Luminar Technologies, Inc. (LAZRQ) Earnings Call Transcript & Summary

Well, welcome, everybody. Welcome to our second annual Luminar Day. Yes. My name is Emily Shanklin, I head up marketing here at Luminar. We are broadcasting live from our new Silicon Valley headquarters here in Sunnyvale, California. Fun fact, this facility used to be the offices of our friends over at Volvo. So it's particularly appropriate that we are having our first West Coast installation of Luminar day here in this facility. We've got a great morning plan with lots of updates across the business. So without further ado, let's jump right into it with our Founder and CEO, Austin Russell.

All right. Thank you, Emily. Amazing. All right. Welcome, everyone. So great to see some familiar faces, new faces, all of the above. And a warm welcome to everyone on the live stream here. So it would be great to get this all kicked off and full speed ahead. So first, I want to be able to just set the stage again, just a reminder of the problem that we're solving and the impact that we can have. So on average, American driver spend on the order of 93 billion hours and 3 trillion miles driving every year. And at a global scale, that ultimately leads to around 1.2 million annual fatalities from vehicle collisions, very, very serious problem that has to be addressed. And one of the leading causes of death and fatalities among particularly below ages 44. It's indiscriminate of age all together, and this is something that affects everyone, loved ones, those and beyond. And not to get too morbid, but I've said this before, statistically, on average, on the order of 1 in 100 of us will tragically lose our life in a vehicle accident. And that means that statistically at least what between here and live stream, 5 people are more likely than not to tragically lose their lives at some point because of this fundamental problem with the automotive industry and with vehicle accident. And we know this is a problem. The thing is, is that even with the most advanced technologies out on the road today with cameras and radar systems, even with some of those basic kinds of scenarios that we're talking about for assisted driving capabilities, they fail -- they can fail at 70% of -- for even to something as simple stopping for a pedestrian in automatic emergency braking type functionality. And that's just one of many examples of the kinds of things that are important. And we're going to get into a little bit more around the statistics as well as some breakthrough new safety research from Swiss Re that is able to help prove some at this point. And show the benefit of what we can do. That ultimately amounts to, for example, in just the U.S. alone, an estimated $1.4 trillion of annual economic societal harm from motor vehicle crashes. And I mean, we're seriously -- we're talking like a material percentage of the U.S. GDP at this stage. I mean this is -- it's a very, very real problem. It has to be addressed. So one interesting comparison I would like to draw that's like a relatively recent example is how safety is handled in other kinds of other aspects of the transportation industry and other kinds of forms of transportation. There's a very recent example in the aviation industry where late last year, early this year, there was kind of a shocking incident of a door plug blowout on a 737 MAX that got everyone into a frenzy public outrage We all saw the headlines around this and ultimately led to the ouster of the CEO of Boeing in the wake of ongoing safety concerns for a nonfatal incident. And I think it just goes to show of the kind of standards that can be set in certain transportation industries about how seriously safety is taken. So you take a look at this compared to the automotive industry. If you take a look -- we talked about global automotive fatalities on the order of 1.2 million per year. Take a look at that equivalent number for the commercial aviation industry. Take a guess. What do you think? 1 million, 0.5 million, 1,000, 72, that was it last year. This is the kind of standard that automotive should be moving towards. And just to take a little bit of an example as a reference, given just how much focus there was on that kind of incident. If you were to take the same statistics for automotive and apply it into aviation, it would be the equivalent of a 737 entirely full of people falling out of the sky every hour of every day of the year consecutively with no survivors every single time, put that into perspective. I don't think any of us would be getting on a plane if that is that were happening. Yet at the same time, we blissfully continue on our daily commutes in our cars every day without a worry in the world. This is the fundamental problem. Now that said, there was a proposed panacea of a solution, a driverless utopia, so to say, because the identified problem was always focused around humans. And the question is how to replace humans altogether. This is what was posed by some of the largest companies that were centered around robo-taxi developments throughout the 2010s and even going into the 2020s. This was -- and with build is basically the way to be able to dramatically improve vehicle safety and prevent accidents out on the road. But here's the thing is that there's more than one way to be able to accomplish that. You don't actually need to be able to replace the driver altogether to be able to start making an impact. But nevertheless, this was supposed to be the way it is. And I've been vocal what, since probably 2017 about how this is going to be -- well, I believe, a great solution over the long term, not one in the near term that you can handle. And this was the promise. This is the reality. So you take a look at these kinds of robo-taxi efforts and whatnot. And it is an incredibly challenging problem. This is where there's really no room for error when you have to have 10 9s worth of reliability and safety to handle every possible edge case when you remove the driver. And again, this is where the vast majority of the industry has largely been focused. And even today, still, companies on the order are deploying on the order of $10 billion a year to be able to try and develop these kinds of robo-taxi projects. It's easy to get confused in these kinds of scenarios. You have the construction scenario, you have vehicles wandering into places, they probably shouldn't be at times, they shouldn't be, get set on fire. Crazy stuff like that happens. And this whole perception issue around autonomy has really led -- if you take a look at the Gartner cycle for autonomous vehicles, a little bit of a business case study. You have technology trigger event that sort of happened back in, call it, 2012. You have the Waymos of the world that are really starting to ramp up. Luminar was founded, other kinds of things there, a lot of different billion dollar deals or whatnot for over the 2010s for these kinds of robo-taxi companies as well, the peak of inflated expectations around 2019, 2020. I think -- and Gartner agrees with us, if we take a look at the star of where we are today, firmly at the bottom of the trough of disillusion at the start of 2024. And the question is, is that how and what will lift out of that trough of disillusionment into that slope of enlightenment to be able to actually realize this technology into commercial production into the hands of consumers and for people to realize the safety benefits of what these kinds of autonomous capabilities can ultimately deliver. So this is where Luminar comes into play. Luminar is all about enhancing the driver as opposed to replacing the driver. This was our big bet. And this big bet is now starting to be able to pay off instead of huge roof racks full of sensing systems and a supercomputer in the trunk and other things required to run something that's not industrialized, our focus was always around, in particular, consumer production vehicles and being able to scale in high volume by working with traditional automakers as opposed to trying to build a new vehicle to service the broader world. So this is a visual of what the next generation of safety and autonomy looks like. And we'll have more to talk about today on the key catalyst to help enable this future. What we're accomplishing with enhancing the driver is twofold. One is significantly enhancing the safety of vehicles and both for the people in the vehicle and well, people around the vehicle, vulnerable road users. Second is enabling autonomy and particularly with a focus on highways. This can ultimately enable a significant amount of time saved, considering when we're talking about those many billions of hours. A huge portion of that, maybe even on the order of half is spent on highways. Here's the thing. We call this Luminar day, a new era, and this new era begins with Volvo. And with that, we welcome Alwin, who leads software at Volvo.

Hello, everyone. I'm Alwin Bakkenes, Head of Software Engineering and Volvo cars. I'm sorry, I could not be there with you in person today. The EX90 is designed to be the safest Volvo car ever, powered by Safe Space technology, and it is the start of a new era for our company, a new era for safety, sustainability and human-centric technologies. The EX90 is smart, connected and electric and it's designed to improve over time through software updates. For the first time, we're integrating a long-range automotive-grade LiDAR into one of our cars, and we do it for 2 purposes, to improve convenience through more advanced assistance systems and of course, always most important to Volvo cars to help improve safety. Now safety is, as you know, core to who we are as a company, we always strive to keep people safe and our choice to integrate LiDAR technology in the EX90 and working with Luminar is part of this. We see that LiDAR technology holds a lot of promise to further reduce accidents and to help keep people safe. The EX90s LiDAR delivered by Luminar, can sense the road in front and is designed to be able to see small objects, approximately 2 football fields ahead both in daylight and darkness at highway speeds, giving us more time to act and avoid potentially dangerous driving situations. By coupling our wider state-of-the-art sensor set with in-house developed software running on the car's core system, we aim to have the best understanding of the environment outside the car that we've had to date. And this will help us enable the car to not only have the opportunity to become safer with time but also improve the functionality of highly appreciated features and functions such as pilot assist. The EX90 was very positively received when it was first revealed a while back, and people couldn't wait to place an order. We're starting production of the EX90 in the first half of this year, and we cannot wait to see them on the road with customers. So to the entire team at Luminar, thank you so very much for the collaboration.

