管理層發言
Thank you for your continued patience. Your meeting will begin shortly. If you need assistance at any time, please press zero, and a member of our team will be happy to help you. If you need assistance at any time, please press zero, and a member of our team will be happy to help you. Good day. My name is Stephanie, and I will be your conference facilitator. I would like to welcome everyone to today's Aeva Technologies Second Quarter 2026 Earnings Conference Call. During the opening remarks, all participants will be in a listen-only mode. Following the opening remarks, we will conduct a question and answer session. As a reminder, today's conference is being recorded and simultaneously webcast. I would like to now turn the call over to Andrew Fung, Senior Director of Investor Relations and Corporate Development. Andrew, please go ahead.
Thank you, and welcome, everyone, to Aeva's second quarter 2026 Earnings Conference Call. Joining on the call today are Soroush Salehian Dardashti, Aeva's Co-Founder and CEO, and Saurabh Sinha, Aeva's CFO. Ahead of this call, we issued our second quarter 2026 press release and presentation, which we will refer to today and can be found on our Investor Relations website at investors.aeva.com. Please note that on this call, we will be making forward-looking statements based on current expectations and assumptions, which are subject to risks and uncertainties. These statements reflect our views only as of today and should not be relied upon as representative of our views as of any subsequent date. These statements are subject to a variety of risks and uncertainties that could cause actual results to differ materially from expectations. For a further discussion of the material risks and other important factors that could affect our financial results, please refer to our filings with the SEC, including our most recent Form 10-Q and Form 10-K. In addition, during today's call, we will discuss non-GAAP financial measures, which we believe are useful as supplemental measures of Aeva's performance. These non-GAAP measures should be considered in addition to and not as a substitute for or in isolation from GAAP results. The webcast replay of this call will be available on our company website under the Investor Relations link. And with that, let me turn the call over to Soroush.
Thanks, Andrew, and good afternoon, everyone. Q2 was another strong quarter at Aeva. We continue to build on our leadership position with more miles and milestones delivered to our customers and partners and important expansion into new areas leveraging our same core technology developed over the last 10 years. In particular, we announced today our entrance into a very exciting new market beyond sensing applications that I will talk more about in a bit. Reflecting our commercial momentum, we achieved another strong revenue quarter with continued strong sensor shipments and NRE higher than the previous year. With demand for Aeva's technology continuing to grow, we are progressing on the targets we set this year to scale our manufacturing, and we are working on ramping across our supply chain from our contract manufacturing partners to foundries and component suppliers to support higher volumes to meet demand. We also strengthened our balance sheet in June with a follow-on equity offering that brought our total available liquidity to over $300 million at the end of Q2. We believe this further differentiates and positions Aeva to support our current customers to commercial deployment and expand our commercial traction across the board. I would now like to highlight recent achievements and business developments. First, we are excited to announce the launch of the Aeva optical connectivity business, which is our expansion to a new market that leverages Aeva's high optical power source and silicon photonics technology to enable next-generation AI data centers. Aeva has spent the past 10 years developing and proving out our proprietary laser-based technology for automotive and industrial applications. Because our technology is built on the same foundational technology used in the telecom industry, it uniquely positions Aeva to repurpose it without significant development to pursue the massive data center market opportunity as the major AI compute companies like NVIDIA, AMD, and hyperscalers such as Amazon, Microsoft, and Google transition to optical connectivity. We believe Aeva can provide differentiated solutions to achieve the increasing performance, power efficiency, and scalability requirements for AI data centers. Earlier this year, we published a paper detailing the industry-leading performance of Aeva's high-power optical source technology. Since then, we have moved quickly to meet growing interest from leaders in AI infrastructure in Aeva's capabilities. In just a few short months, we have been working with a number of AI chip companies and optical solution providers to hyperscalers to validate some of these capabilities, and the results so far have been very encouraging. We have been able to demonstrate how Aeva's technology can deliver across stringent key metrics: the required high optical power, low noise, reliability, and power efficiency in a scalable solution to address the rapidly growing bandwidth and power demands of multiple next-generation AI data center architectures, including near-packaged optics and co-packaged optics (CPO). I am thrilled to share today that we have just signed a key joint development agreement with a leading provider of high-speed optical engines to integrate Aeva's technology into a near-packaged optic solution for a major hyperscaler, with planned commercial deployments as soon as feasible and the potential to start initial volumes in 2027 and ramp in 2028. This is a significant validation of the performance, maturity, and scalability of Aeva's technology, and we believe it will advance the business we are launching today, including our other ongoing engagements across the industry. The