Jan van Eijk (Co-founder Mechatronics Academy)
The 2021 ASPE Lifetime Achievement Award that Jan van Eijk received last year marks the symbolic end of the active career of one of the driving forces behind the Dutch mechatronics industry in recent decades. In part 2 of an extensive interview, Van Eijk looks back and gives some valuable advice for the Dutch high-tech industry. “It is crucial to maintain and feed the network.”
In the Netherlands we like to pat ourselves on the back with how good we are in mechatronics and precision technology. Are we really world class or is it Dutch arrogance?
“I don’t think we need to be too modest about that. In any case, there is little arrogance about it because there are enough points with which you can illustrate that. Not only in terms of knowledge, but also in terms of industrial application, we have built up a major lead over the rest of the world. ASML’s machines are a wonderful demonstration of this, but Thermo Fisher’s transmission electron microscopes are also technical marvels that are hardly unparalleled in the world. And it all comes from the same network, the same community of people who have taken technology to the next level.”
'The willingness, or even the obligation, to share knowledge in the network is crucial.'
How did we achieve that level?
“The basis of this lies with Philips and the Natlab. The optics knowledge that has been built up there forms an important pillar. The machine building and mechanics capabilities also play a major role, as does all the know-how surrounding the construction principles for which Wim van der Hoek laid the groundwork. At its core, it revolves around the Natlab, where people could conduct scientific research at the highest level and link it to implementations in a practical application. The latter in particular has been decisive, plus the fact that that knowledge has been shared with others and that we have built an ecosystem in this way.”
“That willingness, or even the obligation, to share knowledge is crucial. A wonderful example are the famous Thursday morning lectures at the Natlab. There, researchers explained to the rest of the lab what they were doing, and what their plans and ambitions were. And it wasn’t just about what great successes they had achieved, but also what they wanted to do in the future. Employees were obliged to attend, and this created cross-fertilization in a collaborative environment. Colleagues who had virtually nothing to do with a field, nevertheless, asked questions. This curiosity to learn about each other’s profession has led to us not only building up the capacities, but also sharing that knowledge with a larger group, and learning to appreciate and respect each other’s fields.”
“Another example is the BM Landjuweel, where the 150 most prominent people within Philips Bedrijfsmechanisatie met every year to tell each other what they had learned and where things had gone wrong. It was an honor to be invited.”
“When BM got into a frenzy in the mid-eighties, those Landjuwelen stopped. They were followed by annual Philips conferences that alternated between control and mechanical topics. With the same goal: to make contacts and exchange technical information. That role has since been taken over by, among others, DSPE and the training programs of Mechatronics Academy and High Tech Institute.”
Prof. Jan van Eijk.
Is the region sufficiently aware of the importance for knowledge sharing?
“Especially in economically difficult times, it is hard to convince people that you might have to organize such meetings maybe even twice a year, because it is essential to maintain and nurture that network. It is one of the core values that has been an assignment for all executives within Philips since the era of famed director Hendrik Casimir. Philips is no longer such a central player, but the philosophy is quite deeply embedded in our community. The trick is to guarantee continuity in the rat race to earn money. That requires some effort. I think it is sufficiently embedded, but I have often been accused of being too naive.”
In recent years, ASML has taken over the dominance of Philips. How do those companies compare, the Philips of then versus the ASML of today?
“Unlike Philips, ASML is a mono company, focused on one application field. That’s a big difference. Within the old Philips, questions arose from all kinds of domains that were often based on the same basic problems. As a result, a specialized knowledge center could come up with excellent solutions for both electron microscopes and placement machines for discrete components.”
“When working at Philips, I have done pre-development work for ASML. Intimately connected, yet independent. That worked very well. We got a bag of money and could do whatever we wanted. We had a decisive organization that did not have to worry about the day-to-day problems within ASML. If there was a fire there, the siren went off and all ASML staff had to show up. But there was no siren at Philips; we could quietly work on their next problem. Planar motors are a good example of this. At first, ASML didn’t see any point in that, but we just started working on it anyway. When they saw what we had developed a year later, it was immediately snatched from our hands because those motors had to get into all kinds of machines quickly.”
'ASML shouldn’t get arrogant because it’s the best at something. In my opinion, the organization is bureaucratizing quickly.'
“The current ASML is huge. I’ve said before that they have to be careful that they don’t become another Big Blue. That it, like IBM, will neglect some domains. ASML shouldn’t get arrogant because it’s the best at something. In my view, the organization is rapidly closing down and bureaucratizing. If they’re not careful, it will soon clog up with too many people sitting there just to safeguard their own jobs. If ASML falls into that trap, it will be very difficult to keep up the technical drive.”