Thank you. Amazing. So as Alwin was mentioning, Volvo is no stranger to safety innovations. And it's largely credited with developing a lot of the different kinds of modern safety technologies that ultimately became standardized throughout the broader industry. So everything from the original 3-point seat belt that was introduced back in 1959, all the way to other kinds of modern airbag technologies, even the introduction of systems like Mobileye. And that's been core to the brand ever since the beginning. So if you take a look at some of the kinds of innovations that they've had over the years, we're very proud to now be added to that roster of historic safety milestones throughout the broader automotive industry that Volvo has able to help introduce just as they have for other kinds of previous technologies to substantially improve safety capabilities as well as ultimately enabled next-generation ADAS and autonomous capabilities. So this is a lot in the making and not an overnight success. We began working with Volvo what back in 2017 back in the day. And this -- they were one of the first to be able to really identify the significance of what LiDAR could have, not just for robo-taxis or fully autonomous vehicles or replace drivers, but actually to be able to enhance drivers. And this was hugely significant and some groundbreaking research they were doing at the time with us. In 2020, Volvo officially decided to be able to put our technology on to production cars starting off in low volume. This is a huge milestone for the overall industry and really an industry first altogether for this kind of global production vehicle. Thing is things continue to accelerate. By 2021, Volvo cars at selected Luminar as a standard equipment for the EX90, seeing safety as something more of a fundamental right than something that should just be an option. And Luminar is the first and only LiDAR technology to be launched on a global production vehicle and introduce the standard equipment. And are proud to say that ever since they first decided to put us on a vehicle in 2020, we've expanded the relationship and volume with Volvo to date to what, over an order of magnitude, more than 10x. So we're very excited to be able to continue this relationship and continue scaling and goes to show how much opportunity there is with the automakers even that you're currently working with, that's where you unlock so much value. So we're not the only kind of new technology that Volvo is introducing. There's other kinds of key partners that it takes to be able to make this successful. And that's where you see other companies like in this case, NVIDIA which for the first time will be launched with Luminar on the EX90, very excited to do this. And this will be the first time that you see accelerated computing with GPUs, along with high-performance LiDAR and AI systems that can be launched on a global production vehicle in volume. So a very exciting opportunity. But with that, I want to give a little bit of some history to the relationship here, again, going all the way back to 2017 and some of the kinds of steps of what it took to be able to ultimately not just develop the technology, but also successfully industrialize leading up to this very special moment. [Presentation]

Thank you, everyone. Incredible history there with Volvo and thank you for the Volvo guys on here for the incredible partnership over the years. We've come a very, very long way. And it almost say that with this, the past 10 years of Luminar and well, arguably, a lot of the broader industry for this kind of system and technology in terms of its introduction has been leading up to this very, very special moment. And I'm proud to announce that Luminar has officially kicked off our starter production for Volvo cars. Now this is a -- like a quite a ways in the making for quite a time in a key industry level milestone. Like I said, so proud and a huge thank you to the team, a huge thank you to our partners, a huge thank you to everyone from our customers all the way to the supply chain that just work through this to be able to make this all possible and a fantastic outcome. So I would say this is that this is something that what -- back in the day, I think it was almost unheard of to be able to try and build the technology, build a product, industrialize this and get to a stage where we can deliver directly to automakers and do so at global OEM production standards and do so in volume. We were able to achieve that. Despite the odds, despite the naysayers, despite everything along the way, this is an absolutely fundamentally transformative moment for our company and kicks off the new era. There's a lot of different implications for this for our business. The obvious of which is finally getting into the hands of consumers for the first time you can have this kind of capability of technology that the industry has seen. Economically, this is the first time our nearly $4 billion order book will also begin to materially convert into revenue. So this is where it goes from theoretical, yes, cool. Maybe this stuff will happen, maybe not, probably not to, hey, this is actually happening now. And it's easy to forgive also some of the skepticism given that the vast majority of programs throughout the broader industry have it successfully materialized to be able to launch. But that's where Luminar's changing that narrative show how our bet on enhancing the driver is paying off and to be able to successfully deliver, to scale and to be able to get this product out on roads. And it's not just in any single geographic region. Luminar is going global. This is where we are deploying our technology across north -- many, many countries in North America, Europe, Asia, all around the world in key markets. And this is where you'll start to see by the end of next year approaching 6 figures of Luminar-equipped cars driving around, collecting data, being able to actually start enhancing the safety and assisted and ultimately autonomous capabilities on vehicles. So we're very, very proud to be able to have this deployed out on roads over the course of this time frame, and it all kicks off starting with the start of production. And of course, Volvo will have more to say in terms of the specifics of their vehicle rollout and other details there that's there place to talk about. So -- but the question is, what does that mean for Luminar, not just the long term but also in the near term? What this means is that Luminar is expected to ship more than 5x more product in the second half of this year than in the rest of our 12-year history combined, something that we should all be very proud of and something that truly kicks off our exponential growth into the future. Now what this also does and one very important measure of this is that it also helps drastically enable our economies of scale that we talked about for quite some time in terms of scaling that up while also at the same time, helping drive our costs significantly down. And that's really where the scale is needed to be able to make that happen. There's a huge difference between a factory that you're able to have at capacity versus a factory that's had a small fraction of capacity. And the key thing is that we've made the big investments to be able to deliver the capacity that's needed to successfully scale into 6-figure and subsequently, ultimately 7-figure and into the future, 8-figure volumes. So when you take a look at LiDAR specifically, which is one part of our overall story is that it's very easy to forget about just how crowded this kind of landscape was and how challenging this problem really is. If you take a look back, what, 2018, 2019, 2020, there were over 100, maybe even as many as 200 types of companies or efforts, I should say, more generally around LiDAR development. The thing is that turns out building technology that works is really fricking hard. So -- to put it scientifically. So the thing is that -- it's not just that, though, you also have to successfully industrialize the technology and be able to deliver to the kinds of automotive specifications that are needed and also the ones to deliver the performance and safety to be able to operate the vehicle at speed and to be able to do so while significantly improving the safety capabilities. And this is where you're talking about less than 10 companies. They're ultimately able to sort of get their technology work and be able to develop or deliver something. The key thing is now, Luminar is stand-alone in the sense that we've successfully built the fundamental required specifications uniquely by innovating from the chip level up. We've industrialized the technology to be able to successfully scale with the successful launch of our new factory in Mexico. We've been able to work with the leading companies to introduce new kinds of safety technologies. And at the same time, even on the first vehicle launch we standardized for global production. And in all those ways, Luminar is number one. So it takes a lot of work to be able to establish that global supply chain. And that's where we don't do all of this ourselves. Of course, we work hand-in-hand with our partners, but it's important to have the right ecosystem partners from a supply chain perspective around the world to be able to create a diversified global footprint. And this is where we successfully launched our factory in Mexico and started production in collaboration with Celestica to be able to do final assembly for the LiDAR systems and testing. Fabrinet, where we've been collaborating with the facility in Thailand that they've had for Luminar, which assembles the poor optical components and transceiver components from the design and innovations that we have at Luminar. And Fabrinet will either ship it to Mexico or respectively, to China with -- in partnership with TPK, which we announced the partnership with last year that they are successfully working with us to be able to launch and scale with subsequent product generations beyond Iris. So we have some fantastic news and updates with regards to TPK that we're going to also give later today in the presentation. So again, Luminar is unique in being able to have the footprint that can successfully scale across the globe. I also mentioned that it's unique in the sense that it's tariff optimized as well. We're working with very strategic regions, for example, with Mexico, with which has very favorable trade agreements with the U.S. and other kinds of geographic regions, Thailand, same and China having local production is particularly very important for that geographic market in order to be able to successfully win business there. So where do all the sensors go. Luminar is now awarded over 25 commercial programs. So a huge feat and milestone and continuing to accelerate. And the key thing is in this industry is that the first 10 years of everything that we've done has all led up to this launch with Volvo. That's the thing. That's the catalyst because all eyes in the industry have been waiting to be able to see, can this happen? Can Luminar successfully do this. Can they buck the trend? Can they help transform the global landscape? And today, we have our answer. What does that do, sort of with a cheesy graphic kicks off a domino effect for all of these subsequent vehicle launches. We're talking about a flurry of additional vehicle launches happening in just the next 36 months that are ultimately enabled by this EX90 launch. And that's where it's all exponential scale from here. It's a very -- like I said, an incredible journey to get to the first one. But now, we hit a strike. Here we go. One other really important point powertrains. So there's 3 different major categories of vehicle powertrains. There's internal combustion engine, battery electric vehicles and hybrid electric vehicles. And even some subcategories like plug-in hybrids and other kinds of things there. Historically, would have really seen a lot of is that Luminar being particularly conflated specifically with just EVs and EV companies and EV start-ups and all of the above. This is an incredible -- there's an incredibly important distinction here. The Luminar is not an EV company. In fact, while there is a correlation of new platform development cycles with an increasing percentage of overall EVs. Luminar is completely powertrain agnostic whether your car runs on Dinosaur juice or whether it runs on lithium-ion, whether it runs on hydrogen, whatever it may be, it's independent. The important part here is that Luminar unique. Luminar is unique in the sense that it's planned into programs across all major powertrains. And in fact, the majority of the vehicle platforms that Luminar is playing into actually have vehicle models with internal combustion engines. Why is this important? Well, we all know a lot of the trends that have happened in the EV industry lately over the course of the past year. There's been a large -- there's been a significant slowdown of sales, particularly when it comes to like some of the newer start-ups and newer types of companies in this space. And while that is certainly kind of a factor in terms of like macroeconomic types of conditions. This is not something that over the longer course of time, Luminar being powertrain agnostic ultimately materially affects our business or the value of our business. So this is where we're excited to be able to deliver across the industry for all different kinds of vehicles. One last thing that I'd like to point out is around automotive adoption cycles. And I think this is a really, really important one when you can take a step back. Because the thing is that even for what we're talking about now in expanding with our automakers today. And we're capturing -- while we're capturing more value, more vehicles than what all the autonomous test robo-taxi vehicles on the road times 100 probably for what we're launching with Volvo and beyond. The thing is that when it comes to automotive adoption cycles for broader standardization across the board, these things are very long lead cycles. Automotive industry is a very high barrier to entry industry and also very high barrier to exit. So when you take a look at other kinds of safety, fundamental innovations and technologies that have been introduced over the years, largely by specifically companies like Volvo and Mercedes, both Landmark and Luminar partners. It's generally around a 10- to 20-year adoption cycle from when the technology was first introduced to by the time it starts to become standardized on vehicles and ultimately mandated a standardized on vehicles. What we're seeing on this is in a very important trend of where actually there is an incredible and accelerating rate of adoption that not only follows the trend of some of these curves, but actually is even ahead of it. Why? Because the significance of what we can have with the LiDAR and with these adoption curves is so meaningful in terms of the impact, which we'll get to talk about in a second, validated by Swiss Re as leading a new safety study for Luminar. So point is that there's obviously some level of patients required, but when you're able to standardize, when you're able to get even a tiny fraction of the industry, say, for example, we've said before that just 3% or 4% market penetration, that could be $5 billion revenue, exponentially accelerating $60 billion forward-looking order book just a tiny, tiny slots. So it's very easy to build a massively valuable company. That's why you see even examples like the Tesla of this world that even only have, what, a single-digit percentage market penetration yet worth more than a lot of the other auto companies combined. So as we pivot to a related topic around safety and insurance wanted to be able to give nice clip from our friend, Warren Buffett here.

