The automated knitting system uses a neural network to create garments. Gloves created by the automated knitting system. (Image credit: MIT CSAIL)


Knitting is a craft that takes a certain amount of skill, finesse, and patience. But these requirements could be things of the past with MIT’s latest system. Researchers at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) have developed a computer-aided knitting system, which automates the design and manufacture of knitted garments.


The system, called InverseKnit, create patterns using photos of knitted items. It uses a new design software, called CADKnit that allows people with no knitting experience to create their own garments choosing everything from the size, final shape, templates, and decorative details. The system combines 2D images with CAD and photo-editing software to create custom templates. These are then sent to the system. The item is then created out of real wool.


Researchers developed InverseKnit by creating a data set of knitting patterns with matching images which were used to train the deep neural network to generate machine knitting patterns. According to the team, during testing, the system generated accurate instructions 94% of the time. Researchers brought in a group of inexperienced knitters and found they were able to create fairly complex garments, like gloves, and effects, like lacing motifs and color patterns.


While there’s still more to do before InverseKnit can be commercialized, the team envisions it can be a knitting service for customized garments. It’s also ideal for letting designers spend less time learning how to write knitwear patterns for machines and reduce waste in the manufacturing process. 

“As far as machines and knitting go, this type of system could change accessibility for people looking to be the designers of their own items,'' says Alexandre Kaspar, CSAIL Ph.D. student and lead author on a new paper about the system. “We want to let casual users get access to machines without needed programming expertise, so they can reap the benefits of customization by making use of machine learning for design and manufacturing.”


It’s definitely an impressive system, and the fact that it can be used regardless of skill level gives it more appeal, but it removes the human element from the craft. Each knitter is different, bringing their own style and flair to what they’re doing. Automating the process like this will take away that creativity and may make everything more uniform. Not to mention, it may make it more difficult for crafters who sell their items. Why pay someone else to do it if you can just use this system that simplifies the process? It further adds to the conversation of automating jobs and skills.



Have a story tip? Message me at: cabe(at)element14(dot)com