NRI: FND: Human-Guided Robot Teams for Manipulating Large Flexible Sheets in Manufacturing Applications

University Of Southern California, Los Angeles CA

Investigators

Abstract

Large flexible sheets are processed in many manufacturing operations, including composite sheet layup, sheet metal fabrication, and textile manufacturing. The manual manipulation of such sheets is repetitive, stressful and ergonomically challenging and manual operations can also lead to inconsistent product quality. Throughput, quality, and safety of material handling operations can be significantly improved by deploying human-guided robot teams. This enables the human operator to guide the team operation, with robots performing tedious physical tasks. This award supports fundamental advances in decision making to enable a team of robots to work collaboratively with a human operator to manipulate large flexible sheets in small production volume manufacturing applications. The utility of fundamental advances will be demonstrated by applying collaborative robotics to composite sheet layup. Results from this research will enable the introduction of robots in new manufacturing applications and are expected to make the US manufacturing industry more cost competitive in the global market. The integration of the research with graduate and undergraduate courses will enhance the robotics and manufacturing curricula and enrich learning experiences of the participating students. Outreach activities will educate and inform K-12 students about career opportunities in robotics and manufacturing. Performing complex tasks in hybrid work cells that include both humans and robots requires collaboration among agents. Operation of such cells requires coordinated planning, monitoring, and control to ensure that agents can share the workspace and collaborate efficiently by appropriately adapting to the tasks performed by other team members. This award will make computational advances to facilitate collaboration among agents in shared workspaces for manipulating a flexible sheet. The hybrid cell will use a planner to determine the tasks and assign those tasks to agents in the cell. The hybrid cell will use deliberative planning under human guidance to design cell layout, generate a sequence of operations, select grasp locations and generate nominal trajectories for robots. Deliberative planning algorithms will be developed for collaborative manipulation by proactively accounting for interventions by controllers and humans during execution to handle contingencies. Real-time algorithms will be developed for monitoring of the task progress during the execution and initiating interventions to ensure that the part is manufactured in a safe and efficient manner. 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.