Podcast: The future of AI in manufacturing and its real-world applications
Jeff Winter has nearly 20 years of experience in the manufacturing industry, with a focus on automation, safety, controls, and OT and IT systems. Jeff became a thought leader in the fields of Industry 4.0 and digital transformation, and has actively participated in industry associations, academic groups, advisory boards, and industry research teams.
Jeff has teamed up with Scott Achelpohl, managing editor of Smart Industry, to create (R)Evolutionizing Manufacturing, a monthly series of chats about how industrials of all sizes and budgets can embrace technology. The two experts plan to cover a range of topics, including digital twins, predictive maintenance, cybersecurity, IT and OT convergence, automation, and much more. This episode examines how AI can transform existing manufacturing data into valuable insights.
Below is an excerpt from the podcast:
SI: Last month, we talked all about how data is everything. Isn't it especially important for companies to have data that is in order and accessible for any AI-related project or pilot to get off the ground? Are there use cases for AI that actually help companies make use of the data they've always gathered, but never have used to its full advantage? If so, give some examples, please.
JW: Sure. So first, is it super important to have companies have their data in order and accessible for AI projects, especially for pilots? Yes, you bet it is. And in this case, you can think of AI as, I love to use the Master Chef analogy, because AI can create incredible dishes, but only if you have the right ingredients. If your data is messy or incomplete or hard to access, it's kind of like trying to cook a gourmet meal with a fridge full of just random, unlabeled leftover messes. I mean, yeah, you could make food that's edible, but it most likely won't be very good. Clean, well-organized, and accessible data is the foundation for effective artificial intelligence. Also, data is crucial for AI because it's what AI uses to learn from and to make decisions, especially for machine learning and large language models. Having more data allows the model to be more accurate and effective, and more data means better learning, improved predictions, and a deeper understanding of all the little nuances. So essentially the more good data you have, the better your AI can perform. And this isn't true for other logic-based or rule-based systems where you only need a certain amount of data to make the system work.
Now on to the second part, AI use cases for taking advantage of existing data. So, one great use case is predictive maintenance. Many manufacturers have been collecting data from their machines for years. This is what historians do, and they really only look into them when they're on the hunt for something specific. They aren't just automatically taking advantage of it. Now by entering the world of artificial intelligence and by analyzing this historical data, AI can predict when machines are likely to fail and help schedule maintenance before the failure actually occurs.
Another fantastic example on a totally flip side is customer insights. So, retailers and service providers have just mountains of customer data, whether it's purchase histories or even preferences, feedback, you name it. AI can sift through all this data to uncover trends and patterns that were previously hidden. It can also help companies personalize their marketing and improve customer service and even help develop new products that better meet customers’ needs. So, you can imagine AI here is just kind of like a detective piecing together little clues to give you a complete picture of your customers’ desires and behavior.
But then there's also supply chain optimization. Companies have been gathering data on their supply chains also for what seems like forever, but often it just sits there collecting dust. And once again, AI can analyze this data to optimize inventory levels, predict demand more accurately and even streamline logistics. And it's just like having someone behind the scenes making sure that everything is running smoothly and efficiently. So yes, having data in order is crucial, but once you've got that, AI can still do wonders with it. It can turn this dusty old data into gold, revealing the insights we talked about and efficiencies you probably didn't even know were there.
So, here’s some examples where customers can take advantage of data that they already have, but probably haven't yet, that's not departmental special. PDFs. Manufacturers have hundreds, if not thousands, of PDFs. You've got engineering design documents, you even have handwritten notes. You have user manuals, you have maintenance manuals, you have safety guidelines, amongst a whole bunch of other PDFs. Before AI, finding info in these PDFs was kind of like searching for a needle in a haystack. You'd have to know which document to search inside and the exact keywords to find anything, which was therefore super limiting. But with AI, you can do a lot more. So, I'll explain a few examples of it, just with your PDFs. So first you have semantic search. AI isn't just about keyword searches anymore. With semantic search, you can type in what you mean or what you think, and AI can understand the context. And this is kind of like having a secretary, an assistant, understand what you're looking for and then they go and find it, even when you don't know the exact words or potentially the exact document that it's in.
Another big one is summarizing and comparing. So, if you need a quick rundown of a 50-page user manual, AI can summarize it in seconds. Let's say you have five versions of the same machine, and you want to know their differences because they were made in different years. AI compares and contrasts them, highlighting differences and similarities. Another is translation. You may have user manuals in the wrong language that you're used to because of where you bought it or employees who don't speak the native language. No problem. AI can translate them accurately, making sure that everyone on your team, no matter where they're from, can understand what's in them.
Another is creating inspection lists and quizzes. You can use AI to whip up inspection lists from engineering drawings or user manuals to generate an inspection list to go walk around or even employee quizzes, like safety quizzes, just to make sure that everyone knows what they're supposed to be doing on the machinery. And another one is reading drawings and handwritten notes. Computer vision and different cognitive form recognizers can look at complex engineering drawings or even messy handwritten notes and make sense of them, not just necessarily translate the words that are written, but actually tell you what's happening in the particular image. Now, if you go one level up from PDF, AI can then take this document data and then compare it real time to existing data, let's say from your manufacturing execution systems, your CMMS, your WMS, or any other system. It can then also pull in information from external sources, things like textbooks, on, let's say, operations management to help give you insight you never thought was possible. So, we can then compare textbooks to your real-time data, to your PDFs. And we're not even talking about necessarily capturing new data yet, just existing data that's already out there.
About the Podcast
Great Question: A Manufacturing Podcast offers news and information for the people who make, store and move things and those who manage and maintain the facilities where that work gets done. Manufacturers from chemical producers to automakers to machine shops can listen for critical insights into the technologies, economic conditions and best practices that can influence how to best run facilities to reach operational excellence.
Listen to another episode and subscribe on your favorite podcast app
About the Author
Scott Achelpohl
Scott Achelpohl is the managing editor of Smart Industry. He has spent stints in business-to-business journalism covering U.S. trucking and transportation for FleetOwner, a sister website and magazine of SI’s at Endeavor Business Media, and branches of the U.S. military for Navy League of the United States. He's a graduate of the University of Kansas and the William Allen White School of Journalism with many years of media experience inside and outside B2B journalism.