Arbe Robotics Ltd (ARBE) 2022 Q3 法說會逐字稿

Thank you, everyone, for joining us today. Welcome to Arbe's Third Quarter 2022 Financial Results Webcast. Today, we are joined by Kobi Marenko, Arbe's CEO, who will begin with a business update, followed by Karine Pinto-Flomenboim, CFO, who will review the financials. Then Thilo Koslowski, a member of Arbe's Board of Directors, will discuss the state of the automotive industry in light of the changing market conditions. Finally, we will open up to our listeners for the question-and-answer session.

With that, I'd like to turn it over to Kobi Marenko. Kobi, please go ahead.

Thank you. Good morning, everyone, and thank you for joining us. Please take a minute to review the forward-looking statement. In the third quarter, we made significant progress with our Tier 1 partners. In fact, we are excited to share recent news. We have received a mass production preliminary order from HiRain for the next year, which officially marks the beginning of the next step of Arbe's journey into the mass production phase. I will go over this in more details shortly. Today, we would like to share with you a deeper understanding of Arbe's ecosystem and our strategy of focusing on Tier 1 is accelerating our progress.

As many of you are aware, market conditions across the entire automotive industry are changing, causing some delays in customer launch plans. From the beginning, we made a decision to focus on driver assistance and safety while keeping a realistic vision for the future of full autonomous driving, which we believe won't be commercially available before the end of the decade. To improve our resilience in these market conditions, we are reducing our costs, making sure that we are prepared to adapt to the longer time frames of Level 2+ and Level 3 ramp-up.

In the automotive industry, the Tier 1 suppliers do the heavy lifting of radar system development as well as tech integration for OEMs. The combination of Arbe's breakthrough imaging radar chipset with Tier 1 innovative technology results in advanced radar system that we believe will provide unmatched safety to the industry. I want to share with you the perspective of some of our Tier 1 partners in their own world.

After 5 years of extensive research among 4D radar chips and developers across the globe, Veoneer, a world leader in automotive safety, choose to partner with Arbe. The combination of Arbe's patented imaging radar chipset and Veoneer's patented waveguide technology will push the boundaries of perception performance. We are honored to have been selected by Veoneer, who intend to provide the automotive industry with revolutionary radar-based solution, ensuring unmatched safety Level 2+ and higher autonomy.

Let's hear what Chris Van Dan Elzen, the Executive Vice President of the Radar Product Area, had to say about Veoneer's expectations for imaging radar, the role in the industry and the added value of Arbe's chipset.

My name is Chris Van den Elzen, and I'm the Executive Vice President of the Radar Product Area. Veoneer is a world leader in automotive safety. We participate in market segments of active safety and restraint control systems as a Tier 1 hardware supplier and systems integrator. We were founded in 2018 as a spin-off from Autoliv, and we have close to 70 years of automotive safety development. The radar portion of Veoneer dates back into the early 1990s. We were one of the first to launch radar-adaptive cruise control in 1999 with Mercedes and have built the product line up to having 16 customers around the world, in every segment of radar from corners to front looking, rear looking.

Veoneer is looking at this new segment that's growing in the radar world around imaging radars. And we have been tech scouting this space for the last 5 years looking at new technologies and start-ups that are bringing innovations in this space. And after researching multiple companies and multiple approaches to the technology, Veoneer chose the Arbe chipset because of a couple of main points. One is the strength and the power of the capability of this chipset to do unique use cases like trying to get through a toll booth, determining if it's occupied or the arm is down. In addition to that, the business model around being able to use their chips and create a unique radar for Veoneer really fits within our business model of being a Tier 1 and bringing that value to the OEM.

In terms of the imaging radar market, we see this as an expanding new space. Customers today are implementing 3 to 5 radars per car in a base case just to meet the 5-star program. Going forward, to be able to meet the use cases of having the vehicle do more and more levels of autonomy when you push that cruise control button, we see the need for imaging radars looking at great distances with great accuracies. The number of customers that are already asking for information about this leading up to designs that will turn into a request for quotation is numerous. And it's around the world. It's not one location or one segment of automotive. We're seeing this across the board.

