Data & Dimensions | Ep. 15 - Smart Glasses Startup Challenges Industry Orthodoxy on Privacy and Power

View Full Transcript

Thomas Suarez   00:00:00

I think people assume that if they can see the camera, the camera can see them. And they're totally right. And within Silicon Valley and San Francisco, I think it's easy to say, yes, okay, people will be okay with this. But people are generally not okay with this sort of thing. And I'm not convinced that people will ever be okay with a third eye camera.

Neil Barbour   00:00:34

Welcome to Data Dimensions, a podcast at the intersection of entertainment industry and interactivity. I'm Neil Barbour, an analyst with S&P Global Market Intelligence Kagan. So last month, I was on the floor at Augmented World Expo 2026 in Long Beach, and it was all about smart glasses. Snap, XReal, Google, Meta and others were all jockeying for mind share and floor space. Our forecast puts smart glasses shipments at more than 23 million units annually by 2030, growing at a compound annual growth rate of nearly 22%, solid momentum, but there's also some friction built in, cost, camera anxiety, fashion, battery life. The first company that solves those problems for everyday wearers is best positioned to be the next big story in wearables. One device from AWE that stuck with me was one you might not have seen in the headlines. Our report called Raven Prism, the sleeper hit of the show. It's a stand-alone pair of smart glasses that runs its own Linux-based operating system, features hot swappable modular battery packs, a physical camera cover and is designed to be from the ground up, an all-day ambient computing platform. The company behind it is Raven Resonance, founded by Tom Suarez, and we have him here today on the show. Tom, welcome.

Thomas Suarez   00:01:48

Thanks, Neil. It's great to be here.

Neil Barbour   00:01:50

Tom, the Raven Prism is described as an ambient computer in the form of glasses, Linux underneath, real apps on top, built for technical creatives, people who make things. That's a different pitch than what Meta and Snap are making. Kind of spool that idea out for me. What are you guys aiming for here?

Thomas Suarez   00:02:07

So we really see kind of 2 main categories of use case, one being micro interactions, which is things that you -- or things like notifications or driving directions. This is something where you're in and out of the display within 5 to 10 seconds or less. And that has been somewhat covered by other devices, although there are, of course, input concerns, which we'll get into in a minute. Then the other side of this is kind of reference material. And we think that this is where Raven Prism really shines. -- because it's based on Linux, it has an Arm 64 system. It's a quadcore system, super open, you can actually start putting up more applications that are really sophisticated. And the reason we're going after technical creatives at first is that this is a -- I mean, this is kind of who we are. We're kind of in a way, building for ourselves. But there's a lot of instances where you would want to have -- whether you're a developer or an engineer or a musician or a filmmaker, there are areas where you want to have your hands free to do -- to hold things or to like hold a musical instrument or touch like a physical object while also having a lot of really helpful contextual information. And you want to be able to control that application hands-free and ideally silently, which is where the eye control comes in. But yes, building an ambient computer, the idea of ambient computing is that you have technology that's there when you need it and gone out of the way when you don't. And that is the entire design philosophy behind making it an ambient computer as opposed to more traditional smart glasses.

Neil Barbour   00:03:42

Yes. I guess what I think of when I saw the Snap Specs, which also kind of have that all-in-one profile, maybe have a more robust spatial computing applications than where you're pitching right now, but they really look like swim goggles or something. They're really ostentatious from a fashion perspective, whereas you guys -- at least the prototype you guys are showing like you wouldn't necessarily know they were smart glasses first. So I think you're sort of bridging those 2 worlds is an interesting dynamic. So I want to walk through some of these big design choices we were talking about. First, the swappable batteries. It seems like such an obvious solution, but talk about what's going on there and maybe why some companies haven't landed on this yet.

