A new research project from Canada’s Laurentian University has tried to bridge the gap between textile production and robotic automation by proposing a novel deep learning-based pipeline for reverse knitting to integrate vision-based robotic systems into textile manufacturing.
- This work establishes a foundation for fully automated robotic knitting systems, enabling customisable, flexible production processes that integrate perception, planning, and actuation, thereby advancing textile manufacturing through intelligent robotic automation.
- The research project by Haoliang Sheng, Songpu Cai, Xingyu Zheng and Mengcheng Lau, also co-authors of the paper published in the journal Electronics entailed a model that could convert fabric images into comprehensive instructions that knitting robots could read and follow.
- Fully automating the knitting of clothes, this model could realise patterns for the creation of single-yarn and multi-yarn knitted items of clothing.
THE PAPER addresses the challenge of automating knitting by converting fabric images into machine-readable instructions.
- The deep-learning system developed by the researchers reverse-engineers knitted fabrics from images, enabling greater customisation and scalability in textile manufacturing.
- It tackles the problem of producing knitting instructions by completing two main steps.
- The two-stage architecture enables robots to first identify front labelsbefore inferring complete labels, ensuring accurate, scalable pattern generation.
- By incorporating diverse yarn structures, including single-yarn and multi-yarn patterns, this study demonstrates how our system can adapt to varying material complexities.
- Critical challenges in robotic textile manipulation, such as label imbalance, underrepresented stitch types, and the need for fine-grained control, are addressed by leveraging specialised deep-learning architectures.
- A series of tests, using the model to produce patterns for more than 4,000 textile samples, which were set to be made of both natural and synthetic fabrics, found that it performed well, attaining over 97% accuracy in generating knitting instructions.
THE TEAM: Summer interns who worked on the project included Xingyu Zheng, Jiahui Shu, Tong Zhou, Ying Teng, from Shanghai Sanda University.
- Stoll by Karl Mayer and Shima Seiki provided equipment, technical support, and expertise that were critical to the success of this research.