported research.
“The NSF grant compliments this course by allowing me to teach basic robot coordination while also asking critical questions about environment optimization and making that robot coordination more adaptable,” said Li.
With the support of the five year award, Li and her team will continue to refine their methods to make significant advances in the field of multi-agent coordination and environment optimization.
For More Information: Aaron Aupperlee | 412-268-9068 | aaupperlee@cmu.edu