For Yuksel Temiz, photographing extremely tiny subjects is just part of his job as a microelectronics engineer at IBM’s Zurich Research Laboratory.
Temiz works on minuscule devices that use microfluidics: a type of tiny, liquid-based circuitry that, instead of using metal wires, directs the flow of liquid through hair-thin channels like a microscopic canal system. Specifically, Temiz and his team develop the underlying technology for miniature diagnostic tools that can take in and analyze samples, like a medical patient’s blood or saliva. But to actually develop and test those tools, engineers need to rely on extremely powerful microscopes, which can cost tens of thousands of dollars apiece.
Eventually, Temiz got fed up with the limitations of those boutique devices and built his own — and it only cost him $300 because he made it almost entirely out of Lego bricks.
And it works: he’s used his Lego microscope to take images that he’s later published the images in prestigious scientific journals including Science Advances, Scientific Reports, and Biomedical Microdevices. Now he’s sharing the instruction manual, which he designed to look like that of any other Lego set, with the world.
Futurism caught up with Temiz to learn more about his Lego microscope. Here’s our conversation, which has been lightly edited for clarity.
Futurism: So the big question is “Why?” What problems were you facing with a traditional microscope that made you decide to build your own out of Legos?
Yuksel Temiz: It started about two years ago. I was asked to provide really high-quality images of our microfluidic chips. I was asked to provide images, videos, for a big tech event organized by IBM. The organizer wanted a kind of an artistic view from an angle, with nice colors of liquids flowing in the channels. They found the top view from a microscope quite boring.
But none of our microscopes in the lab could take high-quality images from an angle. I took a camcorder, a microlens, and a tripod, and I built a very inexpensive setup to take nice images from an angle. I thought I could do it myself. This was already my hobby, so I started building my own microscopes specifically for taking high-quality images of our microfluidic chips.
These chips have reflective surfaces, meaning it’s extremely difficult to take images with a regular camera. It’s kind of like taking an image of a mirror. You see the reflection of the camera, and if the light isn’t uniform you see the point source of light. I discussed with my managers and we thought it would be a cool tool for labs, for schools, and we decided to open source everything.
And has this Lego microscope improved your work at all? Were the angled pictures just for the sake of a more artistic presentation, or do they actually accomplish something for the benefit of your research?
Originally the idea was to take images of the chips for publications and presentations. It’s nicer than the top view — it shows the depth. These microfluidic channels are 3D structures. When you have it tilted to you, you get much more. It’s like 2D versus 3D. I don’t think it has much advantage in terms of diagnostics. When you have blood or fluid in the channels, you will not need a Lego microscope to quantify the samples. This is more for taking pictures for the public.
But we also noticed you can image biological samples — there are these microscopic glass sides available for biology classes. These are plants, animal tissues, cells, and when I put them under the microscope, I get really nice images. So then I thought yeah, maybe such a microscope would be good for schools, for kids’ education. It’s not just for labs, it’s a nice microscope.
So, I see you have some papers in prestigious journals: Science Advances, Nature’s Scientific Reports, and the like. Are the images that appear in those papers the same ones you captured with your Lego microscope? Did you have to retake any of the images with more traditional instruments for the final proofs of the paper?
Typically, what we have is the tilted view of the chip. Then we have an inset next to it with a top view of the specific structure. In the Science Advances paper, we have this clock. In the video, we have the tilted view of the clock, and next to it we have the high-resolution image from a microscope.
For a scientific journal, we have the beautiful image of the chip, but then we have the data and the image taken from a high-end scientific instrument.
I do not think that the Lego microscope is good for any diagnostic quantification, I have to emphasize. But it works well for checking circuit boards, insects, small objects, anything you can imagine. It’s like a regular camera but at a high magnification.
Did any of your paper’s reviewers or the publishers themselves comment on the microscope, or did they accept it within your methodology? What did they have to say about it?
No, in general, we get very nice comments about our figures. This is kind of my hobby: as an amateur graphics designer, I do a lot of 3D animation and 2D designs. In general, we get nice comments about the quality of the figures, but not the specific angle of the 3D images.
The journals publish our images, but we don’t have the chance to use most of our images in publications. We often use the images in our presentations. I presented our work in the big conferences in our field with up to 500 to 1,000 people, and the first thing they say is we have beautiful images.
Because it’s made of hand-assembled Lego bricks, have you tinkered with the microscope at all to change how it works? For instance, could different setups give you different functionality?
I had three more prototypes before this. The last one had a few Lego bricks, but the first one didn’t have Lego at all. It was all 3D printed. At IBM we have a very nice machine shop and very talented technicians, so they helped me. Initially, I didn’t use any Lego bricks, and the reason I started using Lego bricks is I wanted to have the microscope modular and have it be configurable.
In this microscope we can take transmitted light images — a classical technique with the light at the bottom, then the sample and the camera. We can take reflected images, with the light from an angle and the camera at an angle, and we can take cross-sectional images with the microscope orthogonal to the sample. Sometimes the sample is big, sometimes it’s small, sometimes you need a little more space between the camera and the sample.
I started 3D printing interlocking pieces, and I realized that what I was printing was almost exactly a Lego brick. So I said, “Why should I spend time on 3D printing?” I started adding Lego bricks to my design, a little bit more and more, and it became a hybrid 3D printed-Lego design. And then after we decided to open source everything, I completely redesigned the microscope in a way that people can — if they have a 3D printer at home — build it at home.
The design is much more maker friendly, it doesn’t require any professional tools.
Are all the bricks off-the-shelf Legos?
They’re all Lego bricks you can go and buy individually. Lego has a shop called “Pick a Brick” where you can individually buy them. I 3D printed a few components just to connect a Lego piece to my custom-made component. I 3D printed an adaptor to attach the motor to the Legos, for example.
You mentioned that these could be used to improve access to scientific research in school systems or developing countries. Could you tell me more about what that would look like?
I have goals for that. A lot of people are asking me if they can buy it as a kit. We had plans for that but due to COVID we had to postpone them. The ideal would be we call an electronics company, see if there’s support to provide it along with Lego maybe. The electronics company would provide the electronics as a kit, maybe Lego could provide it as a one-click shop thing.
The advantage for schools is that first, it’s Lego — it’s interesting for them. And then students can use their imagination to put a microscope together with different components. There’s a little bit of electronics, and then there’s the image processing part. It’s all based on Python, so kids at schools can start programming simple things. Those microscope slides for biology classes could be already interesting because the kids can put in the sample, and the software can already tell “Hey this is a cell from this plant,” things like that.
How long would it take to build one from start to finish?
Assuming everything is in front of you and the electronic board is already available because the soldering part is not something a kid should do, half an hour. I reassembled it for the interview and it took half an hour to make it, but for someone who doesn’t have experience, maybe an hour?
Just the assembly, following the PDF document, would take you about an hour.
And you made Lego-style instructions, right?
Yeah, it’s in the PDF. I spent more time on the documentation than the microscope itself, just to make it easily understandable by anyone.
What else do you want people to know about this? Is there anything that I, someone who played with Legos as a kid but certainly isn’t an engineer, wouldn’t know to ask about?
It’s not a scientific breakthrough, obviously, it’s just putting components from the shelf together.
Scientifically it’s not interesting — compared to what we do in the lab, this is not something you’d publish in a high-end journal. But for many labs working on microfluidics, this could be a helpful benchmark tool.
Editor’s note 4/30/2020: This story has been updated to clarify the definition of microfluidics, the size of the audience at conferences, and that Temiz and his team don’t interact with medical patients.
More on Legos: Lego Bricks Are Excellent Insulators at Cryogenic Temps