Episode 46 – Theodor Nielsen, CEO, NIL Technology
Jimmy Carroll: [00:00:06] Hi everybody. My name is Jimmy Carroll. I’m the vice president of operations at Tech B2B Marketing. I’m here at SPIE Photonics West 2024. I have the pleasure of being joined by Theodor Nielsen. Theo, thank you so much for taking the time. I appreciate it. And of course I’m with John Lewis, and John, thank you, of course, for being here with me. Is it NIL Technology?
Theodor Nielsen: [00:00:29] NIL Technology or NILT.
Jimmy Carroll: [00:00:32] NIL and NILT. I like it. For those who don’t know, could you tell us about the company and what you do there and what you’re most excited about here today?
Theodor Nielsen: [00:00:40] Yeah, I can definitely. Thank you. We have a long background in nano-imprint lithography, so that’s NIL. And that’s like our name and our legacy. And we started the company 18 years ago focusing on masters for this nano-imprint process. And then five years ago, we decided to take a new path. We had had some years of profitability, was going well, but we had greater aspirations. And we decided to engage deeply in the world of nano-structured optics. So today we are actually still keeping our mastering business. So that runs, but we are fully engaged in the development of mass production of meta lenses. The vertically integrated strategy from design and all the way.
Jimmy Carroll: [00:01:23] So meta lenses. This is, I’ll admit, this is a term that I hadn’t heard of until today, so I’m curious to hear it directly from you what that is and maybe what that means for different applications that your customers work in.
Theodor Nielsen: [00:01:35] So meta lenses are — essentially it’s a diffractive lens. But you are patterning a surface with a lot of nano-structures that each of them works as a resonator. And eventually you can actually take a curved surface and replace it with a flat nano-structure surface. And because you have lithographic precision over these nano-structures, you can actually create more wave function. You can have more functionality in one surface than you can with a curved lens. So as a rule of thumb, we say we can take the performance of four lenses and integrate it into one flat surface.
John Lewis: [00:02:14] And so the advantage there is a compact size.
Theodor Nielsen: [00:02:17] Compact size. So size. We call it the SWAP: size, weight, and power. So we have when we are building systems and so they also brought a sample of an eye-tracking camera or a very miniaturized camera consisting of only meta lenses here. And we reduced the number of surfaces also in the system. So there’s less stray light. We have more control of where the light is going. Of course you can argue that it’s diffractive. There’s always some higher diffraction always, but we can control where they go. So we control the system and create some very mind-blowing performances with an extreme compact form factor. And I think that also answers the second part of your question. So that is what excites me the most. It’s actually what we are bringing to the show here. We have got a lot of attention. I think it’s the first time someone demonstrates a camera consisting purely of meta lenses where that’s more than one meta lens in it, and I think this is something very new, something we worked on for a long time.
Jimmy Carroll: [00:03:20] So eye tracking. I’m going to ask you about some of the other applications too, but I want to start there. What are some of the benefits of having such a small camera for eye-tracking applications? And what might some of those be?
Theodor Nielsen: [00:03:32] You get very fast inspired from the show on the other side of the street, SPIE AR VR MR, and there you have devices that you have to wear on your head, head-mounted devices. And so anything that’s light, very small, I mean, that’s more comfortable to wear. That’s the first thing. Secondly the full vision of AI. You want to wear glasses like you are wearing and you want to integrate sensors and eye-tracking systems in that, and you basically don’t want to notice them. You just want to benefit from them.
John Lewis: [00:04:04] So by tracking the eye in an AR/VR headset, you can optimize the performance of the . . .
Theodor Nielsen: [00:04:12] It’s a human–machine interface. So you have your hands free and you can basically control whether it’s an imaginary screen and a cursor. You can control whatever panel with your eyes. It’s quite mind blowing when you try it for the first time.
John Lewis: [00:04:30] I think it was 2016 when I was editor in chief of Vision Systems Design magazine. I received a phone call from a gentleman and he was saying, you know we optimize resumes. He was like a professional kind of job placement person, you know, and we want to optimize the resumes we write for people, and we want eye-tracking technology so that we can see how people read these resumes. And this was before — AI probably just reads all the resumes now — but this was back in 2016, I think. So like eye tracking, you know, one would be tracking the eye to see how someone looks at something. What are the other applications for eye tracking besides AR/VR.
Theodor Nielsen: [00:05:16] And another area would be within the automotive industry, you know, build it in to make sure that you are paying attention to the right things when you’re driving the car.
Jimmy Carroll: [00:05:24] And your eyes aren’t going . . .
Theodor Nielsen: [00:05:25] That’s drowsiness. You want to avoid that. But distractions. You should keep your phone in your pocket when driving.
John Lewis: [00:05:32] You’re looking down . . .
Theodor Nielsen: [00:05:34] Exactly. So that’s just what I can come up with. And if you really start thinking about it, it’s also all the touch panels, you know, in public places and then now they place alcohol next to it to disinfect. But what if you could just control whatever you are ordering with your eyes.