Veoneer has just signed an agreement with Arbe to expand into this new and growing market. We look forward to our time together assessing our space and our portion of this fantastic opportunity.

In a separate statement, Chris added, Veoneer is now passing 50 million radars produced, and we see the overall market growing to roughly $250 million per year by the end of the decade. As this market grows overall and the imaging radar segment of the market rapidly grows with it, we are looking forward to the journey together with Arbe.

High-definition radar is an important part of advanced ADAS. Valeo, the world's leading ADAS Tier 1 player, has also selected Arbe's chipset for its radar systems. We are very excited to be partnered with such technological leader in areas that are at the heart of the transformation of the automotive industry and sustainable mobility across the globe. Valeo is a leader in the LiDAR market, adding imaging radar to its offering, contributes to comprehensive sensor suite, which is highly important for OEMs.

Martin Mandry, the Vice President of the ranging sensor product line of Valeo, shares a few thoughts about the opportunity for imaging radars and how Valeo's solution with Arbe's chipset stands out.

My name is Martin Mandry. I'm working for Valeo. I'm Vice President for the product line. Valeo is a global automotive supplier, almost 100 years ago. Valeo is operating in the mega trends of the automotive industry, electrification, ADAS acceleration, lighting everywhere and material experience. Within ADAS, there is sensors required, and we are providing these sensors as a system. The radars that we're providing for our customers are fulfilling different needs of our customers, depending on their requirements for automization. We provide side sensors, corner sensors and front sensors to our customers. The radar sensors fulfill requirements in the near field as well as in the far distance. These sensors work in a system together with the other sensors that Valeo provides like cameras, ultrasonic sensors, LiDAR.

Valeo operates globally and has sales of EUR 17.3 billion in 2023, 103,000 employees and operates in 31 countries. Valeo has shipped more than 20 million radar centers to date and has 15 years of innovation in radar technology.

We believe that Arbe has a specific innovative product with the chipset that we are using. The number of radars per car is significantly increasing, and radars play an important role in achieving functionalities that are not existing today but are in demand of tomorrow. Our ambition is to achieve 15% of global corner radar market share in 2030, to be the #1 in HD radar markets by 2032 through partnership with HD radar chipset provider. Arbe and Valeo in a combination can position uniquely in the market and achieve great success together.

Valeo is a great Tier 1 partner, and we are honored to provide the chipset foundation for the radar solution. It is well known that China has the largest vehicle market in the world and is leading the way for an autonomous vehicle future. Another key Tier 1 relationship with Weifu Group is helping us make great achievements within the critical market. Weifu is developing radar system based on Arbe's technology and OEM requirements and has a radar manufacturing plant in China. This collaboration focused on mass market production, safety compliance and providing customized 4D imaging radar solution to automotive OEMs and for autonomous vehicles, trucks, commercial vehicles and traffic applications.

We are pleased and honored to present the President of Weifu Group, Mr. Xu Yunfeng, describing this outlook on the market and the value of our joint solution.

I'm Xu Yunfeng, President of Weifu Group. Weifu Group is founded in 1958. We are a famous Tier 1 supplier in Chinese auto market, 1 of top 30 China auto parts industry companies and also listed in the top 100 available public companies in the China member stock market. We now have 21 subsidiaries, 2 JVs with nearly 8,000 employees all over the world. The group is a well profitable public company. Our revenue has reached CNY 1.36 billion in 2021.

From the beginning of Weifu foundation, we have been adhering to continuously technology innovation and product upgrading. The group has 4 major business segments, which are covering now power, putting hydrogen energy, [intelligence] and electrification, industrial and others, forming a competitive (inaudible) of auto products. We believe our next-generation imaging radars designed with Arbe's chipset, which uses 48 transmit and 48 receivers to grade 2304 virtual channel line for digital (inaudible) instead of using unreliable figures will deliver unparalleled LiDAR performance by leveling thousands of virtual transmitting and receiving channels, providing customized for the imaging LiDAR solution to passenger and commercial vehicles as well as traffic applications at a competitive price.

