Thank you, Operator. Hello, everyone, and welcome to RoboSense's earnings conference call for the first quarter of 2026.

The company's earning results were released earlier today and are available on our IR website, www.ir.robosense.ai, the website of Hong Kong Stock Exchange, www.hkexnews.hk, as well as on newswire services. Today, you will hear from Mr.

Kelvin Lau, our CFO, and Mr. Johnson Wan, our Deputy CFO, who will take you through the company's operational and financial results for the first quarter of 2026.

After their prepared remarks, Kelvin and Johnson will be available to answer your questions. Before we continue, please note that the discussion today may contain certain forward-looking statements, which involve known and unknown risks, uncertainties, and other factors which are beyond our control and may cause our actual results, performance, or achievements or industry results to be materially different from any future results, performance, or achievements expressed or implied by the forward-looking statements.

In light of risks and uncertainties, the inclusion of forward-looking statements in this discussion should not be regarded as representation by the board of directors or RoboSense that the plans and objectives will be achieved. Shareholders and investors of the company should not place undue reliance on such statements.

RoboSense does not assume any obligation to update any forward-looking statements except as required under applicable laws. Also please note that some of information to be discussed will include non-IFRS financial measures and not be complied or presented in accordance with IFRS.

The IFRS financial measure and information requiring those non-IFRS financial measures to RoboSense financial results prepared in accordance with IFRS are included in RoboSense first quarter results announcements, which has been posted on the company's IR website at www.ir.robosense.ai and the website of Hong Kong Stock Exchange, www.hkexnews.hk. Finally, as a reminder, this conference call is being recorded.

In addition, a webcast of this conference call is available on RoboSense Investor Relations website. I will now turn the call over to Kelvin Lau, the CFO of RoboSense.

Thank you, Grace. RoboSense delivered a solid start to 2026, boasting robust revenue growth while accelerating the industry shift to next-generation digital LiDAR.

The company's leadership center of physical AI powering intelligent vehicles and a broad range of robotics platforms. From robotaxi and robovans to robotic lawn mowers, commercial cleaning systems, humanoid robots, and beyond, with our proprietary SPAD SOC LiDAR chips, which was reinforced in Q1 2026.

All right, first of all, let me share with you some of the key highlights of our Q1 2026 performance. Our revenue reached RMB458.8 million, up around 29.9% year-over-year.

Total LiDAR shipments reached 330,300 units, and increased by about 204.1% year-over-year. Robotics LiDAR volume significantly increased by 1,458.8% year-over-year to 185,500 units, accounting for about 56.2% of our shipments.

RoboSense maintained a leading global share across multiple robotics LiDAR niches and ranked among the top domestic suppliers for intelligent robots, robotic lawn mowers, and commercial cleaning systems. In ADAS solutions, we secured volume orders from leading European humanoid robot manufacturers, advancing robotics platform from verification to full-scale mass production.

Finally, order backlog on hand exceeds 9 million units for ADAS. RoboSense recognized a long featured seasonal pattern of lower sales volume of LiDAR products for ADAS applications in the first half of a year, and higher shipments of LiDAR products for ADAS applications in the second half of a year.

Our overall gross margin dropped slightly to 21.7% in Q1 2026 from 23.5% in Q1 2025, mainly attributable to the higher overhead incurred in manufacturing our new EMx LiDAR products in their initial mass production stage. However, benefiting from concentrated delivery of new automotive vehicles in the second half of the year, incremental market demand in the robotics sectors, coupled with cost optimization of self-developed chips and improved capacity utilization, the company's management team maintained full confidence in the company's future development prospects.

In April 2026, RoboSense launched its self-developed SPAD SOC digital architecture platform together with two breakthrough proprietary chips. The first one is Phoenix.

Phoenix adopts a highly integrated design with single-chip and single optical path. It supports 2,160 lines and has passed the AEC-Q100 automotive grade of qualification certification with full qualification for automotive grade mass production.

Meanwhile, the 4-megapixel high-definition LiDAR solution based on Phoenix has secured design wins from leading automakers and expected to realize mass production and vehicle installation within 2026. Another chip is called Peacock.

Peacock's the world's highest-resolution large-area array SPAD SOC chip, as well as the highest-specification mass-producible area array chip in the current industry, reaching VGA-level 3D imaging standard. Products equipped with such chips has commenced small-batch delivery and is slated to kick off large-scale mass production in the third quarter of this year.

This in-house SPAD SOC chip forms the core technological moat that will drive LiDAR leadership for the next decade. 2026 marks a critical transition years for the global LiDAR industry, defined by a broad generational shift from traditional analog and mechanical LiDAR architecture to next-generation digital LiDAR system.

This industry transition raises the overall technical bar for LiDAR beam count and resolution, which is essential to support safer and more reliable autonomous driving, and pave the way for broader adoption of L3 and L4 functionality over the next few years. Legacy analog and mechanical LiDAR solutions are fundamentally limited by their inherent hardware ceiling in resolution and intensity, making them unable to meet the evolving requirement of higher level autonomous driving.

RoboSense continues to lead the industry's digital transformation, anchored by our proprietary SPAD SOC chip architecture. The company is scheduled to start the mass production of our new in-house developed perception chips, Phoenix and Peacock, in the second half of the year.

With our early-mover technology advantages, RoboSense is at the forefront of driving the industry-wide shift towards pure digital LiDAR. We expect gradual gross margin improvement as more product lines migrate to our fully in-house chip solution, and our cost structure will be optimized step-by-step, supporting steady margin recovery over time.

