Arbe Robotics Ltd.(ARBEW)Q1 2026 法說會逐字稿

Good day, and welcome to the Arbe Robotics First Quarter 2026 Results Conference Call. Please note today's event is being recorded. I would now like to turn the conference over to Kenny Green of EK Investor Relations. Mr. Green, please go ahead.

Thank you. Good day to all of you, and welcome to Arbe's conference call to discuss results of the first quarter of 2026. Before we begin, I would like to remind our listeners that certain information provided on this call may contain forward-looking statements, and the safe harbor statement outlined in today's earnings release also pertains to this call. If you have not received a copy of the release, please view it in the Investor Relations section of the company's website. Today, we are joined by Ram Machness, Arbe's Chief Executive Officer, who will begin with a business and strategic update; Kobi Marenko, President and Co-Founder, who will provide perspective on Arbe's long-term direction; and Karine Pinto-Flomenboim, Chief Financial Officer, who will review the financials. Following management's formal remarks, we will open the call for the question-and-answer session. And with that, I'd like to hand the call over to Ram. Ram, please go ahead.

Thank you, Kenny. Good morning, everyone, and thank you for joining us to review Arbe's first quarter 2026 results. I stepped into the CEO role in April, and I see significant potential ahead for Arbe. During the first quarter, we made strong progress and accelerated on the strategy we set out at the start of the year. Today, I will focus on the main areas where we are making progress. Arbe is growing from a chipset-focused automotive company into a supplier of complete radar solutions across automotive and increasingly into adjacent markets. This quarter, we hit several commercial milestones that reflect this transition. We began shipping chips into China through our Tier 1 Hirain. We received orders from robotaxi customers. We also received orders for data collection programs with global automakers and leading mobility players. I will now go deeper into each of these areas, starting with automotive. Let's start with China, the fastest-moving automotive market in the world. During the quarter, we shipped the initial batch of chipsets to Hirain. These chipsets support the production of 48x48 channel radar, a project that we announced in December in which Hirain is developing a Level 4 autonomous vehicle solution for a Chinese automaker. In parallel, Hirain is developing new radar also based on our chipsets with 24x12 channels designed as a lower-cost system that can be used by a wider range of vehicles. This configuration complements its existing high-end 48x48 radar system based on our chipset. With our technology, Hirain is becoming a key radar platform player in China. It can offer OEMs a path from high-end radar systems that are available today in China to full 2K ultra-high-resolution performance. China sold 34.4 million vehicles in 2025 through a local Tier 1. Arbe has gained direct access to one of the world's largest and fastest-moving markets for Level 2+ and Level 3 deployments. Turning to robotaxis. We received orders for our Phoenix radar system from global robotaxi companies. These systems support Level 4 autonomy and full 360-degree sensing. This clearly shows that our high-resolution radar delivers the performance and coverage required for Level 4. More broadly, we see strong and growing interest for our automotive chipsets. We are taking part in data collection programs with global automakers and leading mobility players. We have progressed into advanced collection processes with specific Chinese and European automakers. These evaluations focus more and more on the specific use cases where cameras and LiDAR fall short and where imaging radar becomes the key sensor. This strengthens industry interest in our radar technology as a core sensing platform for autonomous driving programs. This interest comes alongside an important shift in the automotive market. Recently, several leading OEMs have revisited their Level 3 programs. We don't see this as a rejection of eyes-off autonomy. We see it as a reset, going back to the right basics of autonomy. The first-generation Level 3 systems had clear limitations. They were geofenced. Some were limited in speed, some limited to good weather only, and many never reached commercial deployment. Besides the maturity of the algorithm, we believe that one other main reason for that is related to the performance of the sensor and specifically imaging radar not providing the needed performance. Based on our discussions with many OEMs, automakers are now actively looking at the next-generation eyes-off platforms and looking for sensing that can support Level 3 use cases. This is exactly the gap our high-resolution radar is designed to fill. We also wanted to understand what drivers want from eyes-off autonomy. A survey of 1,000 people across the U.S., Europe and Asia came with a clear message showing that drivers are willing to pay, willing to switch their car and even their brand for fully operational Level 3 and Level 4 autonomous driving. Consumers are ready for eyes-off autonomy, but only when it is safe, smooth and reliable. That is exactly the capability that Arbe's ultra-high-resolution radar enables. You are invited to our website to learn more about this survey. Before moving beyond automotive, let me briefly touch on physical AI because it's becoming central to where our technology plays. We are seeing how the revolution in AI and machine learning is introducing a new generation of algorithms. Those algorithms, and specifically vision-language-action models, are transforming how intelligent systems sense, understand and interact with the physical world. In automotive, it drives breakthrough progress for Level 3 and Level 4 capabilities. And beyond automotive, it opens a wider opportunity across robotics, logistics and other autonomous systems. These physical AI systems are only as good as the real-world data they receive, especially in safety-critical environments such as vehicles. This is where our radar fits. Our radar provides dense sensing, long range, low latency and consistent performance in all weather. This is exactly the type of machine-understandable input that AI-driven autonomy needs. During the quarter, NVIDIA announced it expanded the global DRIVE Hyperion ecosystem to accelerate the road to full autonomy and cited Arbe as part of its platform. Our ongoing work with NVIDIA on radar-based free space mapping and AI-driven automotive capabilities is one example of how we are placing Arbe at the center of this major shift. Now let's look beyond automotive, where our technology is opening meaningful new opportunities. The first important development is that Arbe has begun selling complete end-to-end radar systems in addition to selling chipsets. This allows us to address markets where customers want a full radar solution and where sales cycles are much shorter. We already started shipping these systems to players across defense, homeland security, transportation applications, perimeter security, physical AI and several other applications. To support this, we have also set up dedicated production lines to scale system manufacturing. Taken together, these milestones show that our leadership in automotive radar can extend into perception-critical markets. We are expanding our reach, broadening our opportunities and significantly increasing our total addressable market.

