The AR Show: Ari Grobman (Lumus) on When We Will Get AR Glasses We’ll Want to Buy

12:26AM Jan 31, 2023

Speakers:

Jason McDowall

Ari Grobman

Keywords:

waveguide

companies

glasses

optics

display

smartphone

image

technology

people

big

consumer

terms

product

early

smart

ar

years

market

ari

aesthetics

Welcome to the AR show where I dive deep into augmented reality with a focus on the technology, the use cases and the people behind them. I'm your host Jason McDowall. Today's conversation is with Ari Grobman. Ari is the CEO of Lumus, a company developing a novel optics solution for AR glasses.

Before going any further, let me provide a little perspective upfront when I consider the biggest hurdles to delivering consumer grade AR optics and displays are at the top of the list. Although it's worth noting the display and the combiner optics are two separate but intimately connected engineering challenges. Lumus sells an optical engine that includes a micro display from a third party. But the real magic is in the combiner optics, the part that directs the light from the display system into our eyes and combines it with light from the real world. Lumus has taken a different technical approach than most waveguide combiner optic companies, rather than use the principles of diffraction to bend the light into and out of the lens. Lumus uses the principles of reflection effectively partial mirrors to get the light into and then out of the lens. Proponents of the diffractive approach cite that the manufacturing process is relatively easier and cheaper, and the performance of the lens is good enough with a very thin form factor. Lumus argues it solution provides better performance with a manufacturing process that is good enough to be reasonably priced for consumer grade glasses as the market scales.

Now back to Ari. Ari spent most of the early part of his career in sales, including 10 years as the VP of sales and business development at Lumus before being promoted to CEO about six years ago. Ari has led his team through multiple successful customer engagements, advancements in R&D, and continued success in military and enterprise sales. In fact, Lumus technology is used by spinal surgeons across the US, as well as fighter pilots flying a 10 and F 16 aircraft.

In this conversation, Ari describes what he's seeing generally across their various customer engagements, including some of the product strategy, potential early use cases and the timing of new market entries. As a way to help frame some of the discussion are described some of the similarities and differences between components for smartphones and smart glasses. Ari goes on to discuss recent advancements that company has made in both the waveguide technology namely delivering a bigger virtual image from a smaller display by expanding the image in two dimensions within the lens, as well as progress in improving the manufacturability of the lenses. Ari also describes how Illumina sets themselves apart from the competition at both in enabling enterprise and consumer grade smart glasses. As reminder, you can find the show notes for this and other episodes at our website, V ar show.com. And please support the podcast@patreon.com slash the AR show. Let's dive in.

Ari, I've got two kids working their way through the public school system here in Los Angeles. And I've come to appreciate the prescribed path isn't a great fit for a lot of kids. It's sometimes a struggle to accommodate the needs of kids in the system. You grew up in LA and worked your way through the school system here. Do you have any memories of that time when growing up here?

Yes. Especially, you know, when it comes to education, when it comes to teaching teaching methods? I think, you know, first of all, I would say teachers are the unsung heroes of society. They're quite pivotal to young people, people, you know, coming up in the system. on many levels, it's a big responsibility as well, they could kind of make or break, when you talk about not just educating, but building up self esteem. I think that's something that's very important in our society. So actually something that sticks out and something that I take with me, you know, I remember a distinct memory. You know, I think it was towards the end of seventh grade, it was parent teacher student conferences. So the child is in the room, so there's no way he said, she said on it, everyone's there. And I was, you know, subject to a particular rant from a, I don't want to use the term overzealous teacher. But I think, going a little too far. And I think playing a little bit of a power trip. I'm thankfully my parents were there. And my parents always have my back. And that's something that I think is very important. And you know, right on the spot after listening to the rant, you know, my dad didn't even look at me, he just like, I know the kid. He's like, I hear what you're saying, not buying it. And you know, without going into the gritty details of how articulately he put her in her place. The key takeaway I got from that is having somebody who's in your corner, somebody who backs you up. Now, in many cases that can and should be a teacher, I was very fortunate that irrespective of you know, the good teachers or bad teachers along the way, I definitely had, you know, fantastic parents that were in my corner, definitely helped build my self esteem, you know, helped me push forward. I think, shortly thereafter, within a few short months after that, I ended up saying, okay, you know what, let's move to the next level, and I ended up skipping eighth grade. That's something that I take away, you know, very much I try to incorporate that in my parenting. I've definitely had my run ins with principals, teachers as well, if I thought they were taking things in the wrong direction and very much let my kids know that I backed them up. But that's something that I think we also take to the company and my management style, obviously how hiring good people, I think that's something that's something that's critical. But then also backing them up, you know, giving them the leeway, the space and giving them the support, we're in your corner, you know, you got this go do it, there may be falls along the way, you know, maybe stumbles, maybe screw ups. And believe you me throughout the course of every company's history, hours included, there have been a lot of screw ups. But again, you back your people, because ultimately, companies are built. And really their success is built on the people and the work that the people do. So that's something I've definitely taken away from, from my upbringing and tried to bring it to my everyday life and bring that into my, into my job here as well.

That's so amazing, these sort of early childhood experiences, some of them really stand out. And often, at least in my own memory, it's often the ones that are emotionally charged, that really stand out for me. Yeah. And sometimes for the better, sometimes for the worse, correct. And I try to, you know, learn from all of those. It's so great that you kind of reflect on that and you take with it, this idea that you have to support you have to support your team. I work with horses, as well as your children at home. I think the last time we had a chance to chat on this podcast, it was really right near when we were getting started in the first maybe for 612 months of the podcast. So it's been a couple of years in the industry is evolved, Sam, and you've been now in the role as CEO there at Lumus for half a dozen years or so I think, yep. You kind of reflect on that childhood experience. And you think about the experience you've had now here at Lewis, how have you evolved, like, over just the last six years, how have you evolved as a CEO?