Autonomous vehicles, widespread would hurt us. if they want -- if they spread the trucks, and they would hurt our auto insurance business. I think my personal view is that they will certainly come. I think they may be a long way off, but that will depend. It will probably, frankly, depend on experience in the first early months of the introduction in other than test situations. And if they make the world safer, it's going to be a very good thing, but it won't be a good thing for auto insurers.

So take a word for that. And for what it's worth, I couldn't agree more around the lengthy time. It's going to be a little while by the time full autonomy is able to be introduced. But the thing is what we're changing the game on here is by enabling the safety benefits to already start happening and on global production vehicles. And this is where we're able to transform the industry. And one of the big things and trends that's happening here from exactly what we're saying is something that is honestly completely over -- been -- historically completely overlooked, which is massive disruption to insurance. And we're proud to be able to have Swiss Re as a collaborator here. They're a leader when it comes to insurance innovation. They're the second largest reinsurer in the world globally, reinsuring trillions of dollars of policies. And with that, we'd like to be able to welcome Live Ali over a video here. He is the Global Head of Property and Casualty solutions from Swiss Re to be able to speak more about their safety research and the impact that it's had.

Thank you very much. Thanks, Austin. Thanks for having me. Sorry that I couldn't be there live with you on this exciting day. And I think, again, everything that you mentioned at and the comments on Warren Buffet, it absolutely resonate. So before I go into the details of what we -- I think our collaboration together, I just want to take a step back and set to see how much of an opportunity are we talking about? So as you know, insurance market is usually divided into Property and Casualty P&C, as you highlighted, in my title and Life and Health. P&C, Property Casualty is usually your car insurance, your house insurance. And within that market segment, auto basically is representing about 40% to 50% of that market stake. We're talking about, for example, from a few years ago, statistics of roughly over USD 700 billion of more premium around the world that is above USD 300 billion. So it's a significant portion of our market. And of course, Swiss Re the top reinsurance companies in the world and us being in the business now for over 160 years. We have deep expertise in this segment and a lot of background in making sure that the risks are best capture and represented in the market. Now traditionally, as I said, over 160 years, we've been in the risk transfer business, providing products has reinsurance to the market. But more and more, we've also partnering with our clients on risk partnership and risk knowledge. That's where my business units, reinsurance solution comes in, where we're looking to how we can bring our knowledge, our deep expertise in risk assessment, in underwriting and our expertise on the technology side together in providing solutions and the products that our customers could use to better assess the risk to big decisions smarter and faster. Within my area, I have a team called AMS, automotive and mobility solutions. And this team has been now developing solutions for the auto segment of our insurance industry and our clients for many years. One of the products that we have is called AV risk assessment framework, which is the subject of this conversation today. And what this is, is basically, we've developed based on a large database of accidents and claims information, a very comprehensive testing methodology that reflects the real life scenarios and accidents that can happen with the importance and weights of those accidents based on the information and data that we have. So quite an interesting product and quite a lot of excitement around it. With Luminar and Austin's team, we started working over a year ago to determine how can we actually bring the best of the 2 sides, our capabilities and know-how in the risk assessment and also with the fantastic technology that Luminar is developing lighter sensors, as well as a software stack that goes with it and determine how we can show the fact that these systems are improving the risk landscape and improving the safety of the cars on the road. So what we did, we started the journey by basically selecting a car which was basically a model year 2003 and equipped it with the Luminar Sensor and the Luminar software stack and started testing it for the comprehensive testing protocols that we had and recording the performance of this vehicle compared to the vehicle without the Luminar technology, with its own basically ADAS -- native ADAS site. Now the cars that we chose, this car and then a selection of number of other cars that we basically benchmark against each other. Basically, these are chosen to have basically the best-in-class ADAS systems on them. They're the newest model in the market with the best performance. And just to give you an overview of where they stand compared to the inventory of the car in cars in the market. Right now, in the past, years until '25, the estimate is that the new car that we have ADAS systems of Level 2 and above in the market would basically be around 30% of the new cars. Now the average car in the U.S. is -- has the age of 4.5 years, which means that the proportion of these cars with the ADAS systems in the big population of cars on the road is still a very small portion. And when we do the comparison to the best of the best, you can imagine how you can basically extrapolate that with an average car. So keep that in mind. So we did the test on this one vehicle and then a fleet of different -- basically manufacturers equivalent car with ADAS systems and benchmark the performance of these tests against each other, we did over 800 tests and basically within 35 scenarios for all these cars. And basically, what we observed and a lot of that will be available for you to look at in more detail was that the car with the Luminar radar and LiDAR and the software technology, outperformed significantly the rest of the cars that we looked at. And just to give you a bit more of the details of how much they performed -- outperformed basically. The car that was equipped with the LiDAR system Luminar system compared to the original car that was basically didn't have the LiDAR system showed an improvement in the performance of 25% in avoiding accidents. So basically frequency of these events and about 29% in mitigating, which means that the impact was reduced in the case of an accident. So quite a jump from that perspective. Of all the cars that we did the comparison to the LiDAR car. We looked at all of the other ones and the top 2 cars in that group. The comparison between the Luminar car and the other ones again were above 40% better than the other ones in some cases, the force of impacts. And then what we did was that we created a fictitious car, which basically brings the best performance across all of these cars that we looked at. And this car, of course, we call it best performance performing vehicle. It doesn't exist in reality. It's best of the breed of the cars that we looked at. And then compared to -- car with the Luminar technology, we saw that the carbon Luminar technology performs 14% better in avoiding accidents and 15% in mitigating the impact in general. So very much impressive in terms of performance of what we saw on the cars compared to -- and again, these are the cream of the crop, best-in-class cars that we basically looked at and tested and exciting news, and we were really happy to be a part of this work with you, Austin.

Absolutely. Thank you, Ali. Now this is a fantastic stuff and thank you for sharing the results of this. We know you guys have gone through a really extensive independent process over the course of the past year on this. So very exciting seeing from the last Luminar day, of when they first announced this initiative to now seeing the results of all of this and the downstream implications of this. So we'll be fantastic to be able to see and have an opportunity to disrupt the industry together. So thank you, Ali.