Thomas Suarez   00:04:22

This might be one of those things where it's obvious in hindsight. A lot of companies have been trying to optimize thermals and battery life, of course, and just the amount of power that a CPU is drawing from the system or display or whatever. And we, of course, do that as well. But we have -- there's a certain limit to the battery chemistries right now. And so we started the project very early on actually more in the modular or like with modularity in mind. So there's actually power and data running through the Raven wings, which are the hot-swap batteries on the back. So we're actually going to launch a hardware development kit later this year. We have an SDK out already for the software, but we're going to launch an HDK because we want people to be able to create other types of modules like rearview cameras and cellular modems, GPS modules like that. So you can actually mix and match the types of modules that are on the back. It's not just swappable batteries, although the swappable batteries do help a lot with all the power. We have thought about the power system more holistically as an ecosystem of power, both power sinks and power sources. And the entire power system actually is quite complex in a device like this because you have to handle -- most systems are not happy with just getting their power taken out randomly. Yes, we have to account for that, and there's things in the Linux kernel and application level or application layer tweaks that need to happen even down to device drivers around WiFi and Bluetooth, how do you bring those down gracefully and bring them back up in a state that is deterministically correct from where the user left off. So the cool thing is that on Raven Prism, if you take off both of the wings, the DRAM actually stays clocked. There's a smaller internal cell that keeps that processor or the DRAM running so that you don't lose applications. But of course, there's a lot of other interesting challenges there.

Neil Barbour   00:06:25

Yes. I guess I bring it up because even the Meta Ray-Bans, I think, can go for 6 hours, 8 hours, something like that, and then have to go back in the case or something. And it really cuts down on the idea of an all-day wearable if you have to take off your glasses and charge them. And I think to your point, what they're trying to solve for is just efficiency and better battery life in general, whereas you're kind of coming at it from this unique perspective that maybe they're hoping they can just leapfrog this idea and go straight to just having a battery that runs 12 hours, you put it -- you only charge it when you sleep.

Thomas Suarez   00:06:59

Yes. There's also, I guess, maybe a more cynical answer is it's easier to sell people a new device when their battery dies. -- from a right to repair and kind of replacement standpoint, I think it's important that people can upgrade their batteries. And of course, I mean, there's many generations to come of this technology and Raven in particular, we have a pretty cool road map over the next 10 years or so. But that being said, once someone buys a device, they should be able to use it for a considerable period of time.

Neil Barbour   00:07:36

All right. So let's double-click on the front-facing camera thing, another one that might seem obvious in hindsight. You seem to be the only vendor out there that kind of admits that it's a social issue. I mean, I talked with this -- some of the other guys on the show floor and they kind of hand wave it away, this idea that if you're wearing smart glasses with cameras on the front, everyone you talk to has a camera pointed at them and maybe they're not cool with that. How did you guys come to the decision to just cover those up? And how do you envision people using that in practice? When do you put them on and take them off?

Thomas Suarez   00:08:06

Yes. So this is -- the camera cover is one aspect of it. We also have something called Beacon lights, which are these lights that illuminate the rim on the final generation of Raven Prism One, there's actually kind of a clear rim that lights up when the camera is in use in addition to having that physical cover. And then there's a lot of kind of security and privacy architecture in the device that we had to do from day 1 because it's not really something you can slap on at the end. It really has to be embedded into the architecture. But it really comes from -- many of us have worn these devices for a long time, and I wore Google Glass for 12 years and in all different areas across many different states in the U.S. and Europe, like all different places. And it was very interesting to see the reactions from people. And I wasn't wearing it for reactions. I was wearing it because it was a useful device. But it was really interesting. There was actually one piece of etiquette that's self-developed within the glass Explorer community around 2013 or 2014, which is that if you enter a public space where it's inappropriate to have a camera, for example, a public restroom or something like that, because Google Glass Explorer Edition couldn't fold, like there's no hinge mechanism, you would actually take it off of your face and put it around your neck with the camera facing upward into your back. And so this was a signal that, hey, I'm not recording you, you don't have to worry. But of course, there's -- I think, for example, Meta has the recording light, the LED next to the camera. And we explored LEDs and we explored shutters and all sorts of things. The problem with an LED though is that it's usually not noticeable. And it's driven by software at the end of the day, which means it's very easy. You could -- there's -- we've heard of services where they're opening up Meta Ray-Bans and just...