It's definitely been a journey. I think, you know, over time, you learn, especially as you grow into the leadership role. This was my first time actually managing people. So it was a big step up, I got a big vote of confidence that my team back me up right away and did rally behind me. There's a lot of things that when you're kind of in a different position in bizdev, there's a lot of things you just take on yourself, I think learning to trust the people and to delegate and you know, hands off, okay, you got to take care of this, I can't do everything more than that. You're actually the domain expert in this, go ahead, go run with it. You tell me what you think, you know, people come to me for decisions. Well, tell me what you're suggesting, you know, I am relying on your domain expertise. So over time, definitely evolving and trusting people, I think it's something that that I've grown into, you know, it was kind of going a little hands off and really saying, Okay, you guys are driving this part and go for it.

Yeah, that trust goes a long way, in helping bring up people's best, I believe.

Yeah, absolutely. And we do better. I mean, if I micromanage things, you know, God help us.

So in the last handful of years, you know, Louis was already involved in a number of smart glasses, AR glasses sort of projects back in, you know, 2017 ish or so when we maybe had our first conversation. But over the last handful of years, you've continued to evolve the technology. And you continue to engage with even a broader range of the largest of large tech companies who are pursuing opportunities in this space. And I thought maybe we could spend a few minutes digging into your perspective on what you're seeing from what they're doing. And we start with this, this notion of the overarching strategy, what is the overall strategy or set of strategies that you've observed of emerging from these large tech companies?

So there's a couple of things there. First of all, you know, in terms of our evolution, I'd say, you know, we were always best in class, when it came to image quality efficiency, I think, you know, where we needed to up our game, in order to be relevant for the consumer market is to be able to work on form factor, and also to be able to push hard on manufacturing, and be able to get to consumer price points. I think in parallel, we've seen that migration or we're seeing, at least from our perspective, that migration with the top tier companies, where this was looked at as a, you know, the beginning, more of a enterprise kind of play, and certain companies may still make their foray into, you know, professional applications. But ultimately, the AI is towards a more broader consumer reach. So from the industry side, we're seeing and feeling that push in that direction. And from our side, we knew we had to make a couple of key steps in the technology in order to be a lot more consumer friendly. Namely, when it comes to form factor and manufacturing capabilities.

Yeah, we'll definitely come back to the kind of the evolution in the your product itself. But we can stick just for a second here longer on this. And so there's a lot of investments happening across these tier one, these big tech sorts of companies beyond kind of the shift. And certainly we saw the early introduction from Google was initially anticipated hoped to be a consumer product begin, but then survived as an enterprise product, right Google Glass, and we saw from Microsoft, this consistent investment we have at least had seen in the past consistent investment in the HoloLens product line, which is very much focused on enterprise. But now we see all of this effort behind Scenes, least we get hints at peeks at the effort behind the scenes from these other companies to build something that's consumer oriented within the set of consumer products, are there different classes of products or different target markets that they're anticipating, or you're kind of you are anticipating through your work with him on the consumer side? Well,

you know, what they're thinking based on the specifications, you know, that that are getting put on our table. So we kind of see two camps, there's the super lightweight data snacking AR, what Google Glass, let's say, should have been, you know, just more elegant look like a real pair of glasses, but you know, relatively lower field of view, minimalistic data overlay. And then, of course, the much more immersive type of AR, you know, by an ocular, a lot more processing, the concept of, you know, recognizing what's going on, you know, in the ambient world and tying that information directly to that. And then different levels in between, you have your super immersive where, you know, you're talking about gaming, or, you know, first person shooter games, you know, we'll need something like 80 degrees, you'll have something more like, you know, in between that 5060 degrees sweet spot, and then you'll have, you know, the, the lighter weight, you know, less than that, you know, 40 degrees or less, and in many cases, just one eye. Now, it's interesting to see that, you know, several of the companies we're talking to, are actually attacking that market from multiple angles, you know, the super lightweight, and the more immersive, and quite possibly, it looks like launches, just based on timelines may start earlier with a lightweight version, and, you know, shortly thereafter, come with a much more immersive, and this is something that we're seeing across the board. So I'm not giving any, you know, any company's, you know, particular secrets away. And I think, you know, ultimately that that points to the fact that everyone wants a more immersive AR, it's how we get there, how we educate the market, I think, you know, starting with the path of least resistance, super lightweight, easy on the battery, let's educate consumers to start wearing something, let's make sure that there's value in that product, let's get them used to charging that device every night. And then, you know, they love it a year and a half, two years later, boom, here's your glasses, we're taking it up a notch. And I think that is, you know, relatively speaking, a pretty smart approach. The other thing I'd say is, we're all going to be a lot smarter after these launches happen, you know, we're going to find out what works, we're going to find out what consumers want. Once we open up to developers, you know, we're gonna find the real killer applications. But again, that's a whole nother subject to unpack.

One of the questions maybe on the tip of most listeners minds is when when might we see this first batch of the of the earlier simpler version of these data stacking classes?

So I should correct one thing, we may see launches around similar timelines, depending on which customers we're seeing. And depending, you know, what gets locked in and time not everyone is doing this two step approach. Some companies are just, you know, sidestepping, and going for the more immersive. So I should put that out there. We've heard of launches as early as late next year 2024. Realistically, speaking from what we could see from the inside, you're probably talking about mid 2025. Because, you know, when companies give you their timeline, the good thing is, you know, we're very confident we won't be the bottleneck. You know, you can't blame the optics anymore. We've really come a long way. But I think just putting everything else together, you have to assume deadline slips and stuff like that. But I think 2025 is very reasonable to expect. And if I put a handful of companies together that 2025 2026, I think you'll start to see a lot of I should say a handful, at least the ones that we have a front row seat, we should start to see those and things could get pretty interesting.

Yeah. Wow. For so long, I think for every year that I've been in this industry, it's always three years out. Oh, it's three years, three years. Three, now we moved in now. Yeah. Which is great. It's amazing. What? Well,

welcome to my world, or our world, I should say.