(inaudible) is one curve of our new business strategies. If we need to build up our groundbreaking LiDAR system, we need an advanced and flexible solution. After a lot of market research, we noticed that Arbe has the most advanced solutions. That's why we choose to work with Arbe. Last year, there were 26 billion vehicles sold in China. We estimate there will be 33 billion vehicles manufactured and equipped with (inaudible) in 2030. Almost 40% of the new vehicles will be equipped with LiDARs, of which 20% will use next-generation imaging LiDAR. We believe with the cooperation of Arbe, our solution will assume a leadership position at 15% to 20% share in the market in the future.

Relating to Weifu's President statement, we believe that our relationship with Weifu will enable Weifu to provide unparalleled level of safety to vehicles and become one of the leading next-generation radar providers in the Chinese market. The last key partner I will introduce today, HiRain Technologies, is a leading Chinese ADAS Tier 1 supplier. As I mentioned earlier, HiRain has placed a preliminary order for 340,000 chipset for Q3 2023 until the end of 2024, which marks the beginning of Arbe's mass production phase. In Q3, we announced that HiRain was selected by the Port of Rizhao in Shandong province to provide perception radars based on Arbe's chipset. This announcement is significant because trucks require the highest standard of safety and does have the biggest need for an advanced sensor solution due to their collision record and the high risk associated with the size of commercial vehicles.

As we reported in Q2, HiRain announced that it is undertaking major OEMs and autonomous driving projects with our radar solution, and it has projected that it will reach mass production next year. Our connection with each and every Tier 1 partner broadens our outreach and accelerates our path to market. We thank you all of our Tier 1 partners for choosing and trusting Arbe and for working hard to secure customer wins.

As we mentioned last quarter, our Lynx radar is the industry's first surround imaging radar. Lynx addresses a significant market need for 360 degrees long-range, high-resolution sensing at an affordable price. During the third quarter, Lynx won the AutoSens Award for Hardware Development of the Year. We view this award as strong market recognition in Arbe's technology, and we are confident that this will enhance our market position.

As we look to the fourth quarter, we are proactively adjusting to changes in the market, and we believe we are prepared for any global uncertainties. Most importantly, we believe that our focus on ADAS and safety will prove to be the right strategy, and our relationships with Tier 1s will drive success.

Now I'd like to turn it over to our CFO, Karine, to go over the financials.

Thank you, Kobi, and hello, everyone. Let me review our financial results for the third quarter of 2022 in more detail. Total revenue in the third quarter was $1.3 million compared to $0.6 million in the third quarter of 2021. Backlog as of September 30, 2022, was $0.3 million. This does not include the recent HiRain preliminary order. Gross margin in Q3 2022 was 72.5% compared to 30.3% in the same period in 2021. The gross margin increase was primarily related to economy of scale, revenue mix and lower cost per unit as we progress toward production.

Moving on to expenses. In Q3 2022, we reported a total operating expense of $11.8 million, an increase from $8.5 million in the third quarter of 2021. The increase in operating expense was primarily driven by noncash share-based compensation expenses, labor cost increase and, to a lesser extent, expenses associated with Arbe being a publicly-traded corporation, partially offset by a decrease in research and development material expenses.

Net loss in the third quarter of 2022 was $9.9 million, which included $1 million of financial income, compared to a net loss of $13.3 million in the third quarter of 2021, which included $5 million of financial expenses. Q3 2022 financial income resulted from interest deposits and favorable exchange rate revaluation, partially offset by warrants revaluation expenses.

Looking at adjusted EBITDA in Q3 of 2022, a non-GAAP measurement, which excludes expenses for noncash share-based compensation and for nonrecurring items, was a loss of $8.4 million compared to a loss of $8 million in the third quarter of 2021.

Moving to our balance sheet. As of September 30, 2022, Arbe had $63.2 million in cash and cash equivalents with no debt.

