The High Tech Systems Center has been part of the Eindhoven AI institute EAISI for two years now. The mechatronics center sometimes seems to have been blown away by the AI storm, but the research themes surrounding high-tech systems go much further than just artificial intelligence. There are voices in favor of giving HTSC more of its own identity again.
“Yes, we’re still here.” On the other side of the video call, the familiar faces of the High Tech Systems Center (HTSC) appear on the screen: Marc Hamilton, Jan-Jaap Koning and Ton Peijnenburg. Although the mechatronics center of Eindhoven University of Technology has been part of the much larger EAISI Institute for a few years now, the activities have by no means stopped. “The most familiar faces have left,” acknowledges Peijnenburg, referring to the two former HTSC figureheads, Katja Pahnke and Maarten Steinbuch, switching to Eindhoven Engine.
“It can all seem a bit confusing to the outside world,” admits Peijnenburg, so he’s happy to explain how it works and what the differences are between the two initiatives. “The High Tech Systems Center was created to analyze and allocate multidisciplinary issues from industry within the TUE, which is by nature a monodisciplinary research-driven organization. That catalytic formula works; there’s visibility both internally and externally. Eindhoven Engine has extended that idea. In collaboration with TNO and Fontys, they look more broadly, so not focused on high-tech systems alone, but also on smart cities and medtech, for example. The colocation of academic researchers and engineers from industry is crucial there.”
Hamilton adds: “Eindhoven Engine is also moving to higher technology readiness levels. Naturally, the university revolves around more fundamental issues. The Engine is closer to the practical application.”
HTSC has also always had the ambition to physically bring different researchers together, but until now, the multidisciplinary projects have been housed within one faculty. “It’s our task to ensure that there’s multidisciplinary collaboration because the reflex of a university is to leave it to one department,” says Peijnenburg, giving an example. “In both mechanical engineering and electrical engineering, research is being done on superconducting motors. In one department, they mainly look at the mechanical aspects while in the other, they look at the electromagnetic aspects. It’s not easy to get these researchers together and allow them to benefit sufficiently from each other. The coordination in bringing in joint projects between departments and across the boundaries of faculties can also lead to better and more research applications and collaboration with companies. There lies a task for us.”
More uncertainty may have arisen because the High Tech Systems Center has been part of the Eindhoven Artificial Intelligence Systems Institute (EAISI) since 2019. The combination between high-tech systems and artificial intelligence may not be completely illogical, but there’s also no 100 percent overlap. With data science and AI, you can certainly improve high-end machines, but there are also many aspects that have nothing to do with them.
“That’s a challenge now,” Hamilton admits. “Look, AI is sexy, it sells well. Initiatives in that area are put on the map emphatically. This makes it seem as if we’re somewhat in the shadows. But all topics associated with HTSC remain as relevant as ever, even if they’re not AI related. So we have to continue to position ourselves well.”
Because it sometimes seems to the outside world that HTSC is lost in the AI blizzard of EAISI, there are internal discussions. “Is the current setup a good choice? Are we perhaps closing things off by giving the wrong impression? Or are we making routes that we aimed to open administratively more difficult?,” says Peijnenburg. “Of course, there’s a very pragmatic side to it because we’re now surfing the AI wave. That kind of research attracts money and you’d be crazy as a university if you didn’t profit from that. But we should ensure that other, often more physical aspects of high-tech systems aren’t left out. That is why a reconsideration is underway.”
To create clarity and bring focus, the university has reduced the number of institutes, but for HTSC, this hasn’t worked well in all cases. “You could now conclude that everything in high-tech systems is ultimately an AI problem, but of course, that doesn’t apply to all the challenges the industry is struggling with. At HTSC, we absolutely don’t just do AI. We are merely part of a bigger team where most members have their eyes on AI. We’re the last of the Gauls to bravely hold out, with the pressing questions from industry as our magic potion,” says Peijnenburg with a wink.
That HTSC does more than AI is apparent from a project that was approved at the end of last year. Jan-Jaap Koning explains: “At the beginning of this year, a few PhD and PDEng students started working on the WASPD project about disinfection. Wassenburg Medical, among others, is involved, with a focus on cleaning their medical equipment. So is Sensiks, which makes so-called multisensory experience cabins that must be completely disinfected after use. The next step in that field is plasma disinfection. The research to realize that has nothing to do with AI.”
The WASPD proposal predates the pandemic, but it ran much faster through all funding rounds because of that crisis. “The trend was already visible before, but corona has accelerated the urge for mission-driven research,” Koning has experienced. “In the past, this was still an exception, but now research has to be societally relevant to stand a chance of funding. The Dutch research council NWO is rapidly introducing that requisite.” And it’s not always evident for every research topic. “You can’t justify that easily, for example, if you want to make a lithography stepper more accurate. Mechatronics is often a key enabling technology because without better mechatronics, it’s impossible to get better chips.”
Another non-AI topic for which HTSC has set up a consortium is cryogenic mechatronics. “We’re always looking for the limits of mechatronics. After we got the acoustic vibrations under control, all kinds of thermal effects started to play a role. That’s shifting further towards cryogenic temperatures,” says Koning.
This topic is relevant to the aforementioned superconducting motors. And Thermo Fisher Scientific is also grappling with the problem. Its microscopes need to be cooled to cryogenic temperatures, but the cooling systems it uses at the moment create vibrations. Moreover, it takes a few hours before the system is at the desired temperature, which means that a lot of valuable time of a working day is lost. “There are some companies developing vibration-free cooling systems, but those solutions are still far from the right size; you can’t cool a sample holder in a four-meter-high electron microscope with that,” Koning points out. Because these instruments are crucial in life sciences, financing the research won’t be too problematic.
Systems thinking is still deeply rooted in the DNA of the High Tech Systems Center. Not only is it a mission to soak the TUE curriculum with that approach, but it also stimulates research in that area. And that’s necessary, argues Koning. “In the industry, you see a lot of sub-optimizations. Every link in the chain does its best, of course, and tries to deliver the best possible product, but what you really want is optimization across the whole chain. Think of the CO2 footprint or energy and material use. You’d like to see models covering that, a common language for collaboration and systems-of-systems optimization.”
Hamilton adds: “That’s a major data integration issue, from design to production and operation. What you learn on the operational side has to be implemented in the design, within a chain of equipment that all influence each other. That way, you get many collaborating digital systems, which is a huge challenge.”
Hamilton immediately makes the connection to artificial intelligence: “Data integration is a key word there as well. We can use the knowledge gained in that field in smarter and more automated designs. AI in service of the engineering process.”
Although it’s not yet part of the primary process for many companies, they’ll all soon have to get started with data sharing and digital engineering, says Hamilton. “That data integration is necessary to build advanced applications. To lay a solid foundation for this, we still have a lot of questions to answer.”
To give the new generation of engineers a good foundation in systems engineering, HTSC has developed university-wide systems thinking training for PDEng courses and, together with a number of partners, has secured funding from the National Growth Fund. “There’s no book that describes the Dutch style of systems engineering, which means it’s difficult to teach. The funding is aimed at harmonizing programs across the board,” explains Peijnenburg. “In that regard, it would be good when – just like at TU Delft and the UT – there would also be a professor of systems engineering at the TUE.”
This article was written in close collaboration with Eindhoven University of Technology’s High Tech Systems Center. Main picture credit: HTSC