potential scale for this market is large, and we believe a deployment like this deal, once qualified at the hyperscaler, has the potential to exceed multiple millions of units annually with revenues in the multiple hundreds of millions of dollars per year. To support the significant potential for optical connectivity, we plan to leverage Aeva's existing manufacturing and foundry supply chain for volume production. We have also established a dedicated team with Pradeep Srinivasan, Aeva's VP of Photonics, taking an expanded role as Senior VP of Optical Connectivity. Pradeep has been instrumental in developing and industrializing Aeva's silicon photonics IP, and together with the team will enable Aeva to accelerate the development, deployment, and adoption of optical connectivity. With our technology team and the growing commercial interest, we are very excited about the future of optical connectivity and plan to share more updates on this as we progress. Now moving to automotive. We also made really good progress across our key automotive programs this past quarter. On Daimler Truck, we continue to scale shipments of production-intent Atlas sensors to the OEM for their AV stack validation ahead of series production. Importantly, we have begun manufacturing at our fully automated assembly line at Jabil North America, a major milestone towards Aeva's readiness and ability to scale up manufacturing for automotive and others, including Daimler Truck's production program. In passenger vehicles, we have also progressed on milestones with the delivery of our Atlas Ultra systems and continued integration work for joint stack development with a top-10 European passenger OEM production program, as well as with NVIDIA on the Drive Hyperion platform. Our collaboration with Bendix has also progressed to a critical next stage: Bendix has selected Aeva's 4D LiDAR and perception software and will integrate them into its next-generation ADAS series production system for key added safety functionality for trucks, such as collision mitigation across a broader range of scenarios and nighttime driving. Bendix is the leader in commercial vehicle ADAS. Its current Bendix Fusion, which is a vision- and radar-based system, is the market-leading collision mitigation and active safety solution in North America. It is available on most of the roughly 300,000 Class 8 trucks that are sold annually in North America alone. With Bendix having the goal to make its next-generation ADAS system standard for flagship models of their major OEMs, names like PACCAR or International, this selection not only highlights how the industry continues to adopt LiDAR to improve performance, but it is also a strong validation of the manufacturability and cost-effectiveness of Aeva's solutions for Level 2+ ADAS applications where LiDAR has traditionally not penetrated yet. Now turning to factory automation, we reached an important milestone with another customer launch of an Aeva product this quarter. SICK AG, one of the largest industrial sensor providers globally, commercially launched its first industrial sensor using Aeva's Eve precision sensing system. This is part of our long-term strategic collaboration with SICK to leverage Aeva's advantages, such as immunity to ambient light and sensor-to-sensor interference, to deliver more precise and reliable measurements across a broader range of environments at scale. In addition to SICK, we continue to see strong interest for our precision technology from leaders in manufacturing and factory automation and are working towards converting these engagements to commercial awards in the near term. Separately, in defense, we continue to support Forterra's autonomous ground vehicle (AGV) programs as they transition from time-of-flight to Aeva's 4D LiDAR for long-range and velocity detection, as well as for vehicle positioning and stealth operational capability in GPS-denied environments. This quarter, we also progressed on other opportunities with major defense companies and organizations beyond ground applications, including, for example, aerial autonomy. We look forward to sharing more on these in the coming months. And last but not least, in smart infrastructure, we continue to see good traction with the recently introduced Aeva CityOS, our AI-powered platform for real-time intelligent traffic management. Following our first large-scale deployment in the Atlanta, Georgia area, the city of Fargo, North Dakota, selected us to deploy CityOS to improve roadway safety and traffic management, leveraging 4D LiDAR's ability to operate in inclement weather such as snow, fog, and rain, and in all lighting conditions. Aeva's ITS team is also active with other DOTs and municipalities, and we believe that our differentiated solution will continue to penetrate for additional deployments. So in summary, we achieved a lot this past quarter. Our differentiated technology and balance sheet positioned us to further solidify a leadership position in the industry as we execute on our commercial momentum. Now before I turn the call to Saurabh to walk through our Q2 financial results, I wanted to say a few words on the CFO transition we announced today. Saurabh, who joined Aeva six years ago as our CFO, will be moving on in September to pursue a new opportunity outside the sensing industry. I want to personally thank him for his many contributions to Aeva, including his role in helping to take the company from an early public stage to where we are today, and we wish him all the best in his next role. We have already initiated a search for a permanent successor which will be announced separately once complete. In the interim, Rupesh Maheshwari, our VP Corporate Controller, will step in as interim CFO. Rupesh brings more than 20 years of accounting and finance leadership experience at large and growing technology companies and will work alongside Saurabh to ensure a successful and seamless transition. With that, let me now turn the call over to Saurabh.