Thank you, Ram, and good morning, everyone. Stepping into the President's role, my focus is on Arbe's long-term strategy, our strategic partnerships and guiding investments that accelerate Arbe's next phase of growth and commercialization. During the first quarter, we strengthened our balance sheet. We closed an underwritten registered direct offering that raised $18.5 million in gross proceeds. This gave us the financial flexibility to invest in Arbe's current phase of fast growth. What gives me a great deal of confidence is what this quarter demonstrates. Arbe's technology platform is broader and can focus on multiple end markets as well as automotive. The same ultra-high-resolution radar that we built for automotive can be, and is now beginning to be, adopted much more broadly. We see increased interest and potential and in some cases initial system sales in homeland security and smart infrastructure, as Ram already mentioned. And finally, the rise of physical AI, together with OEM demand for better autonomy solutions, plays directly to our strengths: high-definition, long-range, all-condition environment sensing. Ram has recently stepped into the CEO role with strong execution focus, deep product knowledge and the commercial understanding that is needed. This will take Arbe from where we are today to a scaled radar systems company across many markets. I have worked closely with Ram over many years, and I have full confidence in his abilities as well as those of the entire team to bring Arbe to the next phase. I am more excited than ever about Arbe's potential over the coming quarters and years. With that, let me turn the call over to Karine to review the financials.