Jimmy Carroll: [00:05:55] Yeah, yeah.
John Lewis: [00:05:56] A kiosk. ATM machines, all kinds of things.
Jimmy Carroll: [00:05:58] I don’t know if you’ve ever saw the movie “Minority Report.”
Theodor Nielsen: [00:06:05] I think I’ve seen it.
Jimmy Carroll: [00:06:05] Yeah, there’s a scene where, I think it’s Tom Cruise. He’s walking through a mall. And when you look at this billboard. It’s not a billboard; it’s more just an LED screen ad. But when he looks at the ad, it’s customized to him and then serves up something in particular. And I don’t know, technology like that is maybe not as far away as others.
Theodor Nielsen: [00:06:30] I think that could definitely be be realized today if people put their shoulders to it. I guess maybe that already has been realized.
Jimmy Carroll: [00:06:39] Beyond eye tracking, what are some of the most common applications where your meta lens technology is deployed?
Theodor Nielsen: [00:06:45] We’re looking a lot at 3D sensing, again, miniaturizing the optics sometimes. I mean, we are doing some reference designs and showing what it can be done, but we also have customers that brings in the design. And we’re just realizing lenses for such purposes. But again, miniaturizing the optics, the modules such that you don’t really see it, and there are different technologies there: time of flight, structured light. And also here at the show, we’re demonstrating, again, a multiple meta lens camera for iToF, indirect time of flight 3D sensing. So it’s cool.
Jimmy Carroll: [00:07:24] So time of flight. I know that time of flight sensors in industrial automation, for example, are used in a lot of different ways in the warehouse and beyond. But if yours is miniaturized, what are maybe some applications where a smaller time of flight camera might be used?
Theodor Nielsen: [00:07:42] I think, again, it’s really the areas where you are mostly volume or size restricted but also weight. So we’re all wearing wearables. I mean the phones, for sure, other types of wearables. You can think of drones also, like where weight would become an important factor. Consumer electronics. And again also cars. That’s not that big a difference from some of the optics you are creating, whether it’s a 3D sensing system for a phone or it’s a lidar for a car. So again, many of these 3D systems both have a transmit side, where you are illuminating a scene either with a diffuse light or a dot projector, and a receiver side. So we’re doing both these sides. And that’s very similar in the lidar both for medium-range detection and long-range detection.
John Lewis: [00:08:40] You talked about being vertically integrated. Could you talk more about what that involves in your context?
Theodor Nielsen: [00:08:49] So when we started looking at meta lenses five years ago, we were met with a lot of questions about basically efficiency. How much light do you lose to higher-order diffraction. So that was the first task we really looked into. And we realized we had to design our own simulation tools to design these meta lenses and make sure that we could design according to our prototyping and manufacturing process. And a few years ago, we managed to create a focusing lens with 94% absolute efficiency. It was like, okay, this is so close to the theoretical limit that okay, now we have the fundamental building block in place. So we’ve been expanding that design box. And now we are designing all our meta lenses ourselves. When that is set we can work with other people’s designs. We’ll come back to that. And then a lot of people had the question about, can you just drop in a meta lens into our system? But we realized since meta lenses are diffractive and many existing systems are more refractive, the module has to be designed to benefit from the meta lens properties. So we are looking a lot on systems and making reference to science and showing people what can be done. So that’s the design part of it. We are then prototyping it using all the in-house tools: electron beam lithography, etching, and we have optical characterization labs and testers, reliability testers. And then we do manufacture by nano-imprint lithography and etch. So that’s a very similar process to the EB minutes. So if you do for prototyping. So we are setting up our manufacturing. We have the first line in Europe for prototype volumes and small volumes. And we are building a mass production line as we speak in Asia.
John Lewis: [00:10:45] And what type of volumes are you expecting to be able to achieve?
John Lewis: [00:10:48] We are looking at consumer electronics volumes and whatever shape they come in. So we are ambitious, but I cannot comment much more on the details there.
Jimmy Carroll: [00:10:59] Understood. Understood. Indeed. I want to go back and ask a little bit about AR and VR. Because it seems to be a growing focus. It is a growing focus here at the SPIE show. What’s your take on that? So AR, it’s big in gaming and entertainment and some of these virtual concerts and things like this or experience a soccer game or a basketball game through the glasses. That kind of thing is cool, but it’s expanding into industrial too. Like, what were you talking about, John? Like training, troubleshooting, or engineers might be looking at a part together. What’s your sense of the AR market? Is it becoming more mature?