Thank you, Soroush, for your kind words and good afternoon, everyone. Now on to Aeva's Q2 results. As demand for our unique technology continues to grow, Aeva's financials also reflect the building momentum of the business, our expansion to an exciting new market opportunity, and the continued financial discipline as we execute on our plan. Revenue was $6.1 million in Q2, driven by continued strong product shipments and contribution from NRE as we delivered sensors and achieved milestones with a growing group of customers. Non-GAAP operating loss was $26 million this quarter, which is close to prior year levels and reflects our target to maintain operating expenses at similar levels to up slightly year over year while continuing to scale the business. Q2 gross cash use, which we define as operating cash flow less capital expenditure, was $31.4 million. In June, we raised gross proceeds of $115 million in a follow-on equity offering. This brought total available liquidity at the end of Q2 to $302.9 million. With this liquidity position and our differentiated technology, we believe Aeva is uniquely positioned in the industry to continue building our momentum including into exciting new markets such as optical connectivity, and meet more of the growing demand for Aeva's technology. And finally, as this will be my last earnings call at Aeva, I wanted to say that it has truly been an honor to serve as Aeva's CFO over the past six years. I am very proud of what we have built as a team and I look forward to following the growing momentum and continued success of the company. Let me now turn it back to Soroush for his closing remarks.
Thank you, Saurabh. In closing, I would like to thank the Aeva team for delivering on our key achievements in Q2. Aeva's unified perception platform continues to gain the trust of a growing list of leaders across multiple markets, and our ability to launch optical connectivity and realize commercial traction so quickly is another indication of the scalability of our technology and the execution of our team. As we continue to expand, we remain laser focused on the objectives we set out at the beginning of this year: achieve milestones on existing programs while adding new wins, scale manufacturing to support increasing demand, and maintain financial discipline as we grow. We have made strong progress in the first half of the year and are well positioned to deliver on these objectives in 2026. And with that, let's now turn to Q&A.
分析師問答
Thank you. To leave the queue at any time, you may press 2. In the interest of time, we do ask that you please limit yourself to one question. We will take our first question from Colin Rusch with Oppenheimer and Company. Your line is open. Please go ahead.
Thanks so much, guys. And, Saurabh, congratulations on the transition. Guys, can you talk about with this development agreement what the key technical hurdles are that you are going to be addressing here over the next 12 to 18 months to really get ready for commercial ramp?
Yeah, Colin. This is Soroush. Happy to answer that. So, obviously, this is a major customer win for us here, and rapid progress, I would call it lightning speed in the past few months for us to get to here. The key focus has been, first, showing that we have the capability and working and validating that with the actual end customers, including the chip companies as well as the hyperscalers that I mentioned on the call. We have been engaged with the key players in this period as well. So key for us is we see a massive opportunity using the same core technology in terms of what we call high-power optical sources, using our laser-based solutions as well as our silicon photonics technology. We have spent the past 10 years developing and really proving it out in automotive, industrial, and those applications. Now we are leveraging this to launch this new business of optical connectivity for integration of the solution into the hyperscaler and data center environment. The feedback has been very encouraging because the validation is there. We have a ton of data on reliability in field testing. Importantly, we have a different approach, which I can get into a little bit more later, but the approach allows us to effectively provide and meet the very strict requirements: providing high optical power, low noise, reliability, power efficiency, and importantly, scalability to really meet what needs to be done. The technology is proven. Obviously, we have to do some development to make sure, for example, with this joint development, that the solution and our high-power sources and photonic solution integrate properly. So there is some joint development and integration work to do on that, and we are doing that as fast as possible because the hyperscaler wants to deploy this immediately. With that, we go into qualification. Once qualified, then this is going to go into scale. The timing here: we are doing these activities and the joint integration in the next few months, then we will do the qualification. The hyperscaler's goal is as early as the second half of 2027 to start the initial deployments, and then from there scale up into production around 2028. So that is the short answer. The potential, obviously, here is massive because this is a top three or four hyperscaler with millions of units that they are deploying.
Can I just pick up on that last comment there? Certainly, that volume of units is substantially more than you would have been able to produce in your other end markets. I am just curious about potential impact on cost structure for you guys in terms of being able to reduce cost and pass that on or drive market share in some of your other applications and the capacity of your partners to help you scale up to those volumes?
