Q&A: How is American manufacturing evolving with AI and robotics?

UNIVERSITY PARK, Pa. — Since the beginning of the year, the U.S. manufacturing industry has received trillions of dollars of investment from private companies, the federal government and foreign donors, with the manufacturing sector alone accounting for over 50% of private-sector research and development in the U.S.

According to Ilya Kovalenko, assistant professor of mechanical engineering and industrial and manufacturing engineering at Penn State, manufacturing research is harnessing the rapid developments in artificial intelligence (AI) technology to build advanced robots and simulation technology capable of improving factories’ safety and efficiency. These developments, at the rate they are progressing, would be impossible without the unprecedented support the field is seeing, he said.

In the following Q&A, Kovalenko discussed the state of U.S. manufacturing and where he sees the most room for future research breakthroughs and workforce growth.

Q: What is driving this unprecedented federal and private investment in manufacturing, and how has it changed the field?

Kovalenko: In the late 20th century, we lost a lot of manufacturing overseas, meaning the U.S. is now reliant on other countries to produce goods across a range of sectors. Over the last five or 10 years, however, between the COVID-19 pandemic and associated supply chain shortages, we’ve seen firsthand the importance of domestic manufacturing capability. That's why there's a huge, huge push to “reshore” or bring manufacturing back to the U.S.

The backbone of manufacturing in the U.S. is built on smaller, mom-and-pop shops — businesses that are generally struggling to fulfil all the different orders they receive. Right now, U.S. manufacturers have fallen behind compared to much of the rest of the world when it comes to producing goods, meaning we need to conduct a lot of industrial and workforce research to enable that catch-up. One of the ways that we can help those smaller shops is through industrial automation and AI, building robots and tools that can automate and optimize facilities by taking over repetitive and mundane tasks that they struggle to staff.

Q: How is automation and AI changing how manufacturing is done?

Kovalenko: In manufacturing facilities today, there are a lot of dull, dirty and dangerous jobs out there. We don't want to expose people to toxic chemicals or dangerous processes, so automation is being explored to enable robots and automated equipment to make human work safer, faster and more cost-efficient. In my research, I'm exploring how AI can enable product flexibility and adaptability in manufacturing. It is critical for your production systems to be flexible, reconfigurable and adaptable to customer demands or facility needs. AI and smart robotics allow us to dynamically change system behavior and produce the goods that we need faster and more efficiently.

Additionally, manufacturing facilities are in desperate need of more trained workers. Automation and AI can help solve this problem by taking some of the strain off a system that needs more personnel generally, while also enabling safer and more collaborative systems that work with people on the shop floor. For example, manufacturers are considering custom automation systems and humanoid robotics to help address problems. It's an exciting field with so many possibilities being enabled by new technologies, from building large structures with 3D printing, working with the next generation of small-scale fabrication techniques to power tomorrow’s semiconductors, or helping foundries modernize how they produce the castings for ships, power plants and vehicles.

Q: How are universities like Penn State supporting workforce development initiatives related to these new technologies in manufacturing?

Kovalenko: Right now, the U.S. manufacturing workforce needs help. At Penn State, we train students across a range of ages and skill levels, from undergraduates to doctoral candidates, with hands-on labs and industry examples to prepare them to work in manufacturing. This hands-on approach is very important because it enables engineers to go into the factories of the future and effectively interact with all the equipment.

From a research perspective, universities play a critical role. Academic researchers push forward ideas across manufacturing research, AI and industrial automation, figuring out the challenging questions that companies might not have the bandwidth to focus on. At Penn State, we have a lot of folks who are experts in materials, mechanical engineering, industrial engineering and more. Bringing all these perspectives together allows us to discover new ways to make manufacturing safer, more efficient and better for humans in general.

Q: Looking 10 or 15 years into the future, what innovations are you most excited to see reach factory floors?

Kovalenko: Several pathways in particular really excite me. First of all, I think building manufacturing systems to think dynamically and adapt to changing conditions on the fly holds great promise. These conditions could be as complex as a customer wanting a different part based on adjusted specifications, or as simple as the weather outside changing, but powering systems to think dynamically can, for example, eliminate waste and use less grid energy, enabling significantly more sustainable manufacturing.

The other thing that really excites me is where we start thinking about moving these dynamic systems beyond just the four walls of the manufacturing facility. How do we think about manufacturing things on the fly as they are already in transit? How do we think about what's happening outside these four walls and how it affects our system and, again, adapting our capabilities based on that? So, a lot of ideas regarding adaptability, resilience and flexibility are what really excite me with the new progress we are seeing in automation and AI today.