With respect to our guidance for 2022, we would like to update our forecast based on recent market changes. Revenue is expected to be in the range of $4 million to $7 million. At this stage, our revenue are based mainly on sample sales that can shift between quarters. Despite this revenue reduction, adjusted EBITDA is expected to remain in the range of $34 million loss to $38 million loss.

As Kobi said, we are actively taking certain measures to adjust to market changes, adjusting the time line of production, reducing costs and keeping our focus on ADAS technologies with our Tier 1 partners. We believe that our strong balance sheet and adjusted cost structure will support our progress until market conditions have stabilized.

Now please join us for a chat on the state of the automotive industry in light of the changing market conditions with Thilo Koslowski, a member of our Board of Directors. Thilo is an executive adviser focused on technology strategies in the automotive and smart mobility markets. He previously founded Porsche Digital and served as its CEO. Before Porsche, Thilo founded and served as Vice President of the Automotive and Smart Mobility practice at Gartner.

Thilo, we're thrilled to have you with us.

Okay. So thank you, Thilo, for joining us today and for this conversation about the state of the industry. We're hearing a lot about economic changes. How is it impacting the automotive industry?

Yes. Thanks for having me. First of all, it's an interesting time for the automotive industry. And the industry always goes in cycles. There's nothing new. And right now, I would actually say that the auto industry is still in a very positive operating mode. Why? Because for the most part, most companies still make good money in the automotive industry. They have attractive products that they're launching, that they're selling. They are highly option typically, meaning their margins are pretty high. And there's still a lot of pent-up demand from the pandemic that happened over the last 2 years.

So that's all good news for the automotive industry. And cars are highly desirable for consumers. And of course, at the same time, we do see a little bit of clouding on the horizon, in particular, with regards to sentiment that isn't clear from a consumer demand perspective, how it will evolve in the next couple of years, maybe even a couple of months. And of course, it's due to the economic challenges that we're facing everywhere, including inflation, which is causing interest rates to go up, which makes it more expensive to actually buy a vehicle via a loan in the automotive industry. And of course, there is consumer concern about used car values, which are beginning to soften a little bit. And most people trade in the used vehicle to get a new car, et cetera. So there are some clouds on the horizon, but overall, I think the auto industry is still in pretty good shape.

And in this industry, what is the status of the ADAS from one end and from autonomous vehicles from the other end. So we are hearing about robotaxi companies like Argo AI shutting down from the other hand, companies that are mainly based today on ADAS, like Mobileye went public, and the share price went up. What do you think about the status of the ADAS versus automotive or full autonomous driving?

That's a really important question, actually. And I hope I can shed a little bit of light to this whole thing because to me, this is not surprising what's happening in the industry. Every time there's a new technology, a technology goes through a maturity cycle. And those cycles mean typically that there will be a consolidation in terms of R&D activities as well as market consolidation in an industry.

And the company Gartner has a term for this. They call it the hype cycle. You have an initial technology trigger. Then you have overinflated expectations that then go to the other extreme that's called the trough of disillusionment. And that will be followed by the slope of enlightenment. And eventually, you have a plateau of productivity, where the technology becomes mature, provides real value. So what's happening right now with regards to ADAS autonomous vehicles follows that line. But there's an important distinction to be made between both of these areas, ADAS and autonomous vehicles, which the latter means it's the combination. It's the merging of all of these ADAS technologies into something that then eventually might lead to a self-driving and driverless vehicle at some point.

And I think people had a lot of overinflated expectations, I mean, it takes a long time to figure these things out. It's highly complicated. But as we are progressing towards that goal of having self-driving cars, there will already be value that will be presented to consumers, to the automotive industry, to societies, and it's important to really differentiate between both of these areas going forward.

So where -- would you put the ADAS on the plateau side of the Gartner hype cycle and the autonomous driving down the hill today?

Definitely, I see that autonomous vehicles are at the bottom of this trough right now or entering this trough really quickly, right? And again, this is expected. This had to happen. There were way too many activities. They were doing the same thing and not everybody can succeed. I see ADAS being much more progressed in a positive way, matured in a positive way. ADAS technologies are already relevant today. They provide value to consumers, to the industry, and I anticipate that over the next 3 years, in particular, sensor technologies, advanced sensor technologies like imaging radar technology, will actually come to a price point and capability point where they have the reliability and value that the industry will increasingly use that technology to provide better ADAS functionality and applications to consumers. And that's a big deal.