Yeah, absolutely. I think this is a crucial point. What we have done with the product here is focused on solutions at the chip level that are very cost effective. When you look at what is out there, a lot of the focus has been on investing heavily in laser sources to achieve high power. To do that, there are certain trade-offs you have to make. The size of the die gets bigger, you align multiple of these to get a number of wavelengths or a lot of power out. If these things get larger, more alignment is needed. All of this impacts yields, which affects cost as well as repeatable scalability of the solution. Our approach is quite different. That allows us to have smaller-size dies and the ability to actually provide the high optical power without having to use as many chips for the solution. We see that allows for much higher yield, much higher repeatability, and reliability as well. So that is one of the key differentiations for us. As we go into millions, the cost structure is already competitive given what I mentioned. Our focus next is on foundries and manufacturing partners to scale the volume. These foundries are capable, and we have secured some of the capacity we already need to deploy in the market and are working to increase that given the massive pull potential with this new deal that we have. We will use economies of scale to drive down cost and help with the next generation of sensing products as well for our other products.
Thank you. And as a quick reminder, if you would like to ask a question, please press 1. Our next question will come from Matthew Paciulli with Canaccord Genuity. Please go ahead. Your line is open.
Hi, guys. Congrats again on a great quarter, and best of luck, Saurabh. Maybe just to continue on to Colin's question. Could you just give us a little detail around what you expect the revenue model to be for the Optical Connectivity business? How should we think about it from an ASP perspective? Some details around that would be appreciated.
Yeah, I am happy to answer that. At a high level, as I mentioned on the call, we just signed this deal and we are very excited about it. The key focus here is we see the potential for ramp starting in 2027, the back half, and then ramping into 2028. The volume forecast from the hyperscaler is quite massive, and whatever supply we can provide we have the opportunity to sell. The expectation is this ramp will start into 2028 with a minimum of millions of units going into production, and we are talking about multiple millions of units. That would translate into multiple hundreds of millions of dollars of revenue annually. Those figures are multiple millions of units annually and multiple hundreds of millions of dollars annually in revenue opportunity for us. The key for us is being able to do that without a ton of significant developments or reconfiguration, which is why we see the ROI as being quite interesting for the company and why we are launching this new business. That should give you a rough sense on ASPs, but I will stop there.
Great. Thank you. And then maybe just to switch gears, you've made a lot of progress with Bendix. Could you give us a little more color around what that program will look like, the content per vehicle, and the ASPs to the extent that you can share?
Sure. Bendix is the market leader for commercial vehicle ADAS technology and we have been working with them for some time. They have advanced into a critical next phase and have selected us to leverage our 4D LiDAR to provide new functionality for Level 2+ ADAS on commercial vehicles. These functionalities are specifically aimed to enable new types of emergency braking, passenger braking, nighttime driving, and so on—basically better functionality for the end customer. Bendix is part of a large tier 1, and it is a market leader in North America for ADAS solutions. They have been shipping in volume for many years and have a majority market share with their Bendix Fusion system, which today is vision and radar based. The opportunities for us are significant. It uses the same exact sensing product, Atlas, our sensor intended for trucking with Daimler. We will be using the same automated line at Jabil and leveraging that capacity and economies of scale to ship that product. In terms of content per vehicle, think of this as an ADAS product somewhere between passenger and commercial vehicle products. It will not be thousands of dollars, but it is also not very small. Part of that is because we are helping augment the system and with the aim to replace not just add a new sensor but also replace other modalities between radar and camera. Bendix is targeting to make this a standard feature for flagship OEM models for the next solution.
We will take our next question from Joe Moore with Morgan Stanley. Please go ahead. Your line is open.
Yes. Thank you. On this optical connectivity, you talked about starting a group around optical connectivity. How much resource do you want to put into this? Is this a major pivot for the company? I know you will continue to invest in your automotive and industrial businesses, but do you need to scale up R&D around this opportunity given the size of it?
Hey, Joe. This is Soroush. This is an important launch and investment for us, but we are really reusing the core components we've already developed in silicon photonics and the high-power sources, and also our manufacturing partners and capabilities we've already established—with foundries and contract manufacturers that do module assembly. We will leverage a lot of the prior work and investments, so we do not need to invest a ton of capital. Importantly, a lot of development resources and time go into making something complete from scratch; that is why we are able to move very quickly. From a resource and investment standpoint, of course we are serious about this. We are going to invest and we are hiring in this area. With Pradeep's promotion and the team's leadership, I have full confidence in our ability to execute and scale up this business. We will have dedicated resources so we maintain focus and priority on existing programs in automotive with the top-10 passenger OEM and Daimler, while also having separate resources working on optical connectivity so we do not distract or lose focus on the other side. So yes, we will invest; it will not be massive amounts, but we will take it seriously.
