AR training project in development at NC State

Thanks to funding from the National Science Foundation, an augmented reality (AR) workforce project under development by the North Carolina Textile Innovation Engine may one day be the way companies routinely train their employees. 

Jonathan Crumpler, with Western Piedmont Community College in Morganton, N.C., and Tom Wright, of Drop Punk LLC and lead developer with the Textile  Engine, presented their work at the Advanced Textiles Association Emerging Technologies Conference (ETC) in Indianapolis, Ind., Nov. 4.   

“This is going to be an extremely exciting, transformative time period for this type of training,” Crumpler says. 

Crumpler oversees three digital media creation degree programs that provide training for local college students, and Wright has more than two decades of experience in the film and game industry. They provided background on the project, showed video of what the system looks like and gave a live demo. 

AR differs from virtual reality in that in augmented reality, the person wearing the headset sees the digital “universe” overlaid on their actual environment, which is less disorienting for the wearer and less likely to cause motion sickness than virtual reality, where the entire environment is digitally created and immersive. 

The initial program of the Textile Engine project seeks to train a worker on how to tie a weaver’s knot in order fix a broken thread on an industrial loom—a common work stoppage incident on a shop floor. Or, said a different way, Wright says, their first development task was “How do you make tying a knot entertaining?” This helps workers remember how to complete the task. 

By using photogrammetry and a specific large language model—for example, a machine’s entire operating manual—the augmented reality program can represent exactly what the employee is working with. 

In business, the faster that an employee learns a new task, the better for the worker and the company. It can take two years to get someone seasoned and trained, Crumpler says, and this type of system can “expedite that knowledge base” and potentially even make training something that a worker wants to do because the program gives videogame-like feedback.  

Better yet for a company, the shop floor could be replicated in a conference room and prevent machines (and other employees) from being taken out of service for training purposes. “You can break anything in here [a virtual shop floor] because it’s all virtual,” Wright says. 

A system could also one day have integration with the machines—moving from the simpler augmented reality to mixed reality—and warn a worker if they’re about to do something incorrectly or even dangerous. 

It could potentially be used in inspection, as a loom operator could “see” what the programmed design is supposed to look like and watch for errors as it comes off of the machine. 

Challenges of getting the development underway included getting the CAD drawings of the equipment to build their digital counterparts and Wright’s initial learning curve in the programming. He’s a designer, but “barely knows how to script,” Crumpler says, noting how good that is for students at the school to see—that you don’t have to know everything about the task at hand in your job to forge ahead.  

 Also working in this sphere is the Manufacturing Solutions Center at Catawba Valley Community College, which, with Analog Digital Systems, has created the Thread-X training platform using AR and AI, for training and equipment troubleshooting. They expect the funding for years three through five of the project to be announced by the end of the year or in early 2026.