As you imagine, you kind of think through you hear these conversations, you brainstorm within your own company, this set of use cases and the set of devices that are necessary to support them across these sort of AR glasses. Which do you think will be the more common we can anticipate the earlier ones will be the lighter weights, data snacking, less capable devices, but as we have market adoption, which do you anticipate being the more prevalent, I think

there's going to be more you could do with a more immersive, but the value obviously has to be there that consumers are going to take this with them everywhere they go, or at least it's going to be in their, you know, in their briefcase like their laptop. I'd say jury's still out. We're all looking at it. And actually, when, when we're told to be ready for the big numbers, I can tell you there's a huge variance. You know, it's very hard to get companies to actually commit to significant now mean against significance a relative term, companies believe that at launch, it could be a couple of 100,000. But they also believe it could be a couple million in their first year, they actually do not know. And these are the conversations you have. So when it comes to planning, you have to be ready to snap, and jump into, you know, mass production a lot faster. So you have to already, you know, when it comes to the supply chain, there's gonna be a lot of investments and a lot of like, okay, like, you know, how quickly can we scale this up? Do we want to put so much into inventory? These are, these are the, you know, thoughts, everyone's thinking right now, like I said, earlier, we're all going to be a lot smarter, once product gets into market, there's going to be a learning curve, some things are going to flop, some things are going to be runaway hits, which ones I wish I had that crystal ball, I double down and just stop doing r&d On the other stuff. But we offer the buffet table, the capabilities, whatever the OEMs want, you know, we're there for them. So we're in it together with them. It'll, it'll be a fun journey.

One more on the kind of the the market perspective here, before we shift, but as you either kind of think internally or as you collectively hear the stories from these partners of yours. What is the collective belief that the early use case or set of use cases will be that might drive this early adoption? What's the current hypothesis? Like? We don't really know until they're out there, and people are using them? But what's the what's the initial hypothesis for what's going to take to get people to buy, you know, 100,000, or a million of these things?

That's an excellent question. And it's not something that I can necessarily tell you that companies are telling me what their exact plans are, specifically, because you know, we're position is Intel Inside, you know, enable all the smart glasses makers with our optics, they know that they know, it's an open marriage and these relationships. Again, we don't share data with anyone, we've never had any leaks or anything like that from our company. We're very good about that. But having said that, most of these companies are pretty guarded, in terms of the actual applications that they're talking about. I mean, things that we see based on specifications, or certain companies, other products and trajectories, I mean, you could kind of, you could kind of guess, or you could actually look at, you know, what some of the companies have already telegraphed, you know, in their marketing concept videos, anything from immersive gaming, to telepresence to simple navigation, just driving and, you know, being able to see my ways overlaid on my view, when I'm driving, as opposed to take my eyes off the road. And looking at a phone screen. There's so many different applications, so many different ways to go at it, we'll see which one is the killer application, one thing is clear, every one of these companies has at least one or two, quote unquote, killer apps that they know they have to put at launch. Because that compelling reason for a consumer to buy, it has to be a very well articulated use case, customer is going to shell out, you know, a couple 100 bucks can be the cost of a smartphone, low end high end, again, depending on the offering. But that's kind of the expected range of these things. People have to be convinced from the marketing launch that, Oh, I gotta buy this thing. So that will be there. But you know, the phenomenon that you see with you know, smartphones, you see it with tablets, once you put it in the hands of the developers, things could get really fun and get really crazy and will wonder how we lived without these apps before? What they are, again, nobody's telling me we could all guess I think everyone has their own things that they're betting for. But you know, I don't want to put myself in position to say that I actually no, we just show up with the specs show. But the spec, they say do this, we do it. Was that a very long non answer?

Perfect, perfect. Maybe we can take a step back and and talk about the specs from the perspective of how Lumus fits into the puzzle. So there's a lot of pieces software and hardware and components and in companies have been working at this for a long time, right? We had this we noted Google Glass has been around for a decade, believe it or not, well, and HoloLens been around for what is it eight years now or so these devices have been out there and and there's always this expectation that the hardware is going to progress just like software does, right. Or just like the number of transistors we can fit on a computing chip works, there's a steady progress towards rapid improvement. But it hasn't been the case hasn't been the case with optics isn't the case with battery technology. It's a different sort of grind. So maybe you can just take a step back and describe how Lumus fits into the overall puzzle for the thing that we're in. Hopefully, we're on your face. And we'll take it from there.

So when we look at the everything that goes into a pair of smart glasses, you kind of say, you know, how is this? How is this device similar to a smartphone? And on many levels, we'd posit that essentially every major component, except for the display module is essentially an organ transplant from a smartphone, the batteries, the processors. Now, of course, Qualcomm will tell you obviously they make a very specific processor for that. But by and large, it's a similar or same technology Pack. differently, batteries, sensors, processors, you know, where you need to make something that's fundamentally very different is when it comes to the display, because it's a whole different medium of how you provide the information, you're actually putting it right in front of the eye, you have to make sure it's transparent. And you have to make sure that it gives a good image, you have to now combat ambient light, you're dealing with daylight, so you have to have certain power efficiencies, you're now playing the game of you know, how much field of view can you jam in there. And of course, aesthetics, and aesthetics, you know, comes to form factor, it also comes to how much visibility, you know, you have of the optical elements, you know, whether you're using a diffractive, or a reflective waveguide, like we're using, you need to make sure that those are minimized, you need to minimize glare of the physical world, I mean, all these new problems of aesthetics come in that you know, didn't exist with smartphone screens. And then forward light leakage. That's another big one, most technologies are actually projecting outward to the person you're talking to, more or less what you're seeing. So, you know, in terms of social awkwardness, as well as privacy, they can see that so how you manage that light leakage, these are all critical factors that need to be tied into a system. And then you know, Last and certainly not least, these have to be affordable, it has to fit within a reasonable bomb for a consumer product. So when we look specifically at the optics modules, you have different technologies that that drive the projectors into these, but also that you have, you know, how is the image being delivered in front of the eye? You know, I think the industry has decided that it's waveguide. There are other bulkier optic solutions, you know, such as birdbath, as you see with companies like Enrile, which I think are effective at giving a good image, but I don't think they meet the market standard for something that looks like natural glasses, it doesn't enable the natural eye contact ability, that is a word we made up, but it makes sense. All those key factors have to go into an end product if you want the consumer to actually wear them daily. So now, when we break down, there's only a handful of companies doing waveguides it's us and a couple of startups, you know, various approaches, I'd say the the main camp that you know is the primary competitor is diffractive optics, we're on the reflective side, you can look up curl goo tag, he could explain it a lot better than I can. And I don't want to dwell too much on the other guys, ultimately, our pitch is that we have a much higher image quality, higher power efficiency, much better when it comes to the aesthetics, how we deal with the ambient light, how we minimize forward light leakage. And we have a very strong manufacturing supply chain and manufacturing process ready to scale already working with particular customers to scale.