Neil Barbour   00:09:53

Just disabling it

Thomas Suarez   00:09:54

Taking the LED out or cutting a trace on the PCB. I mean there's a lot of ways to get around it. And I think people assume that if they can see the camera, the camera can see them. And they're totally right. And within Silicon Valley and San Francisco, I think it's easy to say, yes, okay, people will be okay with this. But people are generally not okay with this sort of thing. And I'm not convinced that people will ever be okay with a kind of third eye effect with a camera as they are with microphones, for example. Microphones are a little bit more invisible. We know that they're around. They're a little bit -- it's still really important to make sure people are aware when there's a microphone around. But with the camera specifically, there's this third eye kind of potentially very primal fear of having this extra thing that's staring at you. And there's actually a real -- like it's kept me up at night in the past of what are the downstream effects. It's not necessarily can this company collect this data? Some people would say, I don't really care if this company knows about what I do in blah, blah, blah. But the thing is -- the reality is that glasses are unique in that they will be able to see and hear everything that you see and hear in every aspect of your life. And there's a tremendous amount of responsibility that it's not just on the user. It's on the company. So we have put a lot of effort into that. And we've actually even had debates internally of what is the right answer. Should we remove the camera entirely? There was a version of Raven Prism in development that was -- we were not developing it with a camera. And -- but we felt that the camera was still very necessary for a lot of use cases, but it just has to be done in a responsible way.

Neil Barbour   00:11:46

Yes. I agree. At one time, I was also thinking when this market continues to mature, more and more vendors are going to just decide to not ship with the camera early on. But then I think AI became a more crucial component to the center of the stack and the use cases. And if it can't have that contextual input, then what is the real upside of having the smart glasses on. And so I guess the second part of my question there is like when do you put the covers on when do you take them off? So like obviously, you would leave them off if you were doing something like if you were baking at home and you needed to see what step in the process you were if you were doing home DIY repair, then they stay on. But maybe as soon as you get in the car or leave your front door, you put them on and leave them on all day. Is that kind of what you're envisioning? Or is it more complicated than that?

Thomas Suarez   00:12:33

Yes. I think it's more of just a gesture to people around you that I could be recording, not necessarily that I am, but I could be and maybe the camera is there for contextual information or whatever. But it's really helpful. If you have a face-to-face conversation with someone in a private setting or something like that, you can just pop it on and it's clear that there's nothing recording. The other -- I mentioned shutter briefly. Shutters, we considered electrical and mechanical shutters. And shutters can actually work pretty well for this. But the problem is any photographer knows that like a DSLR shutter, that could be 1/2,000 of a second, and there's no way someone is going to notice. I mean you can't even see that. So I might be able to hear it, but you can't really see it. So Yes. Part of this is the gesture you're making in the public space, not the actual functionality, although it is downstream of that and upstream. Yes.

Neil Barbour

So at the show, I noticed you guys were using LCOS, which is liquid crystal on silicon, which came as a bit of a surprise to me, although maybe I'm a little behind on this thinking. I've been hearing about how great micro LED was for smart glasses, but you kind of set me straight on that. Walk us through some of the trade-offs between LCOS and microLED and other display technologies in smart glasses.