Absolutely. Absolutely. Thank you very much.

Perfect. All right. We appreciate it. Great. So give a little bit of some context into this with Swiss Re. And again, just as a reference point in their study, we mentioned some of this, and this is now in a public report that their accident data base is based on statistics of over 1 billion vehicle years, 70 million insurance claims. So again, just a summary of the study of a 20% reduction in frequency of accidents and other accident still happen, reduction in severity by up to 40%. And I think this is probably well beyond even some of the most aggressive estimates of what the kind of impact that our technology could have in terms of safety performance on vehicles. We're very excited to have this data and the importance they shared with -- I mean the combination of everyone from consumers to automakers to insurers to everyone across the industry, and we know that a lot of automakers have been waiting an eager anticipation of this safety report from Swiss Re. So very excited to be able to get that out there. And I think what this also highlights, most importantly, is that whereas historically, LiDAR has really been thought of in the context of primarily autonomous driving systems and full autonomy systems. It shows how LiDAR has a critical significance not just for enabling autonomous capabilities or even just higher level assisted driving capabilities, but also even the most fundamental basic levels vehicle optic safety. And this is why we believe this kind of technology will ultimately be standardized across the board and something that should be a fundamental right for every driver. So if you take a look at the kinds of camera comparisons and part of the reason why we're able to perform so much better is that because we are able to accurately sense where objects are in true 3D space, with a ground truth approach compared to cameras that really just guess from 2D data. And not only that, in other kinds of inclement weather conditions or even at night, we're able to see dramatically better all the way out past 250 meters into the distance for even very hard to see kinds of objects. So this is something that's very significant. We know exactly where everything is. This ground truth view gives us the level of accuracy and precision to be able to truly understand what's going on. And that's the important -- that's the significance for all of this. So what do these results mean economically? When it comes to the insurance economics, if you just take a little bit of a step back and just outlining a summary here. On average, the average annual U.S. vehicle insurance premium is around $2,500 per year as of today. If you take a look at that over the course of an average 10-year ownership span of a vehicle, that's around $25,000 or what, like half the purchase price of the average vehicle. I mean it's crazy to see how impactful this is. So here's the thing. Even a tiny percentage point improvement in safety makes a massive difference in terms of the economics here. So you take a look at this. Like if you were to incorporate that data from Swiss Re into these kinds of economics, you're talking about up to a $900 insurance savings per annum based on the data. And over 10 years, it's up to $9,000 in value savings. When you think about it for a second, like wait a minute, $9,000, that's like 10x even the LiDAR itself. You're absolutely right, like these numbers are incredibly impactful. And I mean to talk about from a total cost of ownership perspective, to have the LiDAR pay for itself, I mean it would only be like a few percentage points, much less 30%. So that's the significance. Now of course, how that $9,000 ultimately gets distributed to the consumer, that's the question and how that's divvied up accordingly. But that's basically just a pure pool of profit of value that's created with every LiDAR sold that's able to have that level of performance and safety improvement. So Luminar has actually taken steps in that direction to be able to capture that value or a portion thereof with every LiDAR sold. And again, we encourage all this data to be shared with other kinds of insurers out there to already start incorporating this into their models. And that's part of why Swiss Re is releasing it to all these other insurers and the insurance companies that they reinsure. So as of today, I'm happy to say Luminar is now a licensed insurance agent in 12 states with plans to be able to have full coverage of the U.S. and some regions internationally next year. This is something that we've been able to do very efficiently with our key partners. And of course, talked all about Swiss Re about how they've done so much work and investment in kind of into this area and developing these kinds of capabilities, but this enables us to be able to jointly capture that significant safety savings or value upside. It's important to note that from the upside of all of this, none of this is materially incorporated into any of our financial models into any or things like that. So our business is not by any means dependent on this. However, it's a clear huge opportunity for further capitalizing now that we're actually getting these vehicles out on the road. So we'll be able to actually ensure, for example, EX90s that are equipped with Luminar with safety savings and significant profit. So this isn't something that's just theoretical either. In addition to the work with Swiss Re and the fundamental safety data underpinning. We've already partnered with the first 2 digital insurance platforms to be able to distribute Luminar insurance and ultimately generating revenue with these partners in the near term. So very, very excited, just something illustrative as an example. Ultimately, what -- the way that this would work is that when you talk about an insurance platform, it won't necessarily always be labeled Luminar insurance is that you kind of -- you would have like a -- this is the technology platform. You'd have in many of these scenarios, like a white label back end of this, where we're providing the policies where it's actually integrated into the future, would be integrated into the order forms four different automakers. So for example, if you're ordering any kind of Luminar equipped vehicle, check a box, get your quote within 30 seconds. This is something that is not unfamiliar with folks that if you're ordering like a Tesla or other kinds of vehicles, for example, to give this as a reference point. So very exciting development for all of this and some news for us that further exponentially scales our opportunity for profits as we now get these vehicles out onto the road for each vehicle sold. I'll also say that we are in consideration for multiple automakers and there's been a huge level of interest overall, in particular, in this insurance addition in addition to the LiDAR as having this as a packaged approach more generally and something that I think over the long term, if you don't have an insurance product package with the LiDAR, that's going to be a tough sell at least from a competitive standpoint, I'm specifically referring to, if someone doesn't. Otherwise, the business model day isn't fundamental to that. But one thing I'd like to point out when it comes to the safety results and the impact of all of us have, people are taking notice, and we're very excited to be able to share these results with regulators as well. And one important point overall is that when it comes to the regulatory landscape, historically, whereas robo-taxis have largely faced significant regulatory headwinds. Luminar, on the other hand, has faced significant regulatory tailwinds. And the importance of this is that it's very different. When you have a car where you rip out the steering wheel, braking systems, trying to get something approved to be able to get out on the road, replacing drivers altogether, you have everything from FMBS standards to crack down from unions to all these other kinds of issues that are at play, whereas even without prompting from any kind of thing with Luminar and beyond, you talk about folks like NHTSA that are actually already taking huge step function improvements in terms of rule advancements, which are actually even expecting that there will be a decree in the near term in terms of a massive step function improvement of a fundamental requirement for next-generation AAV capabilities for automatic emergency braking, which will be much needed and much welcome. And ultimately, most automakers admit that they won't be able to meet that and standardize that without a new hardware solution. So that really leads up to 4 factors that are driving standardization of the technology. There's the consumer demand for safer cars and the automotive brand value that's created from this for the automakers, automotive adoption cycles and new technologies that naturally standardize over a period of time, starting with some of the leading automakers like the Volvos and Mercedes of the world that historically introduced new systems before it trickles down to the rest of the industry, regulatory pressure and mandates that will ultimately be driving standardization and lastly, reduction in the total cost of ownership of the vehicle and the insurance savings to be captured, which already use a massive business case on its own. Each one of these reasons independently is already a factor for what will drive standardization. But combined together, it's inevitable. So shifting gears a little bit. One thing we want to talk through is a little bit about the technology foundation and ecosystem that we've built. Over the years, Luminar has invested around $1.8 billion to be able to develop the leading technology foundation for our industry. And this is something that is composed of a lot of different areas here of what's required to be able to successfully deliver into a production environment to meet these specifications and requirements to be able to set up the foundation for success for future technologies and products and the broader ecosystem. So everything from the LiDAR engineering to Luminar semiconductor to software systems to industrialization to our IP leadership position to industrialize our supply chain to be automotive grade to global OEM standards, there's a lot of stuff that has to go into that. And frankly, there's no way you're doing half of this stuff without investing at least $1 billion to successfully be able to develop this. So the important point here, though, is that Luminar has made the investments that it needed to make. Now for the first time, we get to ride the tailwinds of this investment to be able to successfully scale with our business, introduce new -- these new technologies that we've invested previously to be able to develop and at the same time, be able to substantially reduce cost overall for the business on an ongoing basis as a result of the fundamental technology platform that we've created. And this technology platform has resulted in an end-to-end ecosystem step with both us and our partners that allows us to be able to truly deliver a massive value stream here. Of course, a lot of the focus has been around these first 3 areas for the LiDAR. But Luminar, again, does not just develop the LiDAR system. So we work all the way from the semiconductor level up, developing our own chips, our laser chips, receiver chips, processing chips and electronics to go into this. And we'll be speaking more to that in a little bit here with Jason as our Chairman of Luminar Semiconductor and then subsequently with the NextGen LiDAR. Talk about the transceiver. This is something that we've really designed, developed and collaborate with Fabrinet as well for the production of this, do the LiDAR system development and collaborate with OEMs in terms of integration, talk about series production, we've effectively established a leadership position there with our global footprint. Software stack. We're excited to be able to talk a lot more about that as well as an exciting new partnership that we have that we're announcing today. We have our platform partners from NVIDIA, Mobileye, Qualcomm, ECARX as from a compute standpoint and beyond in the cases of some of those companies like NVIDIA and Mobileye in particular. From an OEM partner standpoint, those are some of the public ones there. And then consumers, we're very excited to be able to now, for the first time, get this technology out of the road into the hands of consumers. We're going to have again, just as we're shipping more product out there into the world, this is that watershed moment where there's going to be massive consumer visibility for the first time around this kind of technology. People see this on the roof line of these Volvos. It's a conversation starter and something that's a really important piece for this whole new era. So -- and then lastly, of course, we talked about insurance. And the downstream effects of what improved safety can have on vehicles and how much economic opportunity there is as well as that being a fundamental driver for adoption from automakers as standardization from regulators as well as, of course, savings from an insurance standpoint. So with all that, we're going to do a little bit of a zoom in here into the software stack specifically with Aaron Jefferson, we're going to dive a little bit deeper into some AI developments with Annie. And then with Jason, as the Chairman of LSI, we're going to jump into the chip developments as well there. So with that, I'd love to be able to welcome up Aaron to speak more about our top.