One of the big differences, at least when it comes to the core technology from our previous conversation to today is that you went from having a one D, a single dimension kind of expansion of that image coming from the display and projecting into the eye. It's now a two dimensional expansion of that image. Can you explain a little bit what that two dimensional expansion is and why it was necessary for you to kind of add that that capability into the product offering?

Yes. So essentially, what we have with the waveguide is an image is projected into the edge of this, you know, for lack of better terms piece of glass, and the image travels to an area in front of the eye. In order to make sure that the image is large enough in front of the eye, we have what we call a one dimensional expander. So you would have a relatively large projector at the edge. So if I'm coming from the temporal side, I'd already have a large image on the y axis the vertical axis, and I would need the waveguide to expand the image on the x axis on the horizontal axis. That's a pretty simple straightforward approach. And that works very well for our military customers works very well for industrial applications, medical applications, these things are fielded we've sold 10s of 1000s of them. We have our manufacturing partner quanta, already making 1000s of these. However, when we talk about consumer consumer needs something that actually fits in a sleek pair of glasses, you know, a pair of Armani's, a pair of Ray Bans, it has to be, you know, small enough, elegant enough to essentially disappear in the background. And that's where we need a two axis expander. So just like we use the waveguide to expand the image in one dimension, now we're using a waveguide to expand the image in both dimensions. So you have a very small projector at the edge of the temple where the arm of the glasses meets the waveguide meets meets the lens in front of the eye, and that at that edge, very small entrance aperture as we call it, very small window, the image is injected inside and the waveguide starts to do all the magic of expanding expanded in the y axis expanded in the x axis. And then in front of the eye, you have this beautiful virtual Image seamlessly blended with your view of the physical world. Wallah. Easier said than done. Easier said than done. But we've been at it for a bunch of years, I'd say a key transition for us is when we realize that by hook or by crook, we got to make sure that we have something for consumer, you know, we kind of knew that we had a good couple years till the big companies start hitting the gas pedal, because we've always maintained those dialogues. And thank God, you know, when the markets starting to happen, and again, I say happening below the surface where the big bets are being made internally at these companies, we finally do have an offering that meets the key form factor requirements, and we're ready for game time.

Ready for game time. I love it. So when you you introduce this, this new generation of 2d expanders? Yep, right, this magic that happens inside the lens. And I remember the first, the 1d expander was already complicated, because a lot of fancy math that goes into making the image the resulting image that you see, look like a single, unified, perfectly well blended image. But the complication of adding a second dimension to that is not, it's not just twice as hard. I'm guessing it's an order of magnitude harder, both from a mathematical and a manufacturing perspective, to put all these things together all these little expanded pieces together into the lens and have it then result in a one beautiful, seamless image. Is that fair?

That is definitely fair. And you know, I would agree, we definitely had to walk before we start driving at 100 miles an hour. But in the last decade, we've made some major, major inroads in the technology, you know, even with the one D, all the complexities as as you mentioned, we made sure that we're able to get that to perfection. In the earlier days, there were imperfections in that a lot of techniques needed to be honed and figuring out exactly the ideal manufacturing process, what causes, you know, little, little noise, you know, if you will, in the image, root cause analysis of that, and then figuring out how to engineer that out of the manufacturing process. So, you know, way back when, when you first came across our waveguides, if it's about six or seven years ago, we had already come a long way in terms of working out the kinks of first generation, one dimensional expansion, that in and of itself, were major steps the team took upon themselves and really just went through a bug list and engineered everything out. Once that was perfected, essentially, we're able to start looking at the next horizon. Because ultimately, we knew in the long term, you need to access expansion to hit consumer, but you got to clean up the image first and in gen one, because otherwise, you're going to be multiplying your optical errors, once you start doing to access expansion. So once we got that cleaned up, again, same techniques employed, essentially, you know, I will, of course, oversimplify. But when you look at a 2d waveguide, it's essentially two waveguides. glued together, now we have come up with, you know, pretty smart ways of doing it in bulk. So you're not just taking, you know, final waveguides, and just slapping them together, there are smarter ways to do it on the stack level, which makes it a lot more efficient. One thing we will say and we'll reiterate just like with our first generation, we use standard processes, standard machinery, there's nothing new that's been invented in terms of the heavy equipment that's required. There obviously, is some tooling and customized jigs, in certain cases, you can even put robotics, just to make sure that when you put those waveguides together in the bulk stacks, you have the proper precision. But this is all in the engineering of bringing up a production line. So So there's, you know, there's multiple steps. But again, it's extremely repeatable once you've tooled up a line. So we're pretty confident, especially when we look forward and we keep saying consumer, consumer consumer, you can't do consumer if you're not going to be able to hit the price points. And you know, the the engineering and all the work that's gone into it, we've really made those inroads to be able to say, Okay, we're ready, we're starting to get those orders, because we're, we're ready with the process.

This is such a critical part of this industry. We've been, you know, collectively banging our heads against the wall for decades, decades. Yep. Trying to solve these really challenging riddles around how do you make a consumer grade optical system? How do you make a consumer grade display projector system? That makes sense. And a big part of the problem is beyond simply being able to demonstrate that it can be that something can be done in the lab is being able to have something that is manufacturable at scale, and at reasonable cost. And the the innovation that you're describing is really very much as important as the showing that it can be done once because we can do it once because $100,000 a copy. Well then maybe you have something that works for the military, but you don't have something that works for consumer sure if you have something that's great and is reasonably priced. That's right that's what enables the true breakthrough in terms of consumer adoption. Yep. You noted quanta is already manufacturing partner they've been making your your earlier gender ration technology for a while now. Yep. And I remember reading a little while back that that shot the German Glass Company is also now partner. Can you describe maybe some of them the advancements you've made specifically in the manufacturing process with these partners?