Thomas Suarez  

Yes. So I mean, kind of like any aspect of the sort of device, there's no right answer. It's really dependent on the use cases. And for our use cases, we have a mixture, as I mentioned, of micro interactions, which is 5 to 10 seconds or less than the display and reference material, where you actually need to have the display on for a considerable period of time. With micro LED, microLED is actually really good and very power efficient for certain use cases. So if you have like a small number of pixels illuminated on the display, for example, if you're getting a notification and you have maybe 5% to 10% of the pixels on the display just showing text. That is a really good use of microLED because you can get these really high contrast and also high brightness outputs from the microLED system, and you're only turning on the LEDs at the locations where it's necessary. MicroLED does not do as well when you are trying to illuminate a lot of pixels in this form factor because microLED, we measure this in terms of what's called APL, which is an average picture level or average pixel level. And the APL scales linearly with respect to power for a microLED system, whereas it's actually a bit more logarithmic on an LCOS system. And what that means is that around maybe 10% to 15% of the pixels -- or 10% to 15% APL, which could be 10% of the pixels illuminated at 100% brightness or it could be 100% of the pixels illuminated at 10% brightness. That's kind of how it works. At that level, microLED actually is really, really good. But when you're trying to illuminate a lot of pixels, you want to have a sophisticated application on there or you want to show graphics or video for more than a few seconds, that is going to turn a microLED system into a furnace. And you're not going to be able to get the heat dissipated fast enough before the thing just thermally shuts down. So we -- or burns the user, which is even worse. So we opted for the LCOS because there are trade-offs, like there's a little bit lower contrast. The brightness is slightly less, although pretty comparable in some cases. But what you get is this very stable system where you can actually draw to many, many of the pixels at the same time because you can actually push a lot more -- you really just have a kind of an illumination source that is independent of how many pixels are on, which is a bit different in LCOS than it is in a microLED.

Neil Barbour   00:16:27

You've worked around some of the really big names in consumer tech or other people in the team as well, Apple, Samsung, Microsoft, XREAL. Without throwing anyone under the bus here, what do you think the current crop of smart glasses gets right? And where do you think the industry is making some avoidable mistakes?

Thomas Suarez   00:16:45

Yes. So this also is dependent on use case. I think for all-day wear devices, it has to be lightweight and it has to be comfortable, but it also has to have a lot of utility. And if it's not -- if you're going to make something with a, let's say, microcontroller system, there usually has to be something -- some very clear specific use case that, that microcontroller could run because it's not going to have the same level of compute that you'd expect from a smartphone chip or something like that, closer to something that Raven Prism has. So that's one thing. But also, I think there's a bit of an over-index, in my opinion, on the total weight. I think more of the weight that we need to be worried about is the nose weight, no more than 25 to 30 grams in the nose. The rest of it is actually a little bit more -- you have a little bit more leeway, and that's where for us, we're close to 70 grams or so total, but the nose weight is actually very well balanced. So you don't really feel it. And it allows us to put a lot more compute and sensors and things in the device.

Neil Barbour   00:17:53

And that's because of the wings, you get to put the wings on the frames back toward the ear instead of the front. So it's distributing some of that battery weight. Exactly. Is that kind of what you're saying.

Thomas Suarez   00:18:01

Exactly. So you actually get a lot of balance there. But also, I mean, I think there have been some really good examples of devices that have done certain things well. Like actually, for what it's worth, with all of its challenges, Google Glass actually did a few things really well. The comfort was there. They had this titanium band that was one like really felt great. It was wonderful to wear. It was not obviously not wonderful for other people to see. But it's the best they could really do in that form factor given the era in 2012. That had more compute than most of the microcontroller glasses coming out now. And it actually ran a TI OMAP 4430, which was a smartphone chip of the era 32-bit Arm. And so that's one thing. But I think the areas where a lot of the devices are struggling is, a, certainly in utility; b, in power management, of course; and c, in the privacy aspects where there is -- and actually in input as well. So having to touch a capacitive touch pad or something like that, this is...

Neil Barbour   00:19:13

I can never get it to work.

Thomas Suarez   00:19:15

This is a very strange gesture to bring your finger up to your temple. I did it for years, and it's tough socially. But also...

Neil Barbour   00:19:25

What is that too, yes.

