Professor Maarten Steinbuch is famous for getting things done, building informal networks of those willing to tackle the grand challenges in our society.
After a career in precision mechanics and controls at Philips Research and working on projects for the chip-design manufacturer ASML, he joined Eindhoven University of Technology. He now heads the Control Systems Technology group, within the Department of Mechanical Engineering. But he is keen to show that success is linked to what his teams can do with others both on and off the campus.
It's the start of the new academic year 2015. So a good opportunity to investigate what research centres like this are doing with startups, especially amongst their students. Maarten immediately suggests we leave his office and take the tour to see what’s really happening.
On the way down the stairs, I ask why he went from industry to academia?
“Because I thrive on the freedom to be able to think outside the box”, explains Maarten. “When you're busy developing the business strategies of tomorrow, you have to put aside conventional thinking about the next quarter’s sales targets. Merely extending knowledge a step further is not developing science.”
“You see so many industrial research labs today that are only developing incremental improvements. But breeding cleverer homing pigeons did not help in the development of the telephone, nor did breeding faster horses lead to the invention of the steam locomotive.”
This region pioneered open collaboration
“I think that European universities can do a lot more to collaborate with industry and come up with disruptive thinking. I don't see the difference between fundamental and applied research, for me it is all the same. We need to take more risks, helping industry with the longer-term strategies.”
“But it is also true that academic research has to remain closely in tune with real-world needs. Which is why we have never lost touch with all kinds of professions or short-term challenges as well. Indeed, as we see from today's startup world, few answers are found by just sitting in the lab. Talking with surgeons, oil-riggers, car manufacturers is essential. That's why I think we need more startups in this region, especially because we have the "maker spirit" in our DNA.”
“We’re going to walk through three different labs, automotive, autonomous robots and medical robotics devices. There’s a connection between all three and it is because we have a leading global reputation for understanding very precise positioning technologies.”
We arrive in a large space which reminds me of the repair workshop in a local garage. Maarten points to several cars parked in the corner.
Rethinking self-driving cars – connected rather than autonomous
“You may have seen Twizy, the small car from Renault. Here at TU/e we took out the steering wheel, replacing it with a special motor for steering, and added camera's. The goal is to make this one autonomous, but perhaps surprisingly, not because we believe in the autonomous car as such.”
During our conversation, it emerges that Maarten's teams have a different view than Google on the feasibility of the self-driving car.
“We want to show that the future use of the road is the platoon, better known as cooperative driving. We are convinced that autonomous driving, often termed the self-driving car – is not the only and last answer. Such cars require too much road space because you need a comfortable safety margin.
“The final answer, we believe, will be connected cars that can follow very closely behind each other in a platoon. We’ve shown that connected cars, driving 50 centimeters behind each other, make enormous savings in carbon dioxide emissions because of the reduction in air drag. At the same time, they make maximum use of the available road space. We don’t need wider roads to solve congestion!”
Maarten's department is working closely with industrial leaders like NXP in developing connected control systems that are also safe from disruptive hackers.
Opposite the Twizy, three small cars are parked next to each other. Maarten explains that although they look similar, but they are not identical.
“These are very lightweight Volkswagen Lupo 3Ls – this one has been made fully autonomous. Next to it is a benchmark diesel version and the third one is fully electric. So if you look on the back the three has been reversed to look like an "e": Lupo eL. We’re telling everyone that we want to play an influential role in the development of electric driving.”
“Tesla started designing electrical vehicles by starting at the top end. My colleague researchers have taken a very different route, by taking a lightweight car like the Lupo 3L and seeing if they could make it far more efficient than a Nissan’s Leaf. We're using the same battery package inside but our car is much lighter. The Nissan Leaf weighs 1500 kg. This one weighs 1060 kg. That means with the same battery pack the range is more than 200 kilometers. We’ve been conducting comparative tests on these three vehicles for the last three years, during both winter and the summer. The data coming back is fascinating.”
System Thinkers are the future
“I got involved with automotive about 10 years ago. I realized during lengthy discussions with industry, that the automotive industry was starting to move away from mechanical engineering. Today’s cars are becoming a combination of computer science, electrical engineering and human-technology interaction, and just a 'bit' mechanical engineering. The automotive industry has been asking Universities to train more students because their business is growing. So, I asked them what type of engineers they needed. The answer was clear – we need more engineers who can think at the system level.”
“For instance, if you want to improve a classical car to drive more efficiently, you need to understand combustion technology, but also aerodynamics, mechanics and electronics. For an electric car you need an electrical engineer and a specialist in energy management. In other words, you need people who can think in a connected way.”
“So at Eindhoven University of Technology, we started a new Masters programme in Automotive technology. We did so by tapping into the resources and knowledge across six different departments. And for the last four years we also offer a bachelor degree in automotive, also based on the same kind of system level thinking.”
“Now we also offer a post Master degree course in automotive system design. We believe we're the first in the world to offer this cross-disciplinary approach with such connected thinking.”