Absolutely. So quanta is actually a fantastic example of a company that doesn't have optics background, per se, I don't think they've ever done real glass processing. And essentially, you know, we put them on the map. And that's kind of a testament to the scalability of the technology, because it's more a function of buy the equipment, follow their procedures, and that's one thing they're fantastic at, and they've been able to get, you know, high 90% yields on their first generation. And they're also starting to get into the second generation, they're also getting into the 2d waveguides. Now, as we reach that, quote, unquote, extra level of complexity with the 2d waveguides, and we're building up a relationship with shot already a couple of years ago, we decided to work with them in lockstep to develop the process. We said, okay, these guys are the real experts in glass processing, how would you do it? Tell us what are the pain points in a manufacturing? What's going to drive cost down? How can we save costs, so let's let's work on the design together at an early stage. So the concept of design for manufacture was brought in, you know, very early so that that was that was the fun balance of getting the optics guys to just go wild come up with the best freakin technology. And then you know, have that constant dialogue with the with the supply chain and say, well, you're going to do that there's gonna be a cost driver, what else can we do? Okay, fine. Good to know. And then go back to, you know, run simulations that change this a bit, where can we ease up on tolerances? Where can we make things easier? And we're at a point now we're, again, still working through the specific generation. But I'd say we've already gone past proof of concept where the showstoppers have now been removed. So when I pontificate on a podcast and say, We're the only, you know, mass, manufacturable, waveguide, that can hit, you know, high 90% yields, I'd say, with a lot more confidence, I mean, the progress is already there. Now, the one caveat, I would add, is that for any new technology, and any new platform, it is not a, you know, low barrier to entry to set up a production line, you're spending 10s of millions of dollars on equipment, and making sure you're all tooled up. And that is also why, you know, you focus on the big market makers, the tier ones, the people who could actually push things and make those bets, asking smaller companies in between here footing the bill to get into mass production, that's going to be a little tougher of an ask. So you know, I always say it's not for the faint of heart or the week of wallet, you need the right type of backing to get into something like this. But I'd say the same thing was true for smartphone screens. Obviously getting those first fabs, how much do you think the fab cost to get the first touchscreens for smartphones? Not cheap, either? Ours is even cheaper? I'd argue

this notion that manufacturability is the is the current frontier, that I don't want to diminish the amazing amount of engineering that goes into making the 2d waveguide, the 2d expansion possible, yes, that by itself is very impressive. But the manufacturing at scale and reasonable cost. And then and then having the partners and the flexibility to be able to respond to the unknowns that are in front of us, even with these tier ones. As you noted, maybe it's 100,000 units, maybe it's a million units, or maybe it's starts at 100. Yeah, to go quickly to a million. Yep. And having a well defined set of partners and process to do that is essential, ultimately, in order to survive the early tumultuous periods. Yes. One of these challenges that you'd noted is ultimately how to make the cost reasonable. So what does that even look like? Like? How do we even think about from a consumer perspective, you talked about the bomb, the bill of materials, the cost of all the components that go into a pair of glasses, as we think about the full on bomb for a pair of glasses, how much of that come from a percentage perspective is the display an optics units or unit I guess, to behave as a monocular? binocular?

So we've used the analogy of the smartphone. And I think it's a very good baseline for understanding, you know, where your wiggle room is in terms of, you know, pricing where those boundaries are. And if you look at smartphone, there's huge variants, obviously, there's the super high end smartphone, and then there's the budget smartphones. And if you take a bomb breakdown, again, the exact percentage of what display is, I can go look those up later. But, you know, as recent as just, you know, probably about two years ago, the last time I looked at a proper teardown, I remember the low end smartphones or, you know, around the $40 range for the smartphone screen, I should say. And then on the higher end phones, you're coming north of $100, you know, 110 or something like that, you know, if you look at these iPhone tear downs, so that's a pretty big range, but then you could argue that there's, you know, high end versus low end, you know, in different function, I think more or less, that's the budget, you know, we're all going to be given For the total display stack, so display includes waveguide, projection optics, good news is projection optics are getting a lot smaller, so they could get a lot cheaper. That's where again, you kind of shift the burden to the waveguide to do the expansion, as we mentioned, so you can have very small projectors, and then the actual display panel, whether it's micro led, I'll cost laser, that all has to fit within that budget. So fact, you know, so you kind of work backwards from there. And you say, if you know, 110, is your is your budget for by an ocular. Okay, now everyone's got to find out where are they on their on that particular one. Now, I am giving you numbers of something that is already at scale and selling a billion units, that product category sells a billion units. So they're, quote, unquote, commoditized components at that point, there's going to be obviously a certain scale a period, I could argue that, you know, even the first several million will be north of that, maybe double that, again, all the components in there, but we do kind of see, like, let's say once things stabilize, after, you know, 5 million or so units that are out there, or 3 million, I don't know what the magic number is, you kind of hit those economies of scale, and then becomes a volume game, you know, the manufacturers are also taking lower margins, because they're getting better volume, et cetera, et cetera, you've worked out all the kinks in the lines, all that fun stuff. You know, and the interesting thing is, you know, when we talked to the supply chain, they're under a lot of pressure, you know, give us the million unit per year price. And, you know, they're putting out some pretty good aggressive numbers. But the question is, is, you know, is the OEM going to commit to that, you know, who's eating the cost? If they don't sell that, you know, so there's, there's that, that kind of back and forth, that's, that's going on, you know, behind closed doors, but they're good conversations to have the good news is, is put your money where your mouth is, you want to sell a couple million will hit consumer pricing? No problem. You only sell 100,000 It's gonna be expensive consumer. Yeah, yeah.

That's fascinating, in a really challenging and delicate dance at this stage of the market, given there's so much unknown. Yeah, and still some amount of risk, right, as we can continue to refine and perfect these supply chains, these manufacturing lines.

Yeah, I mean, part of the trick also is that we're seeing with the supply chain is, is if you make a bet, and you make a line that can handle multiple products, then you know, because the difference between one version of the waveguide versus another, you know, theoretically, a lot of those, you know, NRAS, or sunk costs of setting up a line can be recruited on multiple product lines. And that should also help, you know, especially if we stick to specific architectures, and it's just a little tooling change here or there. And that's the things that we're working through right now.