Thomas Suarez   00:19:27

Yes, if there's any sweat or rain or anything like that, your capacitive touch stops working. So it's all of these -- or for example, there are some devices that actually put the capacitive touch pad behind the ear, which for people with long hair does not work at all. And even when I put it on some of these devices and they have the touch pad behind the ear, it's useless. It's actually -- because I can't get my finger to the touchpad. So there -- it's really important to test -- to test these devices constantly during development, which is why dogfooding is so important, but also to test it with all different hair styles and different demographics.

Neil Barbour   00:20:09

So what have you guys done on the gesture controls or input mechanisms? The Meta Ray-Ban Display has that wrist… I'm not sure how to describe it, but can sort of read your gestures as you move your hand around and can tell what you're doing. And it works to some degree. But are you guys going down that road? Or how are you solving for some of that?

Thomas Suarez   00:20:28

Yes. So with the technical creative target, we wanted something that was hands-free and silent. And that significantly reduces the number of options that we have. So you can do head gestures, for example, which depending on the social situation could also be tough, but there's different ways to do it. Voice is definitely another way in areas where it can be hands-free, but it's okay to be -- to actually speak the commands. But we wanted something that from a silent standpoint, that's also ties into privacy, if you're on a subway or something, you don't want to have to be broadcasting what you're doing to everyone around you. So just like on a phone, right? If you're looking down to your phone and you're typing no one around you unless they're stalking you and looking directly down to your phone, they can't really know what you're doing. But with voice, a lot of these voice interactions, it's more difficult. And a lot of us on the team have been reluctant to actually use these voice interfaces because we're dictating our messages not only to the device, but to everyone around us that we don't know. So there's that. So we've actually landed on eye control as the primary input. Voice is actually secondary. So -- and then we have a tertiary input, which is buttons, but that's really more of a backup system. EyeControl is very important to us because it's silent and it's hands-free. And there are certainly trade-offs with it. It's very difficult to implement it in a way that's not fatiguing to the eyes and selection is also -- that becomes a challenge. On the Vision Pro, for example, Apple actually did a kind of look and pinch. So you look at what you want to select and then they are downward-facing cameras and it can tell when you're -- it's tracking your hands. We don't have that because we don't want to put a bunch of extra cameras and do a bunch of extra compute on the device. But we also didn't want something extra that you need to charge and take with you. It shouldn't feel like suiting up or armoring up in terms of to actually put on your device and start going.

Neil Barbour   00:22:36

Yes. I will say as an aside to that, I have demoed the Meta Ray-Ban Display multiple times, not necessarily with Meta, but at different places. And more than a few times, they have said, we can't find the wrist band or we forgot to bring the wristband, which is indicative of kind of the challenge there.

Thomas Suarez   00:22:52

Right, exactly. And the wrist band is actually really cool, like the technology.

Neil Barbour   00:22:56

When it works, it works.

Thomas Suarez   00:22:57

Yes. When it works, it's actually really cool. And -- but there's another kind of set of challenges, which is that when you -- and this applies to all sorts of inputs. This is not just the neural band. But when you have an input method that relies upon a lot of training data in order to get right and you need very high population coverage, you tend to end up with slightly less accuracy. And on a keyboard, it's almost 100% accurate. You hit the key and as long as you didn't miss the key, as long as it's not user error, it's going to enter the key, the letter that you selected or the character you selected. It's a similar thing on a touchscreen. Imagine if every -- 1 out of 10 times or 1 out of 20 times even that you're typing a message and you hit the letter S and it actually hits the letter K on the other side of the keyboard. That will be frustrating. And honestly, it might even be unusable. And this is a challenge that a lot of input methods have, and we've seen it with some of the newer ones where like send and cancel are very similar gestures. And if the device doesn't register the difference, 1 out of 20 times, people stop trusting it. So it's a really hard problem even in eye control. There's a lot of challenges in this. But I think that by the time we launch and people get to try Prism and try using the eye interface, it's going to be like, oh, why isn't everyone doing this?

