Designer of the future: Trump on the way to becoming a machine human

From the Handelsblatt archive: For the family company, it is no longer enough to just build the best lasers. In the fight against competitors from China, complete robotic systems count. Thanks to AI, Trumpf is also conquering this field.

Ditzingen. In the Trumpf development hall in Ditzingen, Swabia, the revolution in robotics does not begin with a spectacular-looking robot. It begins the moment machines stop having their every move dictated to them because they start making decisions for themselves.

Robotics is actually not a core topic for the family business, which specializes primarily in the use of lasers - from sheet metal processing to chip production. However, it is in the company's DNA to always have its finger on the pulse of the latest technologies and to be in close contact with research and science. The secret of Trumpf's success is making new technologies suitable for industrial use.

And so Ditzingen is ultimately about making robots smarter. They are intended to complement sheet metal processing machines into an intelligent system and thus create competitive advantages. Because the Chinese competition is now causing problems even for the long-standing world market leader. The Handelsblatt received exclusive insights into the sensitive area of ​​Trumpf for the Future Designers series.

If you think of intelligent industrial robots and think of images of the machine man from the legendary silent film classic “Metropolis” by Fritz Lang from 1927 or of the gold-colored Star Wars robot C-3PO, you will probably look disappointed when you enter the Trumpf development hall.

“In the factory, a robot doesn’t have to look like a human,” says Ralph Lange, head of robotics at the medium-sized company. "If he only needs one arm, he only has one, if he needs three, he has three. If he doesn't need a head, then he doesn't have one." The intelligence could lie somewhere inside the robot or in the cloud. Ultimately, Trump is about making robots smarter.

Today, even with extremely technical devices such as industrial robots, technological leaps mainly occur in the software. Trumpf has around 600 developers for its own software. There is also an AI team with twelve people at the headquarters and another 90 spread around the world. "AI is not a spectator sport," says Sarah Engel, "Head of AI" at Trumpf. "We take all employees with us. There is compulsory training, but we rely on curiosity and play through applications on site."

Ralph Lange: AI software is fundamentally changing robotics. Photo: Ralph Lange, head of robotics at Trumpf

The Trump Method: The Path to the Light

Curiosity and personal development have always been two important driving forces in the Trumpf cosmos. Berthold Leibinger laid the foundation for his first invention in his diploma thesis in 1957. The copying and nibbling machine made it possible for the first time to copy contours and sections of sheet metal parts quickly and precisely using a template. This earned the patriarch, who died in 2018, the nickname “Nibbler King” and laid the foundation for his rise to become a global corporation. He had his patents paid for in company shares until he finally owned the company founded by his godfather Christian Trumpf in 1923.

At the end of the 1970s, Leibinger recognized the potential of lasers for sheet metal processing. Leibinger bought American lasers and was not satisfied with their results, so he refined and stabilized the technology so that it was suitable for mass production.

Berthold Leibinger: In 1956 he began an experimental diploma thesis at Trumpf to improve the cutting process for shearing. Photo: Trump

Today, Trumpf is the world market leader for industrial lasers under the leadership of his daughter Nicola Leibinger-Kammüller. The applications go far beyond cutting sheet metal. They range from processing the glass edges of smartphones to ultra-short pulse lasers to the EUV lasers and plasma generators, without which the latest and most powerful chips could not be manufactured.

No other company can do that. Trumpf has a turnover of one billion euros with semiconductor technology alone. As a result, the company has built a completely new foothold within 20 years. The latest future project with partner Rohde & Schwarz is lasers for drone defense.

Trumpf's show laser at the headquarters in Ditzingen: On the occasion of the company's 100th anniversary in 2023, it shined into the sky. Photo: Trump

However, even at Trumpf, not everything the company sets out to do always works out. Some forays into new technologies have ended. The 3D metal printers were sold, as were the photonic sensors acquired by Philips just a few years ago.

But for the Trump leadership, such failures are par for the course. “We will always adapt to the new technologies and circumstances of the world – even if we don’t peddle them,” says Mathias Kammüller, describing the basic principle of the Swabian company. In management, the husband of the company boss is responsible for digitalization. “We have always incorporated new technologies into our products to move forward.”

Robots help machine tools

The use of artificial intelligence and its application in robotics fit into the picture. At Trumpf, customers’ spare parts orders are already largely handled via AI. But artificial intelligence still has its limits. “When it comes to the causes of damage, direct communication between the customer and the service specialist is safer, as the AI ​​sometimes doesn’t say when it doesn’t know something and starts to fantasize,” says machine tool boss Stephan Mayer, adding “yet”.

» Also read: China's humanoid robots do somersaults, but can they also work?

AI agents in the company are already complementing and controlling each other. This applies to processes, but also to software development, which becomes significantly faster thanks to AI.

“There are applications that were not conceivable twelve months ago,” says Mayer. For example, today algorithms would independently analyze the frequencies of motor currents of an axis with 52 parameters and compare them with the values ​​of the final acceptance tests in the factory in order to remotely diagnose the condition of an axis. There are now more than 50,000 Trumpf machine tools in the world, a fifth of which are connected online in real time. For new machines, the online share is 70 percent.