You've noted that a company to play in this space needs to be strong of wallet, which I thought was such a beautiful turn of phrase. But I don't see Lumus in the news much not announcing the next funding round, or the next funding round. How is it that the company is continuing to fund all of the r&d both in the product itself the design as well as in the manufacturing?

Excellent question, I asked myself this every day. I'm just kidding. We've actually had, you know, a lot of success. Again, a lot of our wins are behind closed doors, I'd say probably the last decade, over 100 million in revenues from customers, that varies between, you know, ongoing military business, we don't do any kind of advertising about our military business. We did do a raise, you know, I think you mentioned it, you know, several years ago. So we've had, you know, here and there some capital infusions, we're not looking for the glitz. And the glamour doing these, you know, big mega rounds, you know, hasn't really been necessary or, you know, to serve our purpose. We've more or less, you know, been plugging away and getting various customer projects, many times you'll get a big OEM or wants to do something very quiet. They'll pay an NRA, they'll do a custom pilot, you know, they'll do a bunch of things, try to figure out what works. Okay, come back, go to the you know, go back to the drawing board. But wouldn't you know, we've been pretty successful in being able to, relatively speaking, not do massive raises and keep the revenues flowing,

customer funded is the way to go. If you can pull it off. That's amazing. Always, always, something we kind of skipped over a little bit Was this some of the details about what is possible from like, A, A specs perspective, with the latest to second generation of the 2d expander? It? I think here, we just had a chance to put I had a chance to put eyes on it at CES here a couple of weeks back. And it is impressive to say the least the second generation device, it's really great. But what what are the numbers right now in terms of efficiency or field of view, or some of the other kinds of key numbers that people pay attention to for these for these systems?

Sure. So feel the view we're looking at right now, the flagship model is about 50 degrees. We obviously could do more than that. So we're discussing with different customers, much more than that, in fact, probably talking about up to 80 degrees while still coming in with a reasonable concern. Number form factor. So you know, those are different concepts were kicking around with different customers, it's something that we look forward to actually implementing in the coming years. In terms of, you know, efficiency. Ultimately, I think for outdoor use, you need to be at about 3000 nits some say a little less, some say a little more. That's, you know, within the sweet spot of the 50 degree waveguide, you go down a notch and field of view, you can actually increase the efficiency significantly, it's actually pretty linear. So we sell to the US Air Force, I think they're running it also somewhere about, you know, 7000 nits or something like that some something crazy, they're driving it beyond whatever we SPECT it out to, but hey, it's gonna go and good. No complaints, no returns? Yes.

That's awesome. And as you you have looked, and you've lived this history of Lumus now for, for a while, right through the early days of doing business development now for the last half a dozen years or so as the CEO of the company. And you've seen this evolution from being primarily focused on on military and enterprise and now really having an opportunity across a number of potential projects to be a major player on the consumer side as well. But as you project forward, maybe there's a couple of things I'd like to hear your thoughts on. One of them is, how do you describe who Lumus is five years from now,

as a company, we look at ourselves on the high level, we're a display company. Now, you know, that could be a lot more encompassing than what it currently is. Right now. It's near to AI display focused on mainly our baseline technology, reflective waveguide. That's our key differentiator. That's what we do better than everyone and you know, focus on your best outsource the rest, as they say, that's something that's, you know, for the near future, definitely going to be a driving force. But in terms of broader vision, as we fast forward, and we start to see, you know, the big market happen. And that could also bring in a lot more opportunities to expand the company, let's say fast forward three years, four years, this becomes mass consumer, you have 10s of millions of consumers using it, I see us going way beyond that. I mean, our technology, we haven't even scratched the surface of what we could do in terms of automotive HUD, using waveguide actually, as a projector, and saving literally 10s of liters of space inside a dashboard and be able to be available on pretty much every car. You know, right now, the heads up displays are relegated to bigger cars that have more real estate for that, that's a key pain point for the automotive. And that's something where we just haven't, you know, had the opportunity to focus on because, you know, we're getting pulled from all sides, just for consumer near to I display. But there's a lot more to do there. Even in military, there's a lot more to expand in terms of ground soldier, more we can be doing in the aviation side of things. But if we look beyond that, I mean, you know, I remember we had a particular customer come to us, and they had a pretty large ask, and I remember one of the top engineers came to me and said, Are we limited to reflective waveguide technology, and my response to him was absolutely not remember, we're a display company, you find a better, smarter way to do it, bring it to the table, let's do it, you know, look at the bigger picture. So we do encourage, you know, people on the team to do that. And we have some pretty crazy advanced r&d stuff that I should be careful what I discussed, just because as we work through these concepts, there's always more IP, you want to make sure you're covered before you share with the world. But, you know, larger vision of the company is to, you know, really, really expand on this beachhead of what we're doing and display and, you know, become a much bigger display company.

Amazing. So as you kind of look out, just maybe the next couple of years, you talked about maybe late 2024, early to mid 2025, we might start to see this next batch of consumer grade display snacking through the simpler, maybe sort of products, AR glasses that are coming to market. What do you think if it's not display, what do you think is the biggest hurdle the biggest, longest pole in the tent, so to speak, of achieving a generally available product?

I think UX is tricky. Making sure it's seamless, because the first thing a consumer is going to ask is, well, what the hell value am I going to get from this, if I could just as easily pull out my smartphone and do this, it has to be that much better. It has to be that more intuitive, that's simpler, it has to be you know, so I think that comes both on application and the interface the the interface really has to be done seamlessly. There are other challenges that go into a system power management, heat management, you know, maybe the lighter weight ones probably heats less of an issue because there's a lot less processing and if it's one display versus two displays, this is all true. Wait, I can tell you on the waveguide side where you know, you fight for every gram. You know, on the projector side, you fight for every gram, the designers, the industrial designers, the mechanical product designers, they're they're fighting you tooth and nail on every little bit. Why does this have to be there? Okay, what do we gain if we, you know, what happens if I just shave this part of the waveguide because, you know, it's got to fit into the, our aesthetic needs. And, you know, those are the things that they wrestle with, you know, like I said earlier, I think at this point, less of the burden is on us. I think we've kind of called the industry's bluff. And, you know, they said, here's a tough spec, and we're like, boom, here you go. So, we'll see. I mean, these are complex products. And that's what I'm saying, you know, even though targets are for end of 2024, they're gonna they're gonna slip, not because because all the other stuff,

because you've kind of studied this market now and live this market so intimately for the last many years. How has your perspective on this market on the AR glasses opportunity? Or the challenges or the players? How's your perspective changed in the last couple of years?