“If you think about it, that's quite a statement because it means that Automotive is no longer dominated just by mechanical engineers working in splendid isolation. There is more and more influence from computer science and electrical engineering. We now have almost hundred alumni who have graduated from this program – and many have ended up working for the chip fabrication giant ASML just down the road. System thinkers are always able to see the bigger picture, which is why they end up leading the larger and more complex projects in the company.”
We walk though large swing doors into the next room, complete with an indoor football field.
Soccer Playing Robots with a Social Goal
“This is also my team, building soccer playing robots. They recently returned from the world championships in China. We have been world champions twice and we proudly hosted the Games in Eindhoven in 2013. We follow the rules of 5-a-side football, but remember both teams are completely autonomous robots – humans can simply only observe. Of course, there are similarities with the technology used in autonomous cars, which explains our interest in both”.
“But the ultimate goal is to build autonomous care robots. Our population is getting older, so these kind of autonomous robots are going to become extremely important. Part of the work involves a lot of indoor mapping and guidance systems so we can build robust machines that can reliably take the place of health care workers. This is the kind of back-breaking work which is best done by robots.”
“The biggest challenge at the moment is to build useful and robust machines at an affordable price. If you want to build a robot to do a simple task, like make a cup of tea and bring it to you, it is so incredibly difficult to do that for a price under €10,000 .”
“Being able to reduce the price of precision technology is hard but necessary. My work started in Philips in the days when they were developing the first Laservision and CD-players. The cassette recorded the sound onto magnetic tape. Building a playback head wasn't that difficult. But building a mechanism so that a laser could reliably track information stored on a spinning disc was a triumph for precision mechanical engineering in the late 1970's. It was amazing that Philips was able to mass-produce such intricate technology – and the first machines were manufactured here.”
“Let me show you the third and perhaps the most challenging area we're working on. How can you build useful medical robots to assist surgeons perform the most of delicate of operations? Eindhoven leads the world in precision mechanics, so it's up to us to show what can be done. Our goal is to build a robot that can be more accurate than a surgeon”.
“Ten years ago we looked at the huge medical machines built by the US company Intuitive Surgical. We examined their DaVinci robot which assists surgeons in very precise operations in the thorax or abdomen which cannot be done easily by hand. The robotic arm can get in at angles which are not straightforward otherwise”.
“The problem then was that this huge machine had only limited accuracy and a lot of training before the surgeon could use it. The apparatus is free standing next to the operating table, and there is a long path between the instruments and the patient's body. So I asked my PhD. student to design a new robot which would operate in close proximity to the patient. This not only dramatically improves cutting accuracy, but it means we can put four sensors in the tip of the robot head to give the surgeon haptic feedback – he or she can feel the forces on the tissue and that is not possible with the Intuitive machine.”
“This is the principle of Master – Slave, with the robot following the actions of the surgeon, but then scaled down so as to be far more precise”.
Learning from failure and restarting the venture
“Now, you might be surprised to discover that we can never bring this device into the commercial market. For a start, you need €50 million to build such a robot for the market. And no investor is going to fund a development whose prime goal is to build a machine that competes with the market leader. So, in hindsight, we were naive in 2010 in thinking we could win from the incumbent.”
“But this disruptive thinking did get us noticed by leading surgeons. They pointed us in other directions where precision mechanics was also needed. So a year later, we had a slave robot designed to assist eye surgeons with difficult operations on the retina. We were able to adapt ideas in the previous prototype, but then focus on the need for accuracy and scaling in such a delicate operation. Our robot operates with an accuracy of better than 10 microns, which means we can do operations in the vein of a retina- which is the size of a human hair.”
“That's very interesting because you can put fluids into veins which might dissolve occlusions that are preventing blood reaching that part of the eye. It means that sight could be restored for patients who could otherwise lose their vision in that eye.”
“So we formed a startup called Preceyes with a team of 5 people. Up to now they were funded by large pharmaceutical companies, so our people could conduct proof of concept tests of the procedure. We think such an operation may mean it's possible to put stem cells behind the retina. I'm delighted that discussions with venture capitalists are proceeding well. We need around 30 million Euro over the next five years. We're optimistic because we get regular visits from the top eye surgeons from the United States and Germany who come to test what we've built.”
The arrival of the plastic surgeons
“We've also been approached by a plastic surgeon from Maastricht. He works on the micro-level – reattaching fingers that have been severed from the hand. They also take tissue from one part of the body and use it to reconstruct a breast after invasive surgery. In both cases, they need to reconnect arteries and veins. They each have a diameter of around 2 mm, and the surgeon has to put eight stitches around the circumference of these arteries and veins.”