Thank you, Kobi, and hello, everyone. Let me review our financial results for the first quarter of 2026 in more detail. Revenue for the first quarter of 2026 totaled $0.5 million compared to $0.4 million in Q1 2025. Backlog as of March 31, 2026, is $1 million. Gross profit for Q1 2026 was a negative $0.1 million compared to a negative $0.3 million in the same period last year. Turning to operating expenses. Total operating expenses for Q1 2026 were $11.2 million, down from $13.1 million in Q1 of 2025. The decrease in operating expenses was primarily driven by lower share-based compensation expenses reflecting earlier grants that are now fully vested, along with last year's award being structured with half in cash and half in equity. The decrease was also related to the Q1 2025 tape-out expenses, which are reduced as we advance toward productization. This decrease in operating expenses was partially offset by the weakening of the U.S. dollar mainly against the Israeli shekel and, to a lesser extent, by labor-based provisions and a merit increase. Operating loss for the first quarter of 2026 was $11.3 million compared to an operating loss of $13.4 million in the first quarter of 2025. As previously announced, during the quarter, we implemented cost reduction measures that are expected to reduce our ongoing operating expenses by approximately 15% with the full impact expected to take effect starting during Q2. Adjusted EBITDA, a non-GAAP measurement, which excludes expenses for noncash share-based compensation and for nonrecurring items, was a loss of $9.9 million in Q1 of 2026 compared to a loss of $9.7 million in the first quarter of 2025. We believe that this non-GAAP measurement is important in management's evaluation of our use of cash and in planning and evaluating our cash requirements for the coming period. Net loss for the first quarter of 2026 was $9.4 million compared to a net loss of $13.8 million in the first quarter of 2025. Net loss in Q1 2026 included $1.9 million in financial income compared to $0.5 million of financial expenses in Q1 of 2025. As of March 31, 2026, Arbe held $53.6 million in cash and cash equivalents and short-term bank deposits. Turning to our outlook. We are reaffirming the 2026 guidance we provided in February 2026. We continue to expect full year revenue in the range of $4 million to $6 million and an adjusted EBITDA loss in the range of $28 million to $31 million. Now we will be happy to take your questions. Operator?

And our first question today comes from George Gianarikas with Canaccord Genuity.

I was wondering if you can go into a little bit more detail around the traction you're seeing in defense-related applications, drones, et cetera.

So thank you, George, for the question. Basically, around defense, there are a few applications where our radar can function. The first is not so different from our automotive application. So it's autonomous vehicles, whether they are supply trucks or even armored vehicles. All of them want to drive safely, especially off-road in environments that are full of dust, rain, fog and other conditions where radar is the best sensor for that. For that, we already have a client, the U.S. Army, that bought from us hundreds of units. We believe that we will receive another order for that as well. The second application is perimeter defense, whether against people trying to enter an area or against drones that can attack or drop supplies or contraband into those areas. We definitely see this as an application that our radar can supply. On top of those, there are a few other smaller applications where radar—especially our radar—can help. A year ago, we were not even considering those markets; today, we are focusing on them. We are trying to do business development with the leading players, including the big names in the defense industry as well as new players. This market is also going through a real transformation. The players that were there 10 years ago, like Lockheed Martin, are changing, and there are new emerging players taking market share. We are trying to work with both the established players and the new players and provide our solution.

And maybe just as a follow-up, you obviously mentioned the NVIDIA reference design. I'm wondering what the conversations have been like, how much additional traction? Or how has that changed any of the conversations you're having with OEMs and other partners in terms of deploying your imaging radar alongside NVIDIA?

Yes. I think that this question refers a lot to who are the players that can potentially bring the AV stack—the autonomous vehicle software stack—to the market because eventually, autonomous driving is all about software, algorithms and machine learning. NVIDIA is the top player in the market today on that front. So this is very important from the perspective of OEMs, whether they have selected NVIDIA or not. I think they are all watching what's going on in the industry and who has the potential to get the AV stack to the finish line and really provide a reliable and sustainable solution to consumers. Eventually, you want to get into your car in the morning, drive safely to work and let the car drive for you so you can do something else and buy your time back by using the autonomous driving feature. There is a race right now. Many players are working toward that. NVIDIA is probably the strongest player. There are other players as well, and some OEMs are trying to build their own homegrown stack. Everybody is looking at what others are doing, and there is definitely a clear race on that. I think it's important to see the progress we've seen recently at the algorithm level and the breakthroughs in deep learning and AI. This progress also applies to autonomous vehicle software, and we'll see this coming into the market pretty soon, starting with China and moving forward to the Western world.