Theodor Nielsen: [00:11:46] I think it is. That’s also other cool applications within surgery and medical assistants, right? I think it is. And I think there’s so many things happening that it’s starting to become so relevant that it’s come to stay now. That’s a big thing with the spatial computing coming out lately. I think the next week they’re going to launch that. And that’s very interesting. I think the Ray Ban and Meta took an approach from the other angle to creating sunglasses with the cameras in. And I think actually that it’s very important that people start to think of sunglasses as something you need to charge. I think that’s actually a small thing that, it makes me smile when I realized it, but actually sunglasses are now a thing that before you leave, you have to make sure that they are charged. And I think that’s actually one of these nudging things of how we perceive things and how we familiarize ourselves with new things. I think it will come.
Jimmy Carroll: [00:12:51] John would hate to charge his glasses, I know that. You touched on this a little, but what are some of the unique optical challenges in AR and how are meta meta lens uniquely suited to these applications?
Theodor Nielsen: [00:13:10] So, I mean, actually we are involved in the AR from two different angles. So we actually still are doing masters, making masters for customers and especially for making AR wave guides for the industry. So we are very much involved in that. And that’s maybe the area where my company is mostly exposed to all the development efforts that they are. And we’re having some deep programs in optimizing how complex and nano-structures we can do on the master level such that people can create very highly efficient wave guides for the glasses. When it comes to these more compact systems where the cameras and sensors, we don’t see requirements that are in many aspects that different than they are from other industries, like there’s a need for, again, the SWAP: size, weight, and power, and people are very interested to see what we can do with this new technology. I mean, now this is a podcast. And so I think maybe people cannot really appreciate how small this camera is. But much fewer surfaces in this whole camera system. Simple to assemble. Very, very robust. There’s no plastic in it. The temperature’s stable and these things. So people appreciate that.
John Lewis: [00:14:36] The camera, I don’t know, it looks like maybe the size of a long grain rice or something.
Theodor Nielsen: [00:14:43] Maybe half a rice. One-third.
John Lewis: [00:14:46] It’s pretty small, for those that can’t watch this. Yeah, but you had mentioned the eye-tracking camera and the depth camera. Others cameras you’re working on, other types of cameras?
Theodor Nielsen: [00:15:04] Already last year we demonstrated here at the SPIE an extremely simple camera, which we call it the 1M camera. So it’s basically one nano-structured surface in front of the sensor. It’s a little bit of a rigid optical system, but for machine vision purposes this is ideal and very, very simple. So that is also something that’s one of our fundamental building blocks in our whole understanding of how to build modules, how to build cameras using this type of technology.
John Lewis: [00:15:37] I went to a presentation once by a Roomba design engineer who developed a machine vision system that’s in a Roomba robot, and I think of these mobile, portable, compact applications where they need a camera. Is that where you might see these types of things?
Theodor Nielsen: [00:16:00] I think there’s a lot of things happening these days. AI is coming. So you can do a lot on the signal processing side. So the the first thing you have to ask yourself is what the camera sees, is that intended for the human eye or is it intended for a machine to interpret it. If you really go that way, you can sacrifice certain optical parameters that you normally would appreciate to maintain in photography that is tended for the human eye and sacrificing such things and then optimizing on compactness. Then you can really create something, but it has only a functionality together with an algorithm that looks at it. So this is where my company is. I mean, we are working with other companies because we don’t have that side of it. We’re looking at the hardware and our focus on the manufacturing of the lenses and demonstrating what we can do and modules. So combining those two things, that’s where the magic really starts to happen. And that’s also in that space where I see meta lenses today is actually mostly in the near infrared, the SWIR region, visible region but still monochromatic. And there’s a lot of interest in bringing meta lenses to broadband visible. And I think that will happen when you combine the meta lenses with the signal processing. So from our perspective, making a meta lens in the near infrared and SWIR, we’re etching into silicon. When we have to make a meta lens in the visible spectrum, we’re working with a different material. So we have the processes to make these lenses, but we are not putting equally much effort into the design and the visible spaceship. But if customers have their own designs there, we can process them for sure.
Jimmy Carroll: [00:18:02] So beyond the chargeable portable sunglasses, do you have any general predictions or trends that you expect to see in the coming years and within the purview of, let’s say, the photonic space, machine vision, industrial?
Theodor Nielsen: [00:18:19] This is terminology of a sense sensing, right? We see more data, more high-accuracy data basically to make better predictions and better decisions. I think that’s one area. Another area is a new type of sensing. I mean hyperspectral. I mean different type of spectroscopy. Whether that is air quality, skin hydration, food freshness, authenticity of certain things. I think that is a very, very interesting area.
Jimmy Carroll: [00:18:54] Theo or John, anything else that we haven’t talked about here that you feel like we should be covering?
John Lewis: [00:19:00] Nothing for me, Jimmy. Thanks.
Theodor Nielsen: [00:19:02] Thanks for having me. And thank you for the opportunity.
Jimmy Carroll: [00:19:04] Of course. I know how busy your day must be, so I really appreciate it. For anyone who would like to learn more, it’s NILT.com. And if you have questions or comments, we would be happy to pass them along to Theo. You can reach out to us at manufacturing-matters.com. And thanks very much.