“As an outsider, you have absolutely no chance if you want to make an improvement proposal. Internally there are so many experts in all kinds of fields that an entire army will jump up to explain why your idea is not possible. For an innovator like me, it is not a nice organization to work for; I don’t get any energy from it. On the other hand, ASML’s successes are undeniable.”
AI no panacea – The Netherlands is at the forefront of mechatronics and precision technology, but we are too small to do everything. Which markets and sectors should the Netherlands focus on?
“There are very interesting possibilities in machines for photonics. A new field that is very related to the production machine problems that we’ve already solved in existing systems. With many of the same companies we can make a positive contribution and claim part of that industry for the Netherlands. However, we shouldn’t try to become the production center. Don’t aim to be Samsung or TSMC in photonics. The focus should be on developing equipment for that sector.”
“Another area in which we can benefit from the technology base that has been built up, is robotics. The football-playing robots from the Eindhoven University of Technology are a nice example of that, but we really have a solid base to score high in robotics. Think of medical and surgical robots, because they also contain that high-tech precision component.”
And automotive, could that be interesting? There are many companies around Helmond, among others, that operate in that market.
“I’m not sure whether the Netherlands is big enough to play a leading role in automotive. Superpowers such as Bosch and the car manufacturers themselves are very dominant. We can certainly make important contributions with smart innovations, but we are not in a position to have a major industrial part. We may have missed the boat there. But even if we had done everything we could twenty years ago, I wonder whether we would have been able to gather enough manpower in the region. We can’t carry another ASML-sized company in the Netherlands.”
And artificial intelligence?
“That is mainly a buzzword for me, just like mechatronics was 35 years ago. You can attract money with it, but above all we have to keep our feet on the ground and not expect that all problems will be solved if we use AI. I don’t believe in that. AI is an interesting tool from which you can certainly derive value, as well as from CAD and finite element methods. I think AI can provide valuable insights that might help you better understand what is happening. This will allow you to improve devices effectively. If you only do data-based optimizations, you will only make limited profit. But if you understand the core of the problem, you can come up with new concepts. In any case, AI is certainly not a panacea.”
From a distance – The award of the 2021 ASPE Lifetime Achievement Award last November marks the end of your active career. What else do you want to do?
“Mechatronics Academy regularly gives fun training courses in Asia or America. If a nice location comes along, I would be happy to contribute. Companies can always knock on my door if they are looking for innovative solutions. Then I would like to join in to think along in the preliminary phase. After half a year I’m worthless because then I’m just going to say that it could have been better. Also when engineers are at a loss because their machine isn’t working properly and they have no idea why, I like to think along. Maybe I can help out. Not because I know better, but because I ask the right questions from a distance that help gain insight.”
“Other than that, not much has really changed. Just like the past ten years, I go on holiday very regularly. With the camper through Europe or to the US. I also play golf and bridge. I can keep doing that for a long long time.”
Jan van Eijk’s most important patents
Jan van Eijk has about fifty patents to his name. What are the two most important?
“I came up with one of my first patents together with Martin van den Brink. If you asked him this question, he would probably mention this patent as well. It is about the alignment of the mask and the wafer in a lithography machine. In ASML’s first machines, this was done with an optical system that compared one point on the mask against two points on the wafer. That means you can never control the rotation nor correct the magnification of the lens. Our invention was to add a second feature. Although that second optical measuring system made it twice as expensive, the big advantage was that you could measure the position and the angle perfectly, without being dependent on the stability of the machine frame. You could also correct the magnification of the lens based on the distance between the two mask features on the wafer. That was a very valuable step for ASML because it made sure that the first generation were guarded as a stable, well-producing machine, with a yield that was a factor of 2 to 3 higher.”
“Another important patent dates back to 1990 and deals with frame motion compensation. If you let a wafer stage make a step, a reaction force is generated on the machine frame. That frame will start vibrating which will lead to image errors. You can solve this with control technology by adding extra sensors and using smart feedforward tricks. But the mechanics themselves are not infinitely rigid. In the second generation of lithography machines, the structure had become so large and heavy that ASML ran the risk of low-frequency vibrations that were detrimental to the quality. You can of course wait for the vibrations to damp out, but that’s killing the productivity. We then came up with the idea of installing three DC motors between the machine frame and the floor. When the stage moves, those motors generate a force that compensates for the movement, causing the frame to stand still. It seems simple, but it has enabled the second generation of ASML machines to reach high positioning accuracy and throughput. That has been of decisive importance because from that generation onwards ASML started to make real money and they were on fire.”
This article is written by Alexander Pil, tech editor of High-Tech Systems.