So as we're all pursuing the self-driving car, in the meantime, there are real technologies that provide real value today and tomorrow. And I think that's really important to understand. I anticipate that by the end of this decade, meaning in 8 years from now, every vehicle will at least have some basic form of ADAS capabilities in the car. Why? Because consumers are expecting that as well and because there's so much benefit in this, and it also presents an upsell opportunity for the automotive industry. Self-driving cars will definitely take longer. No question about this.

I also believe that by the end of this decade, more people will actually at least occasionally experience self-driving technology, maybe even driverless technology, let it be on the trucking side, commercial vehicles, on the highways, or maybe delivery bots that do this. It's not going to be something that we'll experience every single day and every single moment, but that was an overinflated expectation to have to begin with. But the technology is definitely there. And as we're progressing, especially on the ADAS side, there's real value to be had. And I do believe that because technology is maturing in a good way that you will see the technology being democratized, meaning that more and more vehicles will actually be able to have this technology in their cars, especially also since the automotive industry is working on new electronic architectures that will be launched by the middle of this decade, meaning by 2024, 2025, and those new architectures will make it much easier to include those technologies going forward.

So this is a pretty big deal, and it would be a mistake to throw ADAS together with self-driving vehicles in the same bucket and then look at it in a negative way. It would be a mistake from an innovation perspective but even more so from a market perspective.

Thank you, Thilo, for this perspective, and thank you, everybody, for hearing us. Now we will be happy to take your questions.

We'll take the first call from Josh from Cowen.

I guess to start, can you walk me through what's changed materially from your perspective over the last few months that's driving the guide down? And most importantly, how do you feel about the long-term outlook and some of the revenue projections you had given previously for, let's say, the middle of the decade given the changing environment?

So when we are looking now into Q4, what we see that is changed is a shifting of some of the orders for preliminary samples from our customers between the quarter. So as you probably know, in '22, our revenues were not really based on full production. The full production was expected to the second half of '23 and is based on starting of production in China. And today with the announcement on HiRain, we see that our plan that our early revenues and early production -- serial production revenues will come from China, we can say check box on that. The next phase, of course, is winning contracts outside of China, in the Western world. This is something that should happen over the Q1 and Q2 '23. And those winnings will contribute to our revenues by the end of '24 and '25.

So I think you had previously communicated in September to March was a time frame when you would expect to get some more clarity on some proposals you had out there. Is that sort of still the right time frame? It sounds like maybe slipping into the second quarter of the year in '23?

First quarter and second also, yes. What we saw in the last quarter is that most of the OEMs shifted their selection and delayed a bit their selection. Some of them not meaningful, but some of them decided to skip to the next year model. And to stay with the current low-end radars that they have today and to reselect, reevaluate the replacement of the radar in the second half of next year, but majority of the selections moved to Q1, Q2 '23.

Okay. And then my final question, Veoneer, can you walk me through how that came about? And then how material of an expansion of the opportunities is that win? And I guess when would we expect that to start contributing to revenue?

Yes. So first of all, Veoneer is one of the largest players in the radar market, I think the third player in this market, selling today 50 million radars a year and expecting to reach 200 million radars by the end of the decade and shifting from the current radars to imaging radars in the future, a major shift. So for us, we see Veoneer as a main player. They were the first company to introduce ADAS adaptive cruise control based on a radar to Daimler, to Mercedes, and it's still one of their best customers. So we definitely see Veoneer as a major win. And the fact that they are basing their entire next-generation radar suite on our chipset is a major win for us. There -- they will go to production, I believe, in -- by the end of '24, early '25. So it's in line with our expectations. So of course, we will need to supply them with chips before they will be in full production mode. So second half of '24 is like our current -- like our original plan for the Western market is in line with their timetable as well as with the timetable of Valeo.