Our next question comes from Casey Ryan at AmerX. There are other players as well, and some OEMs are trying to build their own homegrown stack. Everybody is looking at what others are doing, and there is definitely a clear race on that. I think it's important to see the progress we've seen recently at the algorithm level and the breakthroughs in deep learning and AI. This progress also applies to autonomous vehicle software, and we'll see this coming into the market pretty soon, starting with China and moving forward to the Western world.

I had a couple of questions. Would you care to describe— it sounds like you have multiple robotaxi opportunities. Is there a regional tilt to those? Or are they global, meaning North America, Europe, China? Or are they focused in one region?

So the opportunities that we see right now are coming from North America, and we see opportunities in China as well. So the main sales are definitely North America and China.

Okay. That's terrific. And then when you talk about supplying the full radar system versus a chipset, can you talk about maybe what the revenue differential or pricing differential is between those two as we think about unit growth in both products?

Yes. There are several factors that influence the change and significantly increase the average unit price (AUP) when we talk specifically about non-automotive products. One is the fact that we are selling a complete system, not just the chipset, and that obviously drives the average unit price up. Also, the volumes are much lower. When we work on a Level 3 project with an OEM, the volumes are really high, and the negotiation power and economies of scale drive the AUP down. For robotaxis, robot trucks, commercial vehicles and defense applications, the volumes are significantly lower, but the AUP is significantly higher because there is no economy of scale to the level you have in the automotive industry. So we see a significantly higher AUP in the business of complete radars compared to automotive chipsets.

Okay. And then just sort of a small thing, but do you think ultimately, long term, the gross margins for both product lines are similar? Or do you think the complete radar system might be lower long term?

Usually, the gross margins on mid-scale system projects are a bit lower than when you talk about chipsets. That's expected because when we charge for chipsets, we charge for the differentiation we bring. When you charge for systems, you also charge for mechanics, chassis and other components. So percentage-wise, the gross margin might be lower, but the dollar profit per system is significantly higher. So percentage-wise it's lower; in absolute dollar value it's higher.

And of course, worth mentioning the volumes, which are significantly different between systems and chips.

Yes. Understood. That's terrific. And then just one last question for me, maybe Karine. On the backlog, are we to take that as being kind of a 12-month backlog number or shorter or longer?

No, this is 12 months, correct.

Okay. Terrific. And like maybe to clarify for myself, that $1 million, it's fair to think would be consumed in '26? Or are you saying it's kind of 12 months from the end of Q1, basically?

It's 12 months from the end of Q1, but I assume all of this will be in 2026.

And our next question today comes from Matthew Galinko at Maxim Group.

I'm curious what the cost was to stand up the full production line for radar. And given your current capacity there, do you expect that's sufficient to meet what you have in the pipeline for defense and AI applications? Or do you expect you'll need to add capacity over time?

We built our production line to supply the current capacity that is needed, which we assume will be hundreds of systems per month, thousands of systems per year. The investment is not dramatic; it is on the low side of the hundreds of thousands of dollars, not more than that. We have the ability to increase this capacity as time goes by and as demand grows during the coming quarters.

We also started discussions with some contract manufacturers for expanding this production line further down the road.

And that does conclude our question-and-answer session for today. I'd like to turn the conference back over to Mr. Machness, CEO, for any closing remarks.

On behalf of the entire management team at Arbe, I would like to thank you, our shareholders, for your continued interest and long-term support of our business. This quarter marked the start of a new chapter for Arbe, a chapter focused on turning our technology leadership into commercial scale across many markets. We have our first system-level order, growing robotaxi traction and expanding Chinese market opportunity through Hirain and a clear position at the center of the next wave of eyes-off autonomy and physical AI. Based on all of this, we believe Arbe is well placed to deliver meaningful value. We look forward to updating you on our progress next quarter, and I personally look forward to taking on the CEO role to bring Arbe to the next level in its development. And with that, we end our call. Have a good day.

Thank you. Today's conference has now concluded, and we thank you all for attending today's presentation. You may now disconnect your lines, and have a wonderful day.