Hello, everyone. The first thing you may notice is that have a little more grade than that picture right there. I ensure you that it's not Luminar driven. There's probably life and kids, the very kids I'm trying to save are the kids that were trying to kill me. But nevertheless, I'm here today to talk to you about software. And what we do with our software is really we take that 3D point cloud that rich data that we've always talked about, and we translate that into actionable safety benefits for the vehicle. We -- what we call our software stack and our software is lead sentinel and sentinel consists of 4 key components: core software, perception and mapping functions, our proactive safety and then tools and simulation. Let's dive into each of those. So core is really just the software utilities that allow our customers to really utilize and communicate with our LiDAR. Some common ones are LiDAR drivers, diagnostics, things of that nature. But what you see here is what we call our intelligent horizon tracking our intelligent, region of interest scanning system. What that allows us to do is make sure that we put our dense points and our dense lines, our density where it matters, independent if we hit a pothole, independent if we hit a heel or we go up any kind of embankment. We basically want to make sure we have that information and that data exactly where it needs to be such that we don't lose sight of what's important, and we can act on the data accordingly. As we go from there, we take all this point cloud information, and we say, "Okay, what the hell do we do with this? What does this mean? So Perception is taking that data and then creating an environment model. And we do that in 2 different methods. We utilize a machine learning AI system that basically is based on a foundational model that takes that information, looks for dynamic moving objects and use a semantic segmentation to classify each point within that scene. So from there, we can understand if it's a vehicle. We can understand if it's a pedestrian, we can understand if it's a sign, the lane and then we create this environment model based off that information. Next, we use geometric say classical geometric algorithms, where we utilize that 3D point cloud data. We're taking it and saying, okay, we understand what this is. We're putting -- we're looking at dynamics of the objects persistence of the information. We're putting bounding boxes around those, and we're trying to understand exactly the 6 degrees of freedom, where objects are heading, and we're tracking that information. We group those together. We put them in a tracker. And with that, we create these -- we create object list. This object list is fundamentally redundant and is also algorithmically sufficient to then enable autonomy as well as safety. So that is the foundation of bringing everything together and creating this environment model with our perception system. In addition to that, we can take that 3D data and the semantic information and create 3D multilayer maps that really gives us a real seen and understanding of what's going on in the environment. In addition to that, we have a low bandwidth localization layer utilizing our fingerprint technology that really allows us to run this on the cloud or in the vehicle and understand exactly where we are, actually better than RTK, our real-time kinematic systems and GPS system today. Usually, in dense cities, tall buildings you lose sight of the GPS systems lose sight of where it is, very -- a lot of talent there that you don't want. And so our system is able to really localize the system, again, for support all the way from safety to autonomy and really drive that capability on a vehicle. And then all that builds up to what we call proactive safety. Proactive safety is as a product person, what I really love because I want to be able to tell my customer how great our LiDAR. Is I want to be able to show these nice point clouds, but I don't want to say, hey, please believe me, it works. It's going to translate into a vehicle function. I really want to be able, as a team, to be able to go to a customer and say, "Hey, this is what it does and take a vehicle like you'll see later today and show you cemematically what the vehicle can do with this technology. And that's what we have here. And so as the market shifts from, say, camera, camera radar fusion systems today, we can improve our AEB system, our safety systems, higher speed, inclement weather, safety maneuvers, we can make safer maneuvers over time. And then also, we can implement better highway assisted driving higher speeds, more availability, again, and inclement weather and things of that nature, we can make sure that our system is always ready to go. To make all this work, our tools and simulation. This is what we use to basically validate our software. And then we also utilize these systems so our customers can put it within their systems and make sure that they understand what our LiDAR can do. It is a tool that helps us validate our capability, but then helps the consumer or the customer integrate our capability into their system and really deliver a full software solution. So we're excited to be able to develop the sentinel solution. We want to be able to make sure that we have a solution that is ready for the market that we are developing such that we've not only proven out our software capabilities but we're delivering a solution to our customers that they can utilize and put into their programs. Next, we're going to take a deeper dive into AI with Annie.

Yes. I'm Annie [ Guan ]. I lead AI and the machine learning at Luminar. As Aaron presented earlier, our software transformed the 3D LiDAR point data into real safety benefits. A key element of this transformation is a detection of 3D dynamic objects, including production vehicles and classification of critical features such as landmarking and road services. This information is essential for the vehicle to make informed decision about its surroundings and how to navigate. In order to achieve these capabilities, we leverage our AI engine to turn raw point clouds and label data into highly performing models that are optimized to run on real-time systems. Now let's dive into the details of our latest foundation model. Our latest model is a transformer-based design. Similar to those powering the large language models, it serves as a foundation model that can be fine-tuned to multi-test perception. The model detects 3D Dynamic objects, including the location heading with and it can also perform point-wise classification for road, lane markings, edges and other static features. As we enter this phase of production and widespread deployment of Luminar LiDAR, we will have access to huge worldwide data sets, which will allow us to train larger and larger models with improved performance. As we navigate through a Dynamic 3D world that evolves with time, the demand on automotive safety system has grown far beyond the basic 3D object detection. Today, I'm excited to share the insights on our upcoming research focus, which leveraged advanced reasoning capabilities of Generative AI, specifically large language models, with the precision of the LiDAR technology. Our goal is to develop an end-to-end AI agent that takes the raw point cloud. And by leveraging the sophisticated reasoning capabilities of large language model generates a 3D contextual understanding, predicts future movements of the objects in the scene, execute the vehicle motion planning and enables meaningful human vehicle interactions. Traditional [ TDC ] understanding is good at detecting objects, but it fails to capture depth and geometric details, essential for completing analysis. Here, you can see the output from the ChatGPT based on 2D camera images. While identified objects and scenes, it fails to provide a comprehensive understanding of the dimensions and precise location. With LiDAR data, the AI agent can take the full measure of the environment, not just detecting the presence and object, but also understanding its dimensions in its location. This level of detail is what we call contextual 3D scene understanding. Similar to how a typical [ AOM ] use the current token to predict the next token, the AI agent coupled with LiDAR technology can also predict how the thing will evolve. For instance, what the pickup truck will do with exact location in our example. Guided by the reasoning price of 3D understanding and prediction, the AI agent can also generate a planning-related language instruction. With AOM and LiDAR integration, the AI agent can also enable meaningful interactions between human and vehicles. Driver can receive auto information on observed environment be alerted to the obstacles and be informed on the decision-making process involved in the collision avoidance. [Presentation]

Human vehicle interaction can also be facilitated through 3D visual question-and-answer where the driver can inquire about potential safety hazard around the vehicle, the AI agent will then identify and describe any detected hazards and recommend appropriate safety measures. To enable those capabilities, we need to transform the intricated geometric LiDAR data into a representation space that can be understood by AOMs. We will extend our current model to learn alignment between the representation of individual 3D objects and corresponding tax embedding of the AOM associated with the objects. These integrations unleashes AOM's reasoning and the localization capabilities of LiDAR technology. Now while we strongly believe that the combination of precise 3D information and large language model will lead to a step change in contextual understanding, as with any safety critical technology, it will need to be validated on large data sets. Manual annotation of existing contacts is time-consuming and can be more aeroprone. In order to accelerate the development of those models, we are partnering with Applied tuition to generate large amounts of synthetic data with rich grand tools and aero-free contextual information. Now I will turn in to Austin to talk more about this partnership.