Now we'll turn the call to Suji from ROTH.

Congrats on the progress here. I had a question specifically about HiRain. I was curious for the unit forecast you gave here. Is that a relatively linear progression through '23, '24? Or will it be more back-end loaded?

It will be back-end loaded, but it will be -- it's not really linear. It will go and progress as we go towards the back end as we stated.

But it's starting strong. And that's been great.

Fantastic. And you put out this forecast there. You have -- it seems like with HiRain, you're working with not just automotive but the trucking and then the ports. Is there a potential upside to this number you've given as more customers later on? How should we think about that opportunity?

I think for '25, '26, definitely. For '23, '24, it's already very good numbers. And we are supporting them and working hard with them to make sure that they will really be in production on the time frame that they want to achieve that I believe it's in line with their automakers customers. And we are trying to support them. It might be that by the end of '24, there is an upside opportunity. But definitely, those are just the beginning. So assuming that this '24 numbers will be on time. I think in '25, we will see even much better revenues from HiRain. And they are just one of our partners in the Chinese market.

Of course, yes. And then maybe for Karine, the strong gross margin in the quarter, were there some onetime elements to that? And what's the ongoing opportunity in gross margin?

So I think, as you stated, yes, it's a onetime mainly in revenue mix as we said, that it's currently early revenues and valuations. So they're not really predictive of the margin in production. And they're very highly favorable margin-wise. When we go further, we're still behind the 60% margin in mass production. So I think that's...

Our model is (inaudible) 60%, not about the 70.

Gary will be joining us from Wells Fargo. We cannot hear you Gary, for some reason.

I was hoping that maybe you can give us an update on where you stand with your foundry partners in their ability to support your ramp when that day comes from commercial-related volumes.

Great. So actually, I'm going to visit the fab this week to see our facility there. We already have the entire testing facility, and it's working, and the chips are in the final stages of full qualification. So all of the testing facilities and all of the equipment that is needed to reach to full production is already there. And the team of GlobalFoundries together with our team, improving the test time in order to make sure that they will be able to test the chips on the capacity that is needed. On the side of the fab itself, we secured the capacity that is needed for our current customers. So the reasoning for HiRain to give us this purchase order, this preliminary purchase order was to make sure that they got the capacity from '23, '24, and we are working closely with our other customers to get from them also their purchase orders for '23, '24. So we will be able to make sure that the fab has the capacity for us.

Appreciate that. And for my follow-up, I wanted to know to what extent in your outlook and sort of your long-term forecast long term figure in fiscal year '23 and '24, any sort of interruptions that may exist in your China customer base from COVID mitigation or some of the geopolitics that seem to be -- I guess, specifically U.S. export restrictions and whatnot. Any thoughts there would be helpful.

So first of all, regarding U.S. export limitations right now, there's no limitations on the technologies -- on chips that we are selling as long as Intel will be able to sell their chips to computers and TI will be able to sell their chips for radar, we as an Israeli company, a non-American company, we believe we won't have a problem. And right now, I think that this is the case. Also, it's good to understand that HiRain has facilities outside of China, where they are supporting their customers that are not Chinese or their Chinese customers that producing cars for exporting. So I believe that not all of this revenues will go directly to China. So it might -- part of them will go to a plant in the U.S. or plant in Europe where they have. Regarding '23 and '24, as you mentioned, our plan was on -- to base '23 and '24 on nonautomotive and on China.

On the China side, we see that we are more or less in line with our plan. On the Western world, we see a bit of a shift that might influence the longer term. And also on the non-pure automotive, we definitely see delays like in delivery robots like in trucks and robotaxi. We all see that they are not there. This wasn't a major part of our story, of our entire revenues in '25, '26. But for the short term, this has a larger impact, we still don't know the exact numbers because since those customers are not like automotive that they give you now a focus for '26, those customers are living between the quarters. So we still need to see with them in the coming weeks what is their expectations for '23. So we will be able to get with the exact numbers of '23 only with our Q4 earnings.

Now Matt from Maxim will join us. We can't hear you.

Matthew, we can't hear you.