“Normally it takes two microsurgeons 2 and half hours to connect 2 veins – this is the textbook definition of incredibly difficult work. If they have just one leak, the consequences are dramatic for the patient. They tried to improve the speed and accuracy of the operation by using the DaVinci machine. But it didn't work because the machine isn't accurate enough. So they came to us, and now a brilliant Ph.D. student adapted the design to make a medical robot for microsurgery. The robot is currently undergoing extensive trials at a hospital in Maastricht. We're capturing a lot of useful data which we can use to further improve both the accuracy and speed of the system. Expect to see much more from Microsure.”
“So what we've done is taken the USP of this region – being able to design very precise mechanics and controls – and applied it to areas that no-one has ventured yet. We have also taught ourselves to scale and adapt. My dream is that by 2020, there will be a whole cluster of companies in this region adapting this research to help solve the big challenges that face both the medical and automotive sectors.”
Next wave of robotics startups from Eindhoven
We walk through into another large well-lit laboratory. Maarten introduces me to three of the latest startups.
One is working on applying precision robotics to the deep brain stimulation operation, a procedure used to help Parkinson patients. This pacemaker for the brain involves placing electrodes in certain specific spots. The more accurate they are placed, the less chance that the patient suffers unwanted side effects. You can't help being impressed by how far these students have progressed, not only in designing a much better solution but also thinking about how they will create a viable business.
Next to him, another Ph.D. student is developing a bone removal robot. Together with the Radboud Medical Centre in Nijmegen, they are developing a new procedure to help isolate and remove tumors in the inner ear. Currently, such an operation takes more than 6 hours before the surgeons have reached the tumor to be removed – and the operation results in permanent hearing loss. By using a far more precise robot, the new startup hopes to develop a procedure which is not only faster, but much less invasive. If the surgeon hits the facial nerve in the course of the operation – half the patient's face is paralyzed for life. The same robot may be able to more accurately place cochlea implants – so that patients who were born deaf can hear again.
The final example is a Ph.D. student working on a different but equally challenging problem. Eindhoven is a world leader in the manufacture of Indium Phosphide integrated circuits. They've developed a technique to allow startups to design and manufacture photonics chips that use light instead of electronics to speed up datacenters. That's becoming important as more than a billion new smartphones want to access news, weather and sports from the other end of a fast Internet connection. The Photonics department at Eindhoven University of Technology has come up with a cheaper way of making the chips. But all kinds of optical fibres have to be connected very precisely to these new optical chips. And that's where Maarten Steinbuch's students come into the picture.
“By never losing site of the "big picture", we're seeing all kinds of applications for this nano-precision engineering says Maarten. “If we can connect up chips with miniature fibre optics, then we can help build the even more sensitive electron microscopes in the world. We have a lot to contribute to world leaders like FEI, who build the world's most powerful electron microscopes.”
From light bulbs to microsurgery
“Part of the understanding that we've developed in designing for "stiffness" (as it's known in the trade) is linked to thinking about the degrees of freedom of a mechanism, its thermal centre, symmetry, and flexures. A lot of this originated in the process Philips was building in the 1950's so they could build factory lines to make incandescent light bulbs. And as technology progressed, the same knowledge with enhanced accuracy found its way into lasers for CD players, silicon chip design and manufacture, and so the list goes on. I'm convinced, but also surprised, that the world has taken so long to apply this knowledge to medical-surgical robotics.
But now we've seen the light, it's time to scale up this industry as fast as possible. That's great news for the region, for startups and students, and most of all for the patients who will benefit from more accurate operations which now become possible.
So how do you raise the money to do what the teams want to do?
We have to think differently these days. In the past, there was a decent flow of money from the Dutch government both towards industry as well as universities on the condition that you worked together. But now the laws have changed with the introduction of what Holland calls Top-Sectors. The money flow to industry has stopped and has been replaced by economic tax incentives and rebates. The problem is that these funds get diverted and diluted into the big financial spreadsheets of the CFO, rather than going directly to the project manager at the company who can use it to build relationships with knowledge institutes. I don't believe it works.
To get around this, in Eindhoven we have pioneered a different model. If industry pays to finance a Ph.D. student for four years, the university matches that investment exactly. In that way, there are currently 250 Ph.D. research places being financed by industry. And that works perfectly. We use this momentum in the new TU/e High-Tech Systems Center. It was established late last year to bring together several disciplines, into what we call Mechatronics 2.0, but in 6 months we have already managed to raise €8 million from industry through this matching programme. The Dutch government also has recognized this to be a smart way forward and have added an extra €2 million – they match 25% of what we raise from industry.
Back in Maarten’s office, I’m packing away my camera when I spy what looks like a mechanical rubber band hanging on the doors of a closet. "That's another example of a local invention – and a best-kept secret" beams Maarten. "The old DAF ("Van Doorne's Aanhangwagenfabriek N.V.") invented what's called the "continuous variable transmission”. We have a factory in Tilburg, now owned by Bosch transmissions, that's producing the modern metal version of them. Remarkably, 8% of all cars sold in the world have a continuous variable transmission that come from this region of the Netherlands. And more than 188,000 people have watched the YouTube video of the thing in action. Nobody knows that. But now you do!