All right. Thank you, Annie. Great. All right. Well, thanks, Qasar. Welcome. Much appreciate it.

Thanks for having me.

Making the trip over.

Yes, it's a whole.

5 minutes.

4, 4.5.

Exactly, Exactly. Yes, yes, yes. Exactly. Exactly. Yes. No, I know there's a lot of folks in the Beijing Auto Show and whatnot. But cool. We're so excited to having here and be able to announce a new partnership together with Applied intuition. It's been in the making for quite some time.

Yes. It's -- yes, super proud to work with you guys.

Awesome. Awesome. -- maybe thought to be good. You can kick off a little bit more about what we're doing. And for context, part of the goal of this is to be able to help accelerate the adoption of the next-generation ADAS and autonomous technologies with automakers across the board. They've Applied tuition has had some incredible success lately, incredible work and also what you guys raised a great round like a $6 billion valuation there just recently to congrats too...

In the last month yes. Just for context for everybody, we provide tools that automakers and other vehicle manufacturers used to develop autonomous systems and ADAS systems. Of the top 20 global OEMs, 18 use our products, and we've been -- that's the business that we're in. And so we often look to partner with strong technology providers because if you're trying to develop an autonomous system or an ADAS system, the sensors become obviously a big component of that. And what we're announcing today is we've taken Luminar's sensor model in a validated form, and it is available in the Applied and tuition suite of tools, which is a fantastic stuff for the industry, and it's a first for the industry.

Absolutely. So yes, very exciting one there, too. And we know that one of the biggest challenges for automakers in terms of the speed of adoption around being able to work with some of these kinds of tools that deliver to be able to have that testing and validation and to be able to create that synthetic data. So it's super meaningful and our software team has done a great job to be able to create those kinds of capabilities. Now getting that integrated into the Applied intuition suite is great and hopefully get to share that now also with your broader customer base?

Yes, exactly. The -- if you're an automotive engineer and you're building an autonomous system and you're using Luminar, typically, you'd have to actually get actual LiDARs for testing and development. And now you can do all this in a virtual environment. And so what a validated sensor means is it's closing that sim-to-real gap. So there's a real LiDAR, and then there's a synthetic version of it, a simulated version of it. What's special today that we're announcing is we're taking not only can you model all of the traditional physics properties of a sensor or a radar or a LiDAR or a camera system. But in today's particular announcement of partnership, we're also integrating Luminar software into the sensors. So it's like a firmware in the loop sensor simulation, which -- why that's so great is so Luminar has some OEM customers. We have some OEM customers. Those customers can now interchangeably use our development environment. For instance, if you're an OEM and you haven't used Luminar's LiDARs, now you can use it in simulation. And it's validated. It has Luminar software in the sensor and in the simulation. And so you can actually try out the LiDAR without actually even having physical samples.

Yes, absolutely. And then that's also huge for subsequent generations of products and things like that. We'll talk more about next-generation LiDAR later on. But for stuff like that, the fact that you can already start working with it in simulation even before the product launches at scale is fantastic.

Yes, exactly. And I think the fundamental reality is the whole industry is moving this way. And it's great to partner with you guys on kind of a real industry first here.

Yes, absolutely, absolutely. And also awesome to see you guys working through with a bunch of major OEMs and generating some great business with -- across the board with the whole tool chain there. So very, very impressive. And I think generally super complementary to all the work that we do at Luminar there. So it'd be a great partnership ahead and definitely transformative for getting these kinds of engagements and getting people off to the races with software development.

Yes. And yes, thanks for the partnership.

Absolutely. Well, cool. Thank you, Qasar. I appreciate you coming by.

Thanks for having us.

Okay. Much appreciated. And with that, we'd like to be able to welcome Jason as the Chairman of Luminar Semiconductor to be able to speak to some of the different next-generation developments that we have for our chip technologies, some of which also enable these step function improvement and capabilities for the LiDAR that we'll be talking about subsequently. All of this goes back to that $1.8 billion technology foundation that we talked about to be able to have this kind of incredible IP mode technology advancements and able to actually transform the industry from the chip level up, which is truly what's required to meet the specifications to be able to deliver, to be able to scale, do so efficiently, do so cost effectively, et cetera. So with that, welcome, Jason.

Thanks, Austin. Yes, as Austin talked about, we've started -- fundamentally, when we started Luminar, we knew that we would never sacrifice performance of what could be done based on what was available commercially off the shelf. So when we started the company, we knew the only limitation we had to the performance of what we were doing was going to be the laws of physics, things like the speed of light being too slow, those types of things that we deal with, but not what we would have to sacrifice some performance based on what we can get from a traditional photonics supply chain. It was the laws of physics. So Luminar Semiconductor, we went back to the drawing board. And we said, what do we need to do to fundamentally advance the product road map for Luminar, starting with next-gen lasers, receivers and ASICs to get the highest level of performance at the best level of system integration to drive the economics at a fundamental level? We do that in 2 different areas. We start on the laser side and the receiver side. Where did we start? We started with the acquisition of Black Forest Engineering and then OptoGration and then finally Freedom Photonics. And bring all 3 of those companies together allows us to offer a level of performance and system integration that was unheard of in the industry. We spent a majority of our early years fighting whether or not we could operate in the eye-safe region. But by directing the work around the receiver and directing it around the laser as it followed, we were able to offer that level of performance. Here's an example of some of the systems integration we're able to do. This starts at Freedom Photonics. All of this was unlocked by the performance of the receiver technology from OptoGration and BFE. So here, you have a standard Gen 4 laser, as we call it, which is an indium phosphide laser. But it's a very specialized laser. That laser is not available commercially to anybody in the world, and it offers a performance in terms of efficiency and power that then unlocks laser architectures in our entire system that could not be done any other way. So we take that laser. But if you overlook some of the opportunities, the afterthought for most companies is the drive electronics. At Black Forest Engineering, we were the, and still are, the leader in mixed-signal ASICs, analog and digital. The mixed-signal ASICs, we're able to then drive that laser chip, put it into a package, and this is -- and the integrated laser is one of our earlier packages. But if you see the ASIC underneath here, that level of systems-level integration where we get rid of the hybridization that's typically required, do it in a new way, and we're able to figure out the packaging so that the packaging can be put together in such a way that we can drive the economies of scale. The drive electronics are now essentially in the chip level. The laser is at the chip level. And that unlocks everything else. Where we started, though, and what was unlocked was the receiver technology. If you look at the receivers, and this is from OptoGration, we're now in our fifth generation of back-thin InGaAs APDs. By taking these APDs and back-thinning them and directly bump-bonding them into the readout integrated circuit, which is in silicon, we're able to offer unprecedented levels of noise and sensitivity at the same time. We talk about budgets. You talk about everything that gets done inside a system architecture design. You look at the economics. We look at photon budget. This receiver technology allows us to offer the highest budget possible because we're operating in the eye-safe range but need the least amount of photons to be able to go out 200, 300 meters and see that 10% reflector at range. This technology that you see here on the screen is what unlocks the Luminar LiDAR. Austin talked about the stack and how we can take the stack and start with them in our semiconductor. That is -- and you're going to hear later today and talk about our future product platforms. Those future product platforms come out of chips that look like this. When you take OptoGration, you take Black Forest Engineering and you take Freedom Photonics, there were over $150 million of government funding that went through those companies prior to the acquisition by Luminar. Those companies had 2 things. They had core technology and IP, and they had incredible teams. The subject matter expertise within those companies was unprecedented. If you look at what we're doing, we are taking this as an example. You take the technology that was developed for LiDAR, but you take it for other markets in the photonics space. Here, this is an example of how we're leveraging Luminar Semiconductor's breakthrough laser chip technology and developing a multi-wavelength source to enable the next wave of AI data centers. This is part of a DARPA program called pipes that Freedom Photonics is part of, 16 different wavelengths. We're talking about trying to get terabits of data off of a chip, picojoules per bit. There's not a single person in this room who isn't aware of the challenges of power consumption within data centers today. The AI revolution is being unlocked by incredible work in data centers. And for years, we've been working on this technology in Freedom Photonics. We have close collaborations with industry partners around the GPUs and the CPUs, both commercially as well as in government programs. That efficient uncooled Luminar laser is just another example of the power of Luminar Semiconductor when we bring all these technologies together. I've talked about that $150 million of funding prior to the acquisition and then even more since. This is the power of the company coming together. I want to talk through 3 different things as we look at this. We've got brand-new laser technology with those 16 wavelengths, very narrow wavelengths to unlock the telecommunications, get the data and those bits flowing. We are one of the leaders in the world in photonic wire bonding. We can take a wire bond in optics and photonics via a 3D printed wire, similar to the gold wire bonds that typically everybody else does, and in [ 3B ] move light from this chip to this chip and bond them together. We have a star coupler, and we're able to get those light via the photonic wire bonds in and off that chip. Three companies who've come together to unlock the technology. This is the future for Luminar Semiconductor. Thank you.