That is a good question. I mean, I think for everyone, it's taken longer than everyone anticipated. I think a lot of the where people thought the pain points were have shifted elsewhere. I think for many years, companies, were seeing the display being the big excuse not to move the program's forward and kind of stuck in r&d purgatory until a proper display came along. And you know, already the prototypes we've been showing the last two years have have really started, like I said, calling the customers bluff, and really put a fire under a lot of teams to say, oh, oh, we have to solve those other issues. So, you know, for all these years, we kept thinking we're the biggest bottleneck, and you realize that they don't have their swears together on a lot of other things. So now, there's that panic to make sure that the electronics are up to snuff. Like I said, the interface, the applications, there's a big blitz right now. So I'd say I was a little surprised to see how, how not ready, they were with the rest of the system?

That's really interesting. Yeah, you'd mentioned one of these other types of technologies, which is the birdbaths sort of display. And birdbath. And reel is just one example. There are a handful of companies that are utilizing, if we go back in time, it was, you know, Lenovo today is utilizing it odg used to utilize it. And there's you can deliver a really great image to the eye using that sort of technology. But as you noted, there's been some struggle on the the form factor that results from the sort of glasses. Is that what you think is holding back that sort of design? Or is there something more that's holding back that type of product from broader acceptance?

Well, definitely, from the form factor in and of itself, the aesthetic factors that come into it, I definitely say that that's a non starter, because I could see how much torture we go through on every, on every aesthetic aspect of our waveguides. When we deal with the designers, the product designers and the industrial designers of these products. So there, there's no way I can imagine they actually consider using these for their product just for the sheer aesthetics, they want to go for the natural look, they want it to be seamless to the consumer, they want to make sure that it appears natural, and nobody could tell you're wearing smart glasses unless they're right in your face. That's something that I think you know, right then and there, that type of technology. Beyond that breaking down to the you know, the specific standalone companies trying to offer smart glasses, whether it's diffractive, or whether it's it's bird bath, I think that is a very difficult task to take on. I think it is really a major ecosystem play. It's a os play. There's just so many so many things that have to converge to an AR product. There's hardware, there's software, there's you know, user interface, even within hardware, there's the various sensor technology, battery technology, silicon technology, how that all ties together, you're talking about, you know, 1000s of employees that actually need to go and make a proper product, in most cases. Now, again, you could probably do it with a few 100, if managed properly, even so, the top people in their game will cost you, you know, a few 100 million a year just to put something together. And then you have to bring the apps, the content, the developers, it's you know, from everything we've seen, and I've been in the industry a very long time, we have supported many startup with a few exceptions of companies just going for specific niches like in, in medical, pretty much, you know, I could show you a whole graveyard full of companies that have tried to do stand alone. And you have to ask yourself, I need these companies doing standalone. What are they going to do tomorrow? When tier 1x put something on the market? What do they have to offer? They don't have the ecosystem. They don't have the developers they don't have the content place. Where are they? So you know, it's on one hand, it's great. They're getting things out there. They're creating buzz, you know, is that a long term viable strategy? Not my bet,

not your bet. One of the other challenges that exist is one around the variability of the human body, both in terms of the size of the eyes and faces and all the rest sizes, noses and all those things as well as the ears. There's any hair Yeah, How far back and how long arms need to be, as well as their ability to see without correct division? Can you describe a little bit that the extra level of challenge that comes into accommodating the human physiology and our ability to see clearly?

Yes, so one of the things we get early on in the specification is this massive eye box. And you're like, why so big the, you know, each, each pupil only needs a couple of millimeters. But you're like, you have to factor in the varying pupillary distance of the vast population. So you so you get these big challenges for massive iBox. That just throws the burden back on the optics, guys. Again, it's something we've been dealing with for a long time, it's something we said Can't you just mechanically adjust or do multiple sizes. In any event, you're probably talking about for mass consumer, just because you know, even when you go to a glasses, store, men, women, Asian, Caucasian, large head, small head, even within those subgroups, you have so many different sizes, there's gonna be multiple SKUs of glasses for that. With that, we probably have to make sure that we still have a very wide IPD. But on the frame side, there's challenges that it say the you know, the tier ones or the OEM has to deal with. And we have to support that that, you know, making sure there's a couple of skews that fit into those fit into those glasses frames. Beyond that, think you mentioned vision correction, that's a big one, you have a huge swath of the population, myself included, that require optical correction, if you need glasses, what's the most elegant way to put that in your smart glasses, one thing that we're very proud of, especially with our latest architecture, as far as I know, we're the only waveguide technology where you can directly bond a prescription on the waveguide itself. That's a key differentiator, because that also leads to huge aesthetic advantages, you don't have that Coke bottle, look, you don't have to have that air gap. Whereas other waveguides, in order not to break the T ir, the total internal reflection, nor not to disturb the image, they need to have an air gap that could create additional reflections and also makes the stack a bit bigger.

So in your case, you can just carve out the bit of corrective optics you need and glue it directly to the lens. Easy peasy.

