Background
Currently, 90% of composite sheet layup is completed manually; this manual process leads to low throughput, high ergonomic risk exposure, and high labor costs. While a Human Robot Collaborative (HRC) approach may seem like an alternate solution, current safety standards in HRC are structured either in a stop-and-go fashion, which can damage the composite materials, or use significantly slower robot movements, decreasing productivity. This project proposes an HRC approach to composite layups that uses the following innovations: a perception-based framework to estimate actions performed by a human, a planning system to select the robot action that better suits the estimated human actions, and an execution monitoring layer to track the state of the cell to prevent errors/collisions.
Objective
Create a safer, more productive process for composite layup through human-robot collaboration. The project outputs are expected to reduce layup time by 20% and improve overall layup productivity by 10% while reducing human ergonomic exposure by 50%.
Technical Approach
The expected project outputs include:
- hybrid cell to support human-robot collaboration
- visual perception analytics framework to estimate actions being performed by humans, drive constraints across tasks, and perform robot pose estimation and environment representation
- planning system to determine suitable robot base positions to perform the tasks, select the robot action that suits the estimated human action, and generate low-level instructions for the robot to grasp and position objects
- an execution monitoring layer to track the state of the cell to prevent errors/collisions
Participants
United Technologies Research Center (UTRC) (Principal Investigator), University of Southern California