New AI Manufacturing Course Launches for Engineers

Artificial intelligence is quickly becoming part of everyday life. From data processing to recommendation systems, AI is reshaping how people and industries operate — and manufacturing is no exception. 

Across factories and assembly lines, companies are increasingly using AI to optimize how products are designed, fabricated and delivered. “The reality is that industry is moving very rapidly in this space,” said Cate Brinson, the Sharon C. and Harold L. Yoh, III Distinguished Professor of Mechanical Engineering and Materials Science at Duke. “They need to improve quality, reduce costs and shorten time to customers’ hands; and AI can help them in all parts of the pipeline.” 

That urgency helped launch a new engineering course this semester: Artificial Intelligence in Manufacturing Systems, or “AI in Manufacturing” for short, developed and taught by Brinson and Shana McAlexander (MEMS research assistant professor) with support from industry and academic guest speakers. 

Open to all undergraduate engineering students, the course explores how AI is being applied across the manufacturing process. Through guest speakers, demonstrations, peer teaching, industry interviews and real-world projects, students examine how modern AI tools can optimize all stages of production. 

“It’s a broad course that covers manufacturing ‘from soup to nuts’ — from conception of an idea all the way to the end of a product’s life, including materials, supply chain, process control, sustainability, and automation.” Brinson said. 

One hands-on highlight is a casting lab, where students produce the same mechanical part using three methods — 3D printing, CNC milling and sand casting — and compare the results. The exercise helped students to see where AI might be able to improve things for each workflow, from helping generate instructions for the different fabrication methods to analyzing finished parts for defects.  

“I’d never sand‑casted before,” said senior biomedical engineering major Aria Kundu. “It was cool seeing how little defects can build up and how many factors you have to think about when you’re making something.” 

Kundu enrolled in the class after spending two summers interning at Medtronic, where she worked in a medical device manufacturing plant producing oxygenators. The experience convinced her she wanted to pursue manufacturing engineering after graduation. 

“Your ideas can’t come to life without manufacturing,” she said. 

Mechanical engineering junior Avihan Jain arrived with a different connection to the field. Growing up in India, he frequently visited factories run by his family, from tea production lines to pneumatic equipment manufacturing. 

“I grew up around factories,” Jain said. “Now I’m learning how those assembly lines can be optimized and improved with AI.” 

Despite the technical topic, the course is designed to be accessible to students without a deep coding background. Early lectures introduce basic machine learning concepts, and students are encouraged to use AI tools while developing projects. 

“I didn’t have a lot of technical experience with coding coming in,” Jain said. “But the professors explained the fundamentals, and then we could explore more on our own.” 

Guest lectures are another core component of the course. Students hear from faculty and industry experts working across fields like robotics, materials science and industrial automation, offering perspectives rarely covered in traditional engineering classes. 

One lecture that stuck with Kundu explored AI’s role in cybersecurity. Miroslav Pajic, a Duke electrical and computer engineering professor, demonstrated how an AI system could reconstruct a physical key from photos and then 3D print a copy that opened the same lock. 

“I’m not going to lie: it was a bit unnerving,” Kundu said. “But it was also a good reminder of how powerful the technology is. Seeing what AI can do makes you realize why it’s important to understand it and think carefully about how we use it.” 

The class also tackles the ethical implications of automation and AI in the workplace. Through case studies led by Siobhan Oca, a MEMS assistant professor of the practice, the students discuss how companies balance efficiency gains with worker safety and job displacement. 

“One of the big lessons we learned from industry was: Trust AI, but also verify it,” Kundu said. “AI can simplify your tasks, but the time you save should be spent verifying the output. Humans are still responsible for the final product.” 

Connecting students directly with industry collaborators is another key element of the course. Each student conducts an interview with a manufacturing professional and presents their findings to the class. The students can also receive mentorship from industry professionals for their final projects, where they develop and propose an AI-driven solution for a real-world manufacturing problem such as defect analysis, machine maintenance scheduling, or consumer demand forecasting. 

“Industry speakers bring invaluable insights to the course,” Brinson said. “It’s just so interesting for students to hear how AI is actually being used in the real world.” 

For Jain, those interactions opened up entirely new ways of thinking about the field. 

“This has been one of my favorite classes at Duke so far,” he said. “It’s exposed me to so many real-world challenges that people in industry are actively trying to solve and that I’d be interested in working on.” 

Kundu agrees. As she prepares to start her career as a manufacturing engineer with the health care company Abbott after graduation, she encourages other students to consider taking the course. 

“I think every engineering student could benefit from this class,” she said. “Whether you’re civil, mechanical, biomedical, electrical…no matter what field you’re specializing in, manufacturing plays a part in it.” 

For Brinson, this pilot semester is only the beginning. She hopes future versions will expand enrollment and continue building partnerships with industry. 

“If there are industry professionals out there who want to talk with our students, we’d love to have them involved,” she said. “Those conversations are incredibly valuable for the class and what our students take away from it.”