All right. So we're going to bring things back to LiDAR. We've been talking about a lot of things. But we're here to talk about LiDAR. And where we're going with LiDAR is really this next topic. We've talked a lot about scale. The question is not how do we start scale. Obviously, we're there. The question is acceleration. How do we start moving towards mass adoption of this technology? Because if we can't get this technology and all the benefits that have been talked about today into everybody's car, not just the luxury lines out there, we won't achieve our mission. And so what are the barriers? And what is the classic priorities to scale in automotive? Now the barriers are obvious ones. We all know these things. Cost, integration and performance. You have to be able to have something affordable. You have to able to get it into a vehicle, and you have to be able to do something with it. Now what's interesting here, though, is this order. And this is the classical way things have gone in technology adoption in automotive. Cost is king in vehicles, historically. But unfortunately, it tends to be cost at all costs. Now that means at every opportunity to cut a little bit of performance to save a few dollars, it's happened. It's really common. Next up is integration. Okay. You've got something that can be affordable, not really impact meaningfully the cost structure of a vehicle. Where can we put it? All right. Well, let's put it somewhere that's convenient. There happens to be a hole somewhere in the vehicle. Let's shove it there and see what it can do. And that really leaves performance last. Unfortunately, that leaves whatever you can do, whatever performance happens to be left after you've cut all the cost out, after you found a place to shove it. This is why vehicles today don't really perform very well, especially consumer vehicles, which are heavily focused on cost reduction and design aesthetics. At Luminar, you might suggest or you might understand that all these things we're talking about, we're talking about inverting this. We have the performance first. The whole point of bringing LiDAR into this industry is to give new functionality to the world of mobility. We need to change this priority list to make sure that we are delivering performance to deliver desirable function. We want to give value, benefit. We want to pick integrations that balance form and function, that balance our ability to do interesting things and make a car look cool and want to be in your driveway. And then finally, cost is still very important. But we want to optimize value. We drive costs down while maintaining performance, maintaining benefit, and that drives value to the consumer in all of its applications. So everything we've talked about today, including Jason from Luminar Semiconductor, we have been doing this for a decade of technology development through hardware, software, integration, industrialization, all of these things that have been talked about focused on the most valuable use cases. This is about value generation for the consumers, which trickle down to the whole supply chain. You might recognize this -- the bones of this slide from last year at Luminar Day. It remains very valuable and interesting to track progress. What we have here is a charting between how far away I can detect a threat and then how fast I can therefore safely drive and avoid that threat. In this example here, specifically, we're talking about small obstacles, a tire in the road. This is the kind of thing that is not very relevant to human-driven systems historically. ADAS, AEB isn't normally going to stop for something like this in the road, although maybe it should in some cases. But it becomes centrally critical to autonomy because when you're not paying attention, you're not driving, hitting a tire or a brick in the road, things like that can be, one, a terrible experience but also can be dangerous. It can lead to secondary collisions. The leading kinds of collisions in the world right now happen to deal with lane departures, which is a combination of paying attention and poor reactions. And what we see here is a nice mapping of really what shows off with Luminar [ specialist ] with LiDAR. We can attack these small, difficult, dark things very far away, and that's what unlocks the vehicle speed. And when we think about those priorities again, performance is one. Integration is 2 and cost driving value is next. And when we look at Volvo and the unveiling of this vehicle, that prioritization is clear. They chose the highest-performing LiDAR sensor available to them. They chose an integration location that was going to maximize its capability on the roadway, embracing that techno-centric design, and we're working together to drive cost and benefit to the consumer. The competitive space is, of course, working, right, in the background, trying to catch up to what we're doing. But at Luminar, we focused on that performance. Now everybody is chasing us while trying to keep package down, while Luminar has achieved performance years ago, quite honestly, and have been able to focus heavily on miniaturization, cost scaling. And so while the market is trying to creep forward, they continue to kind of hit a barrier in how much they can really extend the detection of these difficult objects, where they're still not even in their road maps, able to catch up with what Iris can do today. And this isn't made up. This isn't hyperbole. We were actually taking measurements here. Maybe difficult to see on some of these examples, but that's kind of the point. What we can do here is we can benchmark different technologies available, automotive LiDAR sensors, and kind of show the reality of object real-world detections beyond what you just might see on a spec sheet, right? Some of these sensors specify on paper something that looks or can look something very similar to what Luminar writes on a piece of paper from a sensor specification perspective. But there's a lot of ways to play games with those individual specifications. When it comes to real-world detections, Luminar is differentiated. Availability is a really important part of this. The performance that we do, being able to see 300 meters is great, and it gives us a nice opportunity to plan and be proactive about our reactions -- our actions in the environment. But it also helps us perform in all normal conditions. It's not always fair weather like California in the rest of the world, and our partners in Sweden know this very well. And a lot of time, as people specify a LiDAR, and it happens to be in the best possible conditions, which actually happens to be night, interestingly enough. No background light to deal with. But we turn the lights on. The sun comes out. A lot of times, these wavelengths are going to degrade. The world is much brighter at these wavelengths. These architectures are driven to reject less ambient light. And so they have a range reduction that can be very clear in some data examples. You introduce poor weather like rain, and everybody is going to degrade, but we start with a strong range position that we end up still seeing much further. Things get even worse in dust and fog and our scattering [ cross-section ] longer wavelength helps us maintain performance. And importantly, also here, not only are these wavelengths of operation important to how we perform in different weather conditions. It also is important to think about camera and camera technologies that we're often kind of compared to as what is necessary for the task. And what's important with LiDAR is we degrade slowly in all of these conditions. And so -- and our loss of 3D data doesn't go away. We don't lose the information there. We just detect less far away. With a camera, you commonly approach kind of a binary shutoff point where there's not enough contrast in the scene because the fog's too dense or rain kind of ruins your edges and it's difficult for you to understand the 3D or even 2D understanding of the objects. And so this degradation is really important to understand. There's a lot of misinformation out in the world about it. But our performance nominally gives us a huge advantage in all conditions, which again drives desirable performance of functionality, availability on these different conditions. And so we've been working at this a long time. Now if you're following the Luminar story, you know this. Model G, this very military-looking sensor you've seen here, moving towards Hydra, Iris, which is what's going to be going into the EX90 and all the other programs that have been talked about. This shows our track record. Performance has increased in each of these steps, and package has quite obviously decreased. And we are not done. What we're doing here is continuing a road map of size and cost to drive adoption, break down the barriers of scale while maintaining or improving performance and really doing everything in the rest of the business to prepare for scale. And with that, I'll hand it over to my new colleague, David Foster, to talk a little bit about industrialization with our partners, TPK. Thank you.

Okay. Thanks very much, Matt. So delighted to be here today, and in particular, to introduce old colleague, Leo Hsieh from TPK, CEO of TPK. So have a seat there, Leo.

Thank you, David. Thank you.

And we're going to just talk a little bit with Leo. Perhaps start by introducing yourself and a little bit of what TPK does for the audience, please.

Yes, certainly. TPK started as a touch panel maker. All products are inside the first iPhone, the first Kindle and the first Tesla. We are good at product engineering. So we are a good partner with system product makers in industrialization and mass production. With Luminar, we can help accelerate scaling, factory automation and supplier management for speed, efficiency and cost reduction with high quality.

That's great. So some might say that we're almost approaching a sort of iPhone-type moment with what this technology can bring to the automotive industry. As a company that was involved with that iPhone from its very first generation, how do you think about that move to industrialization and to mammoth scale?

Yes, certainly. Actually, we feel it. The moment right now for Luminar is exactly the same as the moment for iPhone back in 2007 because this year, Luminar is going to mass production for the first model. And we have no doubt Luminar will run fast with huge success.

So why does TPK -- why do you believe so much in LiDAR as a technology that can help advance automotive safety?

Yes, certainly. As mentioned by previous speakers, we all know about the good thing about LiDAR and the good technology and the competitiveness of Luminar. For me, I would like to refer to the experience that we work together with Luminar. I firmly believe Luminar will succeed because first of all, Luminar has the gene of Silicon Valley and the working attitude. They are very initiative, creative, passionate, proactive. And all the top management is paying all the details and hand-ons on the details. Moreover, Luminar has the strong command in all the key components, and some of them are even in-house supplied. So certainly, I have no doubt it has all the success factors with previous established companies. We see all the success factors also in Luminar.

