The subject of Kasper Understrup’s master’s thesis has the potential to reduce DTU’s biochip expenses to a fraction of their current level.
“We’re pretty much in the same place the pioneers were back in the 1960s, when they sat hunched over their workbenches, developing the first computers. Computers as we know them today simply didn’t exist at that time, and no-one was quite sure what the possibilities were. But as soon as we find a specific way to use this, I’m sure it will skyrocket—just like the first computers did.”
That is how Kasper Grue Understrup describes the situation that he, as an MSc student on a course at the Embedded Systems Engineering research section under DTU Compute, has found himself in while developing and describing a microfluidic lab-on-a-chip.
A nanocomputer for laboratory use
This is what is known as a ‘biochip’, i.e. a kind of chemical nanocomputer that can be programmed to mix very small volumes of fluids in specific ways. Once the system is fully developed—which may still take several years—it will prove invaluable in testing chemical reactions, in connection with the development of pharmaceuticals, for example.
“What I have designed is a system for ‘pocket-size’ chemical labs. The intention is that, at some point in the future, it will be possible to run a million identical chemical trials—and have the results ready the next day. I have defined the programming language, a compiler (that translates the programmed commands into actions) and a hardware platform that can be built up for specific purposes,” he explains.
Multi-disciplinary
Working with biochips and nanocomputers is a complicated task where deep understanding of multiple disciplines—including chemistry, computer science and biology—is required to come up with a viable result. And it is precisely this broad understanding that Paul Pop, Associate Professor at DTU Compute and Kasper Grue Understrup’s supervisor, highlights as the very special feature of Kasper’s work.
“I am most certainly proud of what he’s achieved. It is usually difficult to find students who can work with this area, specifically because it demands in-depth knowledge of multiple fields,” explains Paul Pop, who adds that the result of Kasper’s work is quite a breakthrough:
“It’s important to our field, because up until now we haven’t had access to a programming tool like this,” he says, going on to explain that the tool they have been using thus far costs around USD 30,000 (or almost DKK 200,000), while Kasper Understrup’s new model will cost roughly DKK 3,000.
Award-winning master’s thesis
Kasper Understrup was presented with the 2014 Embedded Award for his Master’s thesis, an honour he has already put to good use in practice.
“I was looking for work—without much success—when I received the award. Within a week, I had been headhunted for a system developer position at the Internet provider one.com,” he relates.
As regards DTU Compute, the award helps to turn the spotlight on an area of research for which it is normally difficult to raise money, precisely because it encompasses multiple fields, as Paul Pop explains.
Kasper Grue Understrup himself says that he would love to return to DTU in the future to continue working on his programmable biochips.
Article in DTUavisen no. 10, December 2014.
Kasper Grue Understrup completed his master’s thesis at the Embedded Systems Engineering research section under DTU Compute, working with a team of researchers including Ulrich Krühne, Søren Heintz, Jan Madsen, Martin Dufva and Henrik Bruus, and with supervision from Associate Professor Paul Pop.