Okay. So I wanted to touch on -- you covered a little bit of taking your -- taking costs down in the current environment. So can you maybe go a little bit further into where you saw opportunity to pare? Is it just in terms of engagements that you're paring those down towards ADAS? And yes, just any color on this one.

Yes. So since the beginning, our main focus was on ADAS. So we always thought that it's going to be an evolution and not the revolution. So the technologies will start in Level 2, Level 2+, then Level 3 and in parallel the Level 4, and we believe that Level 4 will start ramping up by the end of the decade. So our main focus -- and this is why we are working with the traditional Tier 1s is on regular ADAS, where we believe that 90%, 95% of the revenues are coming. Of course, we announced AutoX win a few months ago. AutoX is today the largest player in the Chinese market. And as time goes by, they also remain one of the fewest players in this market because many other robotaxis are shut down. So we are supporting them as well in order to show that our technology has a long way to go also to Level 4. But in terms of revenues, ADAS from the beginning was our main focus. And as time goes by, we see that this strategy is really the right strategy because the full autonomous and the robotaxi and the full autonomous trucks definitely will take longer than expected.

Okay. I mean I guess as a follow-up then, specifically, how are you -- or where are you finding opportunities to pare costs? And what does that look like as we move into Q4? Or to the extent you could talk about expenses for early 2023.

So we look very deeply in our cost structure, and we're doing several steps to optimize our capital structure in order to make sure our runoff time is long enough to keep the company going until breakeven. We also are in the process of securing a credit line in order to make sure that we have also stable ability, if needed, again, not now but in the future, and we're looking into every aspect of our costs in order to make sure we're working much more efficiently and directly to production.

But in general, our '23 expenses will be lower than '22. We reduced expenses across the board from our headcount towards other contractors as well as other expenses. So we optimized the cost to make sure that we have sufficient cash to run even without any revenues for more than 2 years from now. And of course, with our revenues -- our expected revenues and our initial purchase order, we should be good (inaudible).

Final question from me. I think it was last quarter that we talked about you restructured your engineering team. I think you had 2 teams. You maybe structure it a little bit differently. Did that contribute to lower costs in the near run? Is that more just an operating and flow change? And how has that worked out so far?

I think it worked out so far great. It improved our ability to take the silicon into production and to focus on the silicon side very well. And from the other end, on the radar side to support our customers. And actually, today, we have 6 different customers with their radar fully operating. And this was because of this focus. So it helped us dramatically with the focus. In terms of cost, it's also a bit helped in cost savings, but the reasoning at the beginning was the focus is also -- the byproduct was that we improved our expense structure.

Also, we have the ability to take in-house some works, which we usually outsourced, and this is also an ability to reduce our cost. Now we have Jaime from R.F. Lafferty.

Jaime, we cannot hear you. Yes.

I think it took a little bit of delay. I mean I'm traveling. So I don't have a good IT connection. So you mentioned the orders will be linear, I mean, not linear, but lumpy. What about the backlog? Are you going to be booking -- how are you going to be booking backlog? Because you -- I know it's going to be significant. Is it going to be to multiple tranches? Or is it just going to be one large order into the backlog?

It will be tranches based on customers and OEMs. But of course, from -- also what we get from a customer, we get it split the PO based on their forecast for the 6 months ahead. So that's the method that the backlog will grow. And as I mentioned, I think it's mentioned in the PR -- in the press release that our current backlog does not include the HiRain preliminary order as well.

Right. Okay. And the second one is more of a comment. Congratulations on partnering with Veoneer. That's a major player concerning. They were looking to be acquired for Magna and it's a good foothold into the Western market. So congratulations on that accomplishment.

Thank you. So thank you, everybody. We are very pleased to have you join us today. To our employees and partners, your continued dedication is deeply appreciated. We look forward to updating you further on Arbe's progress in the coming months. Look out for updates as we prepare for several investor events as well as hosting a booth in CES. We'd love to meet you in person for further discussions. Please contact us at investors@arberobotics.com or visit our site to schedule a meeting. Thank you all.

Thank you.