Thank you. Tell me a little bit about TPK's capabilities in industrialization and advanced automation. You're a world leader in this space. How did you grow that business to that point?

Yes. That would -- I need to thank our previous and existing established, successful system product makers like Apple, Amazon, Microsoft, et cetera. You can imagine that we need to help industrialize the product within 2 years, and right now, shrink down to 1 year. And for the time to market, we need to bring up the volume from thousands pieces a day to 1 million pieces a day. So you can imagine the speed and the efficiency that required.

Exactly. No small feat to help scale things like the iPhone into the sort of volumes that it achieved. So one of the really exciting things for us is the industrialization center that we are just setting up jointly with TPK. And it's part of a sort of globalized industrial model that I've certainly worked with in prior companies that works very well. Tell us a little bit about how you think that can improve the efficiency of the way our teams work together.

Yes, it's a perfect match because Luminar has very good product design, system integration, and we are good at product engineering and automation. So there is a perfect match to help the product shorten the time required for industrialization and also help the cost reduction a lot.

That's perfect. So again, thank you, Leo. We really appreciate the partnership, and we look forward to together bringing this technology into much higher volume at scale with TPK.

No doubt. Thank you.

Thank you.

All right. Thank you, David and Leo. Great cut there and very excited about this expanded TPK partnership. Improving efficiency, improving costs and being able to scale further is huge for us. So a lot of great stuff ahead. So in closing out Luminar Day, I want to talk about one more thing. So the overall goal of what we're doing here for our next-generation LiDAR is to be able to enable a transition from premium vehicles towards mainstream vehicles. We've already demonstrated leadership when it comes to the premium market segment, which is important with working with some of the companies that you see the kinds of technologies that are introduced ultimately trickle down to the rest of the landscape. And this is important to fulfill our ultimate vision of democratizing safety. And what we've done to be able to do that starts with the LiDAR. We talked about some of the other components of the stack, which is absolutely fantastic and critical. But there's more to be done to be able to enable that to be successful. So our next-generation LiDAR is -- starts with a simplified elegant architecture. It's leveraging the Luminar Semiconductor breakthroughs that we've spoken about, step function improvement in performance, integration and cost and, of course, automated assembly and integrated supply chain and working hand-in-hand in particular with TPK, among others, to be able to make that happen as part of the expanded partnership. So very excited for the future ahead with this. And with that, let's let it roll. [Presentation]

So this is Luminar Halo. For the first time in 5 years, Luminar is introducing a fundamental step function change in technology, a new technology platform that has been in the works for quite some time, and we're very excited to be able to finally have the opportunity to be able to bring this to reality, to bring this to fruition and to be able to truly enable mass adoption to democratize safety. Could talk a little bit about the kinds of benefits that we have as well as the technology that's gone into this to make this possible. So first off, what's clear, super small. It's really half the height of Iris, 1/3 of the volume. And when it's integrated into a vehicle, you're talking about 1/3 of the total height as compared to some of the current vehicle platforms. It's ultra-light at a ratio of 3x lighter than Iris. We have a huge step function improvement in thermal efficiency, about 2x as compared to the Iris family and, at the same time, are able to have integrated thermal controls that are able to maintain the full performance all the way from negative 40 Celsius all the way up through 85 Celsius. So huge temperature range that's required for automotive and to be able to deliver all of these capabilities. We're able to have a sleek design, aircraft-grade aluminum, IP68 water resistant, et cetera, among other things. So we start off with the integration. We're talking 3x smaller, 3x lighter and 2x the thermal efficiency, huge step function that allows us to be able to get this into more cars more efficiently. And simplified, it's about 1 inch, 1 kilogram, around 10 watts. All right. So jumping in into some of the different aspects of the technology. There's a lot of stuff that had to go in here to make this all possible, again, for the semiconductor level up, 15-nanometer technology and elegant integrated architecture. But of course, we'll talk for a second on some of the semiconductor enablements that have made this possible. So there's the analog component and the digital component of this. We have the Luminar analog processor that we successfully developed. This is part of the design. We have a digital ASIC as well that does an incredible amount of processing on board as part of this that is able to make some of this performance possible. And also combined with some of the next-generation integrated circuits and laser diodes, we have our indium phosphide laser diode or indium gallium arsenide photodetector. We have our laser driver that was talked about. And all these next-generation technologies have -- that are part of this $1.8 billion technology foundation that we talked about make this all possible to achieve these kinds of metrics. So if you take a look in aggregate, there's about 5 unique chips that we have that we've successfully developed that make this possible. But of course, it's not just about the integration. It's also about the performance. And if you take a look at a -- see a nice visualization of the kind of things this can see. So we're talking, in aggregate, a 40x improvement in optical efficiency for photon signal-noise ratio. So massive step function in terms of the kinds of core performance for the transceiver system into the LiDAR. Now we're using a lot of that performance to be able to improve efficiency, reduce size, get better integration, a few other factors. And where we end up at the end of the day is around, in aggregate, about a 4x improvement in the LiDAR point cloud performance compared to Iris. So about 400%. And last but not least, we're doing all of this. Again, we're dramatically improving the integration. We're dramatically improving the performance, doing all this while improving the cost by a factor of 2 as well. And the key is that as Iris is designed for hundreds of thousands of sensors, Luminar Halo is designed for tens of millions of vehicles. And again, we made this so sleek that it's able to fit into all different kinds of roof lines. So you can see a little bit of a visualization of this. Question for the audience. Anyone recognize what this is, this image on the bottom? Yes. What does Luminar and a 1967 Pontiac GTO have in common? It's a hood scoop. Looks remarkably similar, doesn't it? It's so interesting, is to see the design trends. Really, as it should be, form follows function. And if you take a look back in the day through like muscle cars and other kinds of vehicles, you see these kinds of air intakes and other things being developed actually as a feature, something that's a premium feature on the vehicle that indicates better performance for something. And it's funny seeing even the trends today of what you see with companies out there. Like a lot of times, these kinds of events that you'll see on cars will actually literally be like fake vents like they don't actually even go anywhere. That's where -- that goes to show the power of design and design elements and features there. I think the point is that it actually can become something that's meaningful and attractive, and it's something you get used to over time but something that when you know that you have this Luminar Halo on the roof of the car that you can be safe in your vehicle or that much safer. It makes all the difference. So Luminar Halo stands only 25 millimeters in height and just 16 millimeters once integrated, again, roughly about like a 3x improvement in total integration height and sleek factor, so to say, compared to some of the current vehicles that you'll see out on the road. So this is a huge step function improvement here. But on this, there's also an additional mode of integration that we're excited to share because the thing is that while the roof line integration will ultimately give the best performance capabilities for the device, it's important for us to be able to give freedom and flexibility to vehicle designers to be able to do what they see fit. And with that, we're excited to share that Luminar Halo can now be seamlessly integrated behind the windshield. All right. So with that, we have a windshield from a Korean automaker here with it like integrated into the roof line and underneath, so you can actually see it. We can see even it's like a reference camera design against it there, for example, like a Mobileye system or equivalent. That's part of the actual frame and body itself. And this is something that gives a very, very sleek view of the future of what it can look like. Very good. Lastly, we're also offering an option for integrated cameras into the LiDAR. This is very meaningful when it comes to sensor fusion capabilities, being able to blend this. We believe that while, obviously, LiDAR is able to take things significantly further, there are, of course, benefits to having a camera on a vehicle when it comes to RGB sensing. And there is a benefit to being able to have it integrated into the product. So this is something that we believe should be a great capability for automakers, an integrated option. It allows us to be able to do incredible sensor fusion to be able to even further enhance the performance of the LiDAR. So in summary, Luminar Halo, 3x smaller, 3x lighter, 2x thermal efficiency, 2x cost improvement, 4x improved overall point cloud performance altogether, powered by all the new semiconductor technologies that we have in here, enabling new modes of integration and first available starting in 2026. Note, this is just the start. We have an opportunity to be able to fill our long-term mission with this, saving as many as 100 million lives and 100 trillion hours of people's time over the next 100 years. This is what everything that Luminar ultimately rolls up to as part of the long-term vision, and we're taking step by step the ways to be able to ultimately realize this. Now of course, with the launch of Volvo, with the launch of the next-generation LiDAR, with the new partnerships with everything that we have going for us, the reality is that Luminar has never stood in a stronger position than in our 10-year history to be able to fulfill our mission, our value and our business, to be able to realize the goals of what we've all been waiting for. So very exciting, very proud to be here. And thank you for joining us. Welcome to the new era.