The Future of AI

Stephan Mayer discusses the opportunities that artificial intelligence creates for manufacturers

July, 2024- Face Time

Q: Why is TRUMPF interested in AI?

A: In manufacturing, AI is the key to greater productivity and competitiveness. It can help companies evaluate machine data more efficiently, detect errors and automate production. If you’re like TRUMPF and you have a product using big data, you need the infrastructure and ability to connect to machines to collect data. At TRUMPF, we’ve invested a lot of time and capacity in evaluating the potential AI could bring to our products and processes. We are investigating how to apply AI to products for the benefit of our customers, as well as how to use it to improve internal processes and become more efficient as a company. AI is not magic, but it’s very good at recognizing repeating patterns, actions and processes, and it is this ability that makes AI interesting in industrial applications.

Q: How does AI help improve industrial production?

A: When we apply AI, it’s as a tool to automate and replace manual steps and speed up production. For example, AI can help with preventive maintenance. When AI measures different laser parameters, such as temperature or light reflections, it can identify an issue before it becomes a problem resulting in a down machine. This allows us to schedule preventive maintenance to keep machines up and running. Another application is quality control. For example, AI can monitor system data and visually inspect products during the welding process and use camera data to detect any weld seams that aren’t good enough. Another example is the application of a large language model to a database containing thousands of service reports. AI speeds up the research necessary to find the right resolution for customers facing similar issues.

Q: How can AI help companies address problems created by labor shortages?

A: AI is helpful with tasks that repeat or follow a certain pattern. One typical repeating task is machine loading and unloading and sorting of different parts. There are a lot of repetitive tasks done by humans today that can be handled by automation. The problem with automation is that it only runs when all issues are resolved. Automation is not flexible  you must define every possible mistake and include it in the algorithm so that the automation knows what to do. Previously, only easy things could be automated. Now, AI makes it possible to be more flexible and develop a solution without 100 percent perfect input, which increases automation’s robustness. So, AI helps automation, and automation helps alleviate labor shortages.

Q: Could you provide an example of how AI helps automated production?

A: AI can determine which parts are better run during the day when people are in the shop versus at night when it’s unattended. When laser cutting metal parts, you want to be able to automatically unload cut pieces without them sticking to the skeleton. AI ensures a robust unloading process. Every time a machine successfully unloads a part, the machine records the shape as successful. And every time it is unsuccessful, that data goes into the database, too. We have collected millions of shapes that have been successful or unsuccessful in automatic unloading and sorting. Then, we trained an AI algorithm to recognize different shapes and determine probability of successful unloading. Now, this information is available to help customers plan when parts should be run. AI creates a probability-based optimizer of day and night planning, so the machine runs continuously.

Q: What role is AI playing in TRUMPF’s corporate strategy?

A: Our ambition is for customers to run TRUMPF machines and products with a highest possible uptime and achieve the maximum overall equipment effectiveness (OEE). We’re reaching a limit with conventional technology. If we want to further increase machine uptime and OEE, we need a direct connection to the machine in the field. We already established this a couple of years ago when we built a platform for direct data exchange with the machine. Now, we want to apply different methods to monitor how the machine is doing in the field. As I mentioned, we are already doing preventive maintenance, but the next step will be to look at how our customers are doing with our machines to see what they need to do to reach maximum OEE. That’s the guiding principle, and AI is moving us further in this direction.

STEPHAN MAYER is CEO, Machine Tools, and member of the Managing Board of TRUMPF SE + Co KG.

TRUMPF Inc., 860/255-6000, trumpf.com