Great. Thank you. Is there a difference in content between near-packaged optics and co-packaged optics as you migrate the technologies?
That is a good question, Joe. The way we think about this is we are forming this team and group to generate products and solutions for optical connectivity in general. This will be a combination of optical sources, including on-chip integrated laser sources, as well as, over time, external laser small form-factor pluggables (ELSFPs). Those represent massive opportunities with different levels of ASPs, and we are excited about that. Both near-packaged optics and co-packaged optics actually use the same core technology and source that we have developed, with some additional integration that we need to do on the back end. That is why we are able to use the same IP. With this deal, the initial intention is near-packaged optics, but we are also working on other things for co-packaged optics as well.
We will take our next question from Richard Shannon with Craig-Hallum. Please go ahead. Your line is open.
Hi, gentlemen. First time for me, but very exciting series of updates. Can I just ask you in the defense space, do you see that as being a market for just newly developed platforms, or do you see some sort of larger retrofit opportunity for current weapons platforms or vehicle platforms?
Happy to answer that. We see both. We are working already with companies like Forterra that have ground vehicles on the ground. Some of those are retrofits, which is basically taking a vehicle and retrofitting them with an autonomy kit that includes three or four Aeva LiDARs for 360-degree coverage. We are making a lot of good progress and are excited about that. Vehicles are already being deployed to the military and are starting to be commissioned and used, which is pretty fast for the defense space given we just announced this earlier in the year. That has resulted from advantages we have: velocity sensing, long range, the ability to operate in GPS-denied environments, and no interference with night vision goggles compared to what they have used prior to us. We are seeing additional traction in other areas, and I think both ground and aerial space opportunities exist. The team is working on these, and I am excited about the potential for more meaningful opportunities that can piggyback on this. We are seeing demand from defense grow, and as we lock those down, we will share more in the coming months.
That is helpful color. On trucking, I understand Class 8 is the biggest opportunity, but we're seeing autonomy progress in Class 6 and 7. Do you think Aeva has a place in those markets as well, and does that opportunity look different than Class 8?
Yes, we are engaged in multiple programs across different vehicle classes. The way I would think about it is a common platform used for autonomy regardless of vehicle size. Once a product is qualified—especially with large players that set industry standards, like Daimler Truck—that allows us to get other programs. The team is working with others for on-road and off-road or yard environments. These do not all involve the same class of vehicles. Initially our focus was long-range sensing and velocity, but as we have proven the automotive capability and product, we are seeing more demand from other vehicle classes and applications, which will plug into various programs.
We will take our next question from Richard Shannon with Craig-Hallum. Please go ahead. Your line is open.
Great, guys. Thanks for letting me ask a couple of questions. First off, Saurabh, congratulations on your new job. We are going to miss you. I jumped on the call late, so apologies if I am repeating prior questions, but the JDA you signed in optical connectivity is very interesting. Since you phrased this as a JDA, what needs to happen for you to be fully qualified and to hit the time frames you are expecting to deploy in the second half of 2027? Also, is this solution expected to be an ELSFP format or something more component in nature?
Richard, happy to expand on that. With the optical connectivity solution and this deal, the key effort is integrating our core optical technology, including high-power optical sources and photonics technology, into the optical engine that will go to the hyperscaler. The capabilities are around high data transfer for near-packaged optics, which is the first application where we see a lot of volume. The effort is working together with the chip companies and the integrated solution and the optical engine, then qualifying that with the hyperscaler in the next few months. The goal is to have initial release and initial volumes in the second half of 2027 and move into production ramp in 2028. The team is kicking off efforts now, and importantly, we will be using the core technology and the chips we have without making a ton of changes, which is crucial to our ability to move fast. That is part of the JDA: integrating in that way to make this happen.
A follow-on question on SICK: they had a first product launch. Can you give a sense of the overall contribution you expect with them over time? I assume more product launches will follow. How should we think about SICK's contribution medium to long term?
SICK is a strategic partner and collaboration over the past several years. We are pleased this is resulting in product launches and portfolios. SICK released the first product using our Eve sensor in a 1D configuration for distance sensing and displacement sensing, and feedback has been promising from end customers. We expect over the next 12 to 18 months to scale additional volumes and launch follow-on products together. We are also working on other products together that can replace devices like encoders and other speed measurement devices leveraging the same sensor, Eve, in different configurations such as measuring velocity (1V). That is another follow-on we are working on as we ramp 1D. Separately, we are engaged with others in the space showing interest, and that should help create a meaningful segment for us. We will share more as we can in the near term.
This concludes our question and answer session in today's meeting. We appreciate your time and participation. You may now disconnect.