Yep. And beyond that, you could add other functions on let's say, the external lens, you go outside, that could be transition lenses, it could be, you know, photochromic, or could actually be electrochromic. That's something that's, you know, very compelling. If you have an ambient light sensor, and you want to manage the power budget better, or, Hey, you're outside, you anyways, need a pair of sunglasses, and you want to make sure your image is visible, you could turn them into sunglasses. And with electrochromic, you could do it with pretty quickly, you can make that transition going indoors to outdoors pretty quickly. And transition lenses are pretty competitive, but not as good as what you could do with electrochromic. But, again, key feature we enable that

phenomenal. Let's wrap with a few and lightning round questions here. What commonly held belief about this area AR glasses, or spatial computing more generally? Do you disagree with,

you know, a lot of the headlines and the clickbait says this is replacing your smartphone? I believe eventually, Yes, tomorrow, the next four or five years? No, I think that's, you know, there's going to be an easing into it. There's a lot of things that need to be worked out. Like I said, we still need to know what's going on what applications are going to work best, what people like about it, what people don't like about it, I anticipate it's gonna be very cool and amazing at the beginning, but people will get things wrong. And we'll have to be, you know, those next iterations will, you know, take those learnings and make them better. So, you know, immediately throw out your smartphone. I don't think that's happening so fast,

right? Besides the one that you're building, what tool or service do you wish existed in the AR market?

I'd say standardization of specifications would be fantastic when you go to different companies, you know, 16 by nine, four by three, you know, just aspect ratio resolutions. These are things that are changing very much customer to customer, do you want to full square, you want it longer this way that way, I mean, we keep getting this variance. And when you try to also, you know, most of our work right now is really customization. But it's nice to be able to have products that we put in the hands of let's say, our ODM partner quanta and say this is the plain vanilla this is for, we'll call it the tier twos, the smartphone manufacturers who aren't necessarily going to make a custom system but really want to hit the ground running, the more I'd say, you know, standardization we have, the easier would be to start anticipating and making something that you know, it's going to be plug and play, it's going to work with those OSS etc. So that would be a great thing. I'd like to say,

I would have to imagine that it would help with manufacturing costs as well. Yeah,

absolutely. Yeah. And also picking displays and telling you display manufacturers. Here's the spec of the industry be ready, you know, yeah, right now you go to Display manufacturers and pretty much everything is a customization as well because everybody wants something different.

That I guess works fine for now. And maybe with an X scale it'd be it'd be fine, but the in between part is really difficult. To be able to do a bunch of different skews like that? Yeah. Yep. Kind of shifting gears a little bit in terms of like information absorption. But what book have you read recently that you found to be deeply insightful or profound?

Most of my reading time actually is dedicated to my ongoing learning of the Babylonian Talmud. So that's actually that's, that's a massive body of work, where basically takes the, you know, the biblical laws from the Torah, the Old Testament, and gives you the breakdown of the entire legal process, you know, whether it's a case law and all the legal arguments that that goes into why we do what we do, you know, it's something that I have, you know, already reviewed, it's a multi year cycle to actually finish that massive body of work. It's about, you know, on average, seven and a half years, I, you know, do it a little faster, six to seven years, I'm on my third cycle as well, I'd love to say I remember everything, but it's difficult what I gained from it is, I mean, one of the things, I get a lot of things from it, one of the things I do feel that's actually fantastic and applies to, you know, how I get my work done here, is going through these, you know, legal arguments going through the legal cases, it really sharpens your brain. So, and that very much translates to when I'm sitting down and working on a very long detailed contract for licensing deals, that skill set very much translates. So it's probably less about the information gathering versus the sharpening of the brain. That's something that I really like to do in my spare time, and something I do pretty consistently.

That's pretty amazing. Yeah, if you could sit down and have coffee with your 25 year old self, what advice would you share with 25 year old dairy hoof

patients, patients, especially, you know, going back 25 years that make me you know, relatively younger person ready to conquer the world on a, you know, what I thought was a very fast track and, you know, retrospectively was a pretty fast track, I would, you know, again, advise and to any young person, that every step is a key stepping stone, you know, and every accomplishment is just the beginning. And every failure is a major learning experience. That's something that I think is held true. You know, I see it even in my time at Lumus, every experience here has been a major stepping stone. And I think our patience has really paid off. I think it's it's positioned us very well, to really, you know, like I said, get ready for game time. But I think in general with with everything in life, the other advice I'd say is, you know, family, family, family, you know, spend as much time with the wife and kids as possible, you know, we're always busy and wrapped up in this professional world, and, you know, pushing very hard and pushing the limits. And it goes by very fast, kids grow up much, much faster than you think. Now, at this point, I'm a grandfather of two already and kind of trying to figure out how that happens.

That's amazing. That's incredible. Yeah. Any closing thoughts you'd like to share?

I'd like to use this platform to really, you know, give kudos in general to my team, I think many times, you know, I get the opportunity to be the face of this company, the frontman for the band, but without such an awesome team, you know, with a really strong shared vision, we wouldn't be where we are. And, you know, it's it's a very small company relative to the rest of the market. You know, we're, we're being outspent and we're outsized. In terms of manpower, you got 1000s of people, top level engineers in the fields, and really, you know, top quality guys, 1000s of engineers, with directly competing technology, and billions of dollars going against our budgets, and our team, you know, time and again, our iterates and outperforms, you know, when I took the helm couple years ago, six years ago, a good friend of mine, Josh wine, he had worked at McKinsey a bunch of years, then ran a bunch of companies. He actually told me, it's like, you become CEO, Ah, he's like, just realize, and I'll spare you the imitation of the his British accent, he's just realized 80%, of being a good CEO, is having a great team. And, you know, I'll challenge that and say, no, 95% of being a good CEO is having a good team. So have an amazing team. And again, we are where we are and the opportunities right in front of us. And we're going to keep kicking butt and taking names because we have good group of people here. Well said,

Where can people go to learn more about you and the efforts there at Lumus and the rest of the team?

lumusvision.com. Lumus Vision on Twitter. And you can follow me Ari Grobman on LinkedIn.

Amazing. Ari, thank you so much for the conversation.

Thank you.

Before you go, I'm going to tell you about the next episode. In it I speak with Cayden Pierce, Cayden describes himself as a transhumanist hacker working to enhance our intelligence using AI smart glasses and eventually neurotech Kaden posts regularly about the current state of smart glasses and is actively developing open source smart glasses, hardware and middleware solutions. He's also working on a contextual search engine to help deliver meaningful utility to the smart glasses of the future. I think you'll really enjoy the conversation and please consider contributing to this podcast@patreon.com/theARshow. Thanks for listening.