Neil Barbour   00:24:32

How are you guys doing eye control without inward-facing cameras?

Thomas Suarez   00:24:35

So there are actually inward-facing cameras for the eyes. There's one sensor...

Neil Barbour   00:24:40

There's not downward facing for the hands. -- got it. Okay. Let's talk about your investment story, if you can, for a little bit. I see Draper Associates, Ecosystem Ventures, IAG Capital. What does that capital journey look like? And are you still looking for funding?

Thomas Suarez   00:24:56

Yes. So raising for this type of company is interesting because it's kind of at the intersection of a lot of relatively unpopular areas, right, augmented reality, consumer hardware. So there's -- but the cool thing is that we've actually -- we've been able to partner with some really great investors who have been awesome and some of the names that you mentioned and some others as well and who are also on board with the mission of the company in terms of new computing interfaces and privacy and these sorts of things. And the -- I would say it's a relatively -- relative to the kind of the hot areas of investing typically.

Neil Barbour   00:25:44

You mean AI... You have an AI adjacent story.

Thomas Suarez   00:25:47

Absolutely. I mean there's a lot of physical AI on the product. We're making all sorts of technology that are around there. But there -- I think from an investment standpoint, there have been -- it's kind of more in a niche -- and we found investors who are really on board with this company over the next 10-plus years. So we're pretty excited about that. Yes, I mean, we're always open to investment. We are entering the next era of the company right now. So it's a pretty exciting time. But yes, we have...

Neil Barbour   00:26:22

Yes, speaking of which, do you guys plan to take preorders this year? Are you shipping product this year? What's that road map look like?

Thomas Suarez   00:26:29

Yes. So we're actually -- we're shipping this year. We are doing a full launch later this year. Our preview at AWE was really just kind of an announcement and with the full launch later this year. But yes, we will have places where people can come and try the device on for themselves. And then the developer -- we also have a place like a physical place for developers to come and experiment with early prototypes and have direct access to our team. So all of that is coming online later this year.

Neil Barbour   00:27:00

Great. So I guess last question, it's a few years from now, Raven has shipped. The developer ecosystem is building. What does success look like for you?

Thomas Suarez   00:27:09

Yes. I mean -- so to me, the Raven is kind of a journey that started personally back in 2013. And I've been -- I actually built my first AR prototype in that year before I got my hands on Google Glass. And there's -- we've had this vision for a long time around what a spatial device looks like and a spatial operating system and bringing us out of the 2D screens and into the 3D world again. One thing that we've noticed is that screen time between the phone and the ambient computer actually goes down, not up. Total screen time compared to just the phone goes down by about 25%. This has been consistent in our team. And it tells us that we're on to something in terms of this mission of having technology be there when we need it and gone out of the way when we don't. And there's something just really powerful about that. And it's a feeling as well. It's -- I think it's hard to explain how magical that feeling is, but it's so wonderful to be able to do that. We -- over the next 10 to 15 years, we have a road map for all sorts of really cool technologies coming out, and we want to maintain the form factor. I think that's really important to make it look and feel like something people would actually want to wear. But there is going to be a tremendous amount of technology getting packed into these systems that allows it to be a really like a true stand-alone computer over time. And we're excited for all the creative aspects that come out of that, also the productivity aspects that come out of that. But also there's a lot of kind of human factors and social aspects that come out of hopefully building this device right over the next few years. So...

Neil Barbour   00:29:10

That's great. Pulling people back into the physical world. So that's Tom Suarez, Founder of Raven Resonance and the engineer behind the Raven Prism. Links to our overview of Augmented World Expo and our Smart Glasses forecast in the show notes. I'm Neil Barbour. Thanks for listening, and thank you, Tom, for joining.

Thomas Suarez   00:29:25

Yes. Thank you.