Future designer

Now AI should also make laser-controlled sheet metal cutting systems with robots more intelligent. For a long time, Trumpf was the world leader in the core business of machine tools for sheet metal processing. But the position is threatened by emerging Chinese competition. “Lasers are increasingly becoming an interchangeable commodity,” said Supervisory Board Chairman Peter Leibinger five years ago. "We have to offer more in the future. The customer wants superior systems." As head of the Federation of German Industries (BDI), Leibinger is also one of the most influential voices in German business.

And so there seems to be no way around intelligent robots. The biggest risks of Trump's robotics offensive are not doing it or doing it too slowly. The robotics market is growing quickly and the competition from Asia is fierce and is developing new products ever faster.

In order to show customers new possibilities, Trumpf is building complete digital factories, for example in the USA. And today this includes increasingly intelligent robots, even if this has not previously been Trumpf's domain. The fact that this is changing can be seen in the development hall.

Here, two mechanics are working on a new kind of industrial robot. One and a half years ago, Lange came to Trumpf from Bosch to, as he says, “use software and AI to make industrial robots more flexible and powerful.”

What is less known to the public is that Trumpf has been working on this type of robotics for years. There is a reason for this: When processing sheet metal, Trumpf needed robots that have the most power in the last joint and at the gripping point, for example to bend sheet metal. Since such powerful grippers didn't exist, Trumpf built them himself. Now it's a completely new dimension.

Software and AI as the new trump card

Between laser cutting systems, bending cells and robot arms, Lange talks about a change that he has never experienced in more than two decades of robotics development. What changes robotics are neither more powerful drives nor more sophisticated grippers. It's the software. The bottleneck was called programming.

Lange stops in front of an older bending cell. It's now around 25 years old. It works reliably, precisely and to this day does exactly what it was built for. But just that. “Back then, practically every movement had to be programmed,” he says. If something went wrong, the optimization started again. That could take days. Manufacturers of small series were often unable to use robotics economically.

Just a few steps further, Lange stops in front of the SortMaster Vision. On the outside, the difference to the older generation seems manageable. But this is exactly where a new stage of development begins. Every movement is no longer dictated. Three small cameras hang barely visible above the work area.

There are sheet metal parts in front of the robot - some clean, some at an angle, one half on top of the other. In the past, this could have been a problem. Today, AI uses the images to analyze the scene within a few moments. It recognizes the position of each individual part, decides which one needs to be picked up first, calculates the optimal movement path and then sorts the workpieces and places them in the correct stack.

The three cameras above the SortMaster Vision are not expensive special cameras. Only the AI ​​uses this to calculate a spatial image of the situation - even with reflective or reflective metal sheets, which have long posed considerable difficulties for classic image processing. Trumpf does not develop all of the software itself. When it comes to complex platform technology, the Swabians use the Google subsidiary Intrinsic, which specializes in robotics and AI, as a partner.

Guest commentary

AI-based robotics is becoming a global power factor

But Trumpf doesn't just use artificial intelligence for sorting: in the new welding systems, the AI in the software recognizes where a programmed weld seam needs to be corrected and makes the correction independently. People no longer develop every single solution. The person describes the goal. The machine finds its way there itself.

The dark secret of robotics

For the developers, the work begins long before the first real use. Tens of thousands of artificially created sheet metal parts serve as training material. New algorithms are first created in simulations before being tested on real machines. Experiences from the field then flow back into development.

Despite all the progress, Lange clears up a common misconception: "There is a dark secret in robotics. Modern robots are not yet constantly learning on their own." Instead, the machines collect data from practical use. Developers analyze these, improve the models and then install new software versions.

Robotics is increasingly developing like software. Not every few years with a completely new machine, but continuously - through data, models and updates. There are several reasons why this development is gaining momentum right now. AI processes have become more powerful, large-scale computing power is available, and modern development platforms significantly shorten the path from the idea to the finished application.

For Lange, everything comes together at the right time. “Five or 10 years ago we simply wouldn’t have been able to do many of these things.” That's precisely why he's optimistic about the future. For him, the focus is not on humanoids, but on intelligent industrial machines. This could have far-reaching consequences for sheet metal processing. If robots can sense their environment themselves, plan movements independently and deal with changing components without complex reprogramming, automation will for the first time become economical where small batch sizes and frequently changing orders were previously considered an obstacle.

This is important for Trump. Because major customers in the automotive industry who are in crisis are ordering less and less. The smaller customers are becoming more important for the company, which grew only minimally to 4.34 billion euros in sales in the 2025/26 financial year. The increase in incoming orders from 4.2 to 4.5 billion euros has recently caused more optimism.

Related topics

TrumpfSoftwareArtificial Intelligence

In 2025, Trump responded to a slump in sales with a rigid austerity policy, reduced costs by 350 million euros and cut 100 domestic jobs for the first time in two decades. But the company doesn't want to save on the future. Development costs remain high at twelve percent of sales. The robotics solutions are expected to come onto the market in the next few months, by 2027 at the latest.

More: Drones, robots, cars – AI in end devices offers new opportunities for Europe’s chip manufacturers