RII Track-4: NSF: Enabling Synergistic Multi-Robot Cooperation for Mobile Manipulation Beyond Individual Robotic Capabilities

Complex robotic missions may require access to and exploration of more challenging sites than typical workspaces, including steep and broken terrain, subsurface voids, and other extreme locations. These missions often surpass the mobility and perception capabilities of individual robots, such as force and payload capacities. However, these missions could potentially be accomplished by a multi-robot team that coordinates actions and applies effective forceful interaction. To specifically address the problem of multi-robot cooperation in challenging sites, the project will develop an efficient coordination mechanism for cooperative mobile manipulation in unknown, dynamic, or demanding environments. This project has potential to transform and expand the investigator's research scope from dexterous multi-fingered manipulation to mobile manipulation based on a team of collaborative mobile robots. Additionally, the project will advance the research infrastructure and capacity of the investigator's home institution, and the transformative approach could be scalable to general problems of interest to NASA and local companies. This Research Infrastructure Improvement Track-4 EPSCoR Research Fellows project will provide a fellowship to an Assistant professor and training for a graduate student at the Wichita State University. This work will be conducted in collaboration with researchers at the Jet Propulsion Laboratory. The project will develop a Synergistic Robotic Coordination framework that enables formal representation, learning, and planning in cooperation strategies to accomplish physically challenging robotic missions. The framework is characterized by tightly coupling and symbiotic coordination strategies of multi-robots, with the aim of generating synergistic effects through creative coordination strategies, including anchoring, tugging, tilting, wedging, sliding, spinning, and pivoting. To establish mutual awareness, a multimodal multi-rate fusion algorithm for heterogeneous sensors will be designed for positioning, and temporal synchronization algorithms will be developed to coordinate multi-robot actions. A reinforcement logical learning architecture that utilizes a "knowledge network" to expedite learning and incorporates first-order logic constraints will be designed to implement SRC strategies. The project has the potential to transform the understanding of multi-robot coordination by introducing a perspective of formal strategy representation and knowledge-driven learning. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.