AI in Manufacturing Makes Impressions at INTECH

April 17, 2024

Artificial intelligence was the hottest topic discussed at INTECH, which Trumpf hosted in Germany earlier this week, and at a post-event press conference yesterday.

AI is the key to greater productivity and competitiveness in industrial manufacturing because it helps fabricators and metal formers evaluate machine data more efficiently, detect errors, and automate production, the technology manufacturer says.  

Trumpf“In the last couple of years, we have invested a lot of time and capacity to evaluate AI’s potential and brought AI into our products and our processes, says Stephan Mayer, Trumpf CEO for Machine Tools. AI will play a bigger role in Trumpf’s future strategy so that customers can run the company’s machines at the highest possible uptime and maximum OEE, Mayer claims. “We see that, with our conventional technology, we are reaching a certain limit. We need a closer connection to the machine, and that means we need a direct connection to the machines in the field.

“AI isn’t magic,” Mayer adds. “It’s very good at recognizing repeating structures, repeating patterns—pictures, texts, videos, sound … and based on this recognition, AI can make statistical decisions. It offers a method to automate and to speed up operations.”

Mayer offers five examples where AI can be effectively applied to manufacturing: preventive maintenance, quality control, composing reports, improving automation, and in determining which parts can be run unattended.

Preventive maintenance. AI can recognize a problem by watching parameters, such as if the laser heats up faster than expected, or if dirt is reflecting light, indicating that it needs preventive maintenance. “AI can alert us so we can contact the customer that we detected an error, if mistakes in programming are being made, or that the machine speed can be adjusted for optimal speed. AI is an enabler to this next step.” Mayer states. 

Quality control. AI can help monitor system data and visually inspect products with a camera, such as welding seams to check that they are not defective.

Language composition. A third example is applying language to write reports post-service, similar to Chat GPT. Those reports can be accessed again to provide assistance to another, similar service problem to speed up the right resolution.

Improving operation. “AI always helps when it’s about repeating tasks and following a certain pattern such as unloading, sorting, and loading of parts. A lot of those tasks are still being handled by humans. Sometimes they are replaced by automation. The problem with automating these tasks is that it only runs when all kinds of issues are resolved. Automation is not flexible. Every possible mistake must be included in the algorithm so that the automation knows what to do in any possible situation. That is extremely complex, so only the easy things could be automated. AI can be more issue-flexible, so it increases the robustness of automation because it can adapt if all the instructions are not perfect." 

Selecting lights-out or daytime operation. AI can help determine which parts can be run unattended versus during the day with human supervision. Typically, one of the most challenging areas is separating the laser cut parts from the skeleton completely. Now, AI has accumulated data for each blank shape that shows which have been successfully separated for automated unloading and which have not.

Trumpf is applying AI to accelerate its own product development and distributes software updates over-the-air (OTA) to its customers, similar to Tesla. “What was an acceptable lifecycle of 5 to 10 years now is too slow. AI helps us to be more agile in developing new software products on a three-month cycle,” Mayer says.

Industry-Related Terms: Blank, LASER, Run
View Glossary of Metalforming Terms


See also: TRUMPF Inc.

Technologies: Cutting, Pressroom Automation, Quality Control


Must be logged in to post a comment.
There are no comments posted.

Subscribe to the Newsletter

Start receiving newsletters.