Arbe Robotics Ltd. Q2 FY2025 earnings call

• Arbe received a strategic order for its chipsets in large-scale data collection projects, with chipsets shipped for fleets of over 100 vehicles by leading automotive manufacturers.
• There's growing industry recognition of high-end imaging radar for safe hands-free driving. Arbe is in imaging radar selection programs with leading OEMs, and Magna and HiRain are competing with Arbe's chips on these opportunities.
• Collaboration with NVIDIA is helping with customer engagements. Non-automotive applications are scaling up, with Sensrad delivering radars for defense applications and a new support agreement signed with Sensrad.
• Progress in bid processes with target OEMs, with some advancing to final stages and aiming for 4 OEM bid wins in the coming year. Revenues expected to begin in 2027 and ramp up in 2028.
• HiRain launched the LRR615 long-range imaging radar system powered by Arbe's chipset. China's new ADAS testing regulation creates a market opportunity for Arbe's high-definition radar technology.

Revenue for the second quarter of 2025 totaled $0.3 million compared to $0.4 million in Q2 of 2024. As of June 30, 2025, backlog stood at $0.5 million. Gross profit for Q2 2025 was negative $0.2 million compared to negative $0.04 million. Total operating expenses for Q2 2025 were $11.3 million, down from $11.6 million in Q2 2024. Operating loss for the second quarter of 2025 was $11.5 million compared to $11.6 million loss in the second quarter of 2024. Adjusted EBITDA was a loss of $8.9 million in Q2 of 2025 compared to a loss of $7.5 million in the second quarter of 2024. Net loss in the second quarter of 2025 was $10.2 million compared to a net loss of $11.7 million in the second quarter of 2024. As of June 30, 2025, Arbe held $26.3 million in cash and cash equivalents and short-term bank deposits as well as $35.7 million in long-term bank deposits.

• Arbe's leading radar technology remains a priority, with decision timelines extended but continuing to engage with industry leaders. Goal to pursue 4 design-ins with automakers in the coming year.
• Reiterates 2025 financial expectation: annual revenue weighted towards end of year in range of $2 million to $5 million.
• Adjusted EBITDA for 2025 expected to be in range of $29 million loss to $35 million loss.

Q: I'd like to ask first about the 4 design wins that you expect in the coming year. Curious as to how that number has changed at all over the last few months since we last spoke. And what the competitive set looks like? It sounds like you feel like that has tightened a little bit. And if you could share any details around how you feel that set has changed as well.

A: Yes. So I think that overall, the amount of OEMs that we are working with them towards meaning didn't change. What really changed is the environment and the understanding of every OEM that's basically starting to develop a real, let's call it, L3, hands-free, eyes-off driving that they understand that current imaging radar, low-end imaging radar cannot solve the problem. And with the high level of channels radar that has at least more than 1,000 channels, which basically is our chips that support it and the chipset of Mobileye support it are the only candidate that can really solve the problem for L3. And basically, what we saw is that even in programs that they were considering to take an off-the-shelf imaging radar as an option, they decided to focus and narrow the competition to this, I would say, two horse race between the two companies. The only two companies today that has a real imaging radar, high-end imaging radar that can solve the problem for L3, as I mentioned. The other thing that we are seeing is that when the OEMs beginning the data collection with our radar, of course, with Magna's radar based on our chipset or with HiRain's radar based on our chipset, they see that for a bit of a lower end application, what is called L2+, L2++, which is only hands-free driving, imaging radar can actually solve the problem, and they are looking to expand the selection also to those kinds of applications. So I think overall, I don't see today the customers in the market that will say that imaging -- high-end imaging radar is not marked for a hands-free, eyes-off applications. The fact that there is very few amounts of competitors in this market, I think, it puts us in a very good position to be leading in this market.

Q: Is it possible for one OEM to pick multiple vendors or the imaging radar?

A: Not really, not really. I think that, first of all, the data collection because we are today in a world of full AI stack, you need to have a data collection. In order to have an effective data collection, we are talking about millions, not even millions, millions of kilometers of driving, even miles, not just kilometers. So you need 10 or even 100 cars that driving for a year to collect the data and to train the algorithms. So to go with two radars on the same application makes no sense. It's possible to select a different radar for L3 stack and a different radar for L2++ stack. But I don't see an OEM selecting two radars for L3.

Q: You mentioned in your release potential for non-automotive growth. You listed a couple of applications, including defense sector and smart infrastructure. Can you sort of expand upon that a little bit, if possible, what other non-automotive applications are you seeing potential traction with over the next few years?

A: Yes. So first of all, defense, I think, is a growing sector in the last 2 or 3 years. Of course, it's coming from the different side of it from detection -- from border protection to autonomous trucks that move supply to the troops or so on. So detection of drones attacking and so on. So there is, I think, a strong demand on the defense side for a low priced radar. It's not the monsters of the big industries are building today, radar that cost millions of dollars. There is a need for radar in $1,000, $2,000, $3,000. Of course, it's not in the volume of automotive, but the margins are better. Smart infrastructure, as you mentioned, is good. There is a few other verticals that we're already seeing the traction that we cannot yet announce. We believe that we will have also -- we will be able to announce things that are growing in September or October after the summer vacation. But we feel that the non-automotive market is growing right now better than we expected, and we will be able to see the nice revenues shorter than in automotive. Of course, it's not in the volume of automotive, but it's still -- we see it now as a strategic leg to the company.