To date, industrial robots have struggled to find a place in garment sewing applications due to their difficulties handling limp textiles. This project seeks to advance robotic technologies for applications in the apparel industry.

Project

Current robotic advancements have struggled to address the need for handling flexible materials. This project addresses this gap, working to automate the assembly of garments. Limp textiles will be stiffened, but remain compliant and flexible, so that they can be molded, welded, and sewed. The final project demonstration will showcase the sewing of a uniform.

Objective

Advance robotic technology for automated garment sewing.

Technical Approach

Key to this project is its modular nature, where each individual building block can be replaced by different building blocks that are more suitable for different use cases.

Project outputs are expected to include CDIP focused on the following:

  • Simulation-based robot planning module for thin sheets that need to be assembled and have a wide range of variability
  • Shape identification and pose estimation of 2D/3D sheets
  • Modular ROS-architecture for overall functionality
  • Adjustments to visual servoing library ViSP for handling flexible sheets and multiple cameras
  • ROS packages for recognition, pose estimation, and grasping of thin sheets

Participants

Siemens (Principal Investigator), Bluewater Defense, Sewbo, University of California Berkeley