Ultimately, our full-stack technological upgrade enables RoboSense to deliver higher resolution, safer, and more reliable sensing solution to our OEM and robotics customers, strengthening our long-term product merit and market positioning amidst the rapid evolution of physical AI. Driven by new vehicle model ramps, explosive robotics demand, ongoing chip cost optimization, and rising capacity utilization, RoboSense maintains a confident long-term outlook as an AI technology platform executing its dual-core automotive and robotics strategy.

We are well-positioned to capture the multi-year growth of the physical AI market, delivering sustainable returns for shareholders while advancing a safer and smarter world. Looking forward, RoboSense remains dedicated to innovation in AI algorithm and hardware technologies, while continuously reinforcing its leading position in advanced driving and robotics industry.

We are committed to becoming a global leading robotic technology platform company and creating long-term and stable market value and social benefits for the industry. Now I would like to turn the call to Johnson.

Thank you, Kelvin. You've just heard from our CFO, Kelvin, that walked you through our Q1 highlights.

Now I wanted to first address some of the key questions I know you guys will ask. A lot of the people online are also my friends, I wanted to answer some of the top questions that I know I'll be getting from you guys when we open the floor for questions.

I also want to set the scene on how one should think about RoboSense going forward. Now first and foremost, RoboSense is a physical AI company, right?

Humans perceive 80% to 90% of the world through vision. Our LiDAR and next-generation LiDAR camera solutions are the eyes, and we're delivering superhuman vision to power the physical world for AI.

That is all of course built upon our in-house SPAD SOC that we've already spent 8 years researching and developing. This breakthrough will unlock substantial new TAM across many sectors that actually not explored previously for LiDAR, which will enable us to capture market share from traditional cameras in a wide range of industrial scenarios, including factory automation, humanoid robots, passenger vehicles.

Now the market right now is heavily focused on upstream AI infrastructure names as a core investment theme and has largely overlooked the booming physical AI segment. As you can tell from our strong shipment growth, all our business lines are gaining strong momentum, and the LiDAR adoption cycle is still in the early stages.

It's only a matter of time before the market fully recognizes this tremendous opportunity. Now turning back to our Q1 results, we have very strong top-line growth, 40% growth year-over-year, shipment more than 200%.

Now simply put, the strategic priority for RoboSense in 2026 is market share expansion. We'll continue to extend our competitive advantage, especially in robotics LiDAR, which is a segment that we see a structurally healthier gross margin of around 30% to 40%.

Now we're very happy to announce that we are number one in market share across many robotics runways, such as robotaxi robotaxis, lawn mowers, commercial cleaning, robodogs, mining trucks, and many more. We plan to steadily scale these high-margin businesses to strengthen the overall profitability of our company.

Now on the ADAS front, gross margins face temporary gross margin pressure, and that is a function of a deliberate and strategic product transition that we've undertaken. We're actively proactively driving all our major OEM partners to migrate from legacy analog and mechanical LiDAR to our SPAD SOC-based solutions, and we've actually made substantial progress so far.

Now today, approximately 50% of our ADAS shipments are actually SPAD-based units. Our first-generation SPAD chips are, however, currently sourced externally, which will, however, push up the component costs and weigh on short-term gross margins.

Now additionally, we've actually ramped up factory automation to prepare for in-house chip production, and that has also contributed to the year-over-year margin compression. Now that said, we want to reiterate that the current margin pressure is temporary.

Our self-developed SPAD SOC chip will enter mass production in the second half, and this will help us bring notable chip cost reductions. Now even so, and we have to admit that our ADAS gross margin stayed below 20% in Q1, and it's expected to remain below this level for most of 2026.

Now as a result, RoboSense will strive to achieve profitability for the full year of 2026. We view these investments as highly value-additive rather than a negative outcome.

Onboarding top-tier OEMs onto our SPAD architecture will actually build lasting customer stickiness, raise switching costs, and further fortify our long-term technological moat. We're already in active discussions with OEMs on the next-generation vehicle LiDAR projects, and all our partners have expressed to us a clear demand for higher line-count products for next year, which will create a distinct technological gap between RoboSense and our peers by 2027.

It is clear that our OEM partners are positioning themselves for higher line-count products, and our external chip procurement strategy in first half will lead us to a cost headwind. You can think of us as taking the pain first.

Now I know you guys will also ask about our ASP, which fell 55% year-over-year in the first quarter, and that is a function of the higher Q1 base from last year, but the pace of decline will moderate as we go towards the second half of the year. Broadly speaking, our ASP decline, even though it's compared to peers, the underlying causes are fundamentally different.

Our ASP compression is actually driven by mainly changes in the product mix, and not so much a function of the price competition. First, if you look at our lawn mower LiDAR volume, that was close to zero in first Q 2025, but that contributed meaningfully to our robotics volume in Q1 2026.

Second, we shipped a meaningful amount of our E1 blind spot LiDAR in this quarter. Now these two product lines collectively accounted for more than 50% of our Q1 total shipments, and those products carried a lower average selling price compared to our flagship automotive LiDAR products.

With that, I want to open the floor now for questions.

Hi management, thank you for taking my question. This is Joey Young from BOCI Securities.

I have one question regarding your ASP. We see as a big shift towards higher performance product or more premium product made penetration into overseas or joint venture clients.

I was wondering is there is any chance that there will be a potential inflection point for ASP in the second half of 2026 or in 2027? Thank you.

Thank you, Zoe. It's a it's a good question and also it's commonly asked by quite a quite a lot of investors.

Zoe, as as I think you I think you you may be aware that, okay, LiDAR LiDAR's ASP, I think the past 2 years keep on degrading from, for example, in 2023, I think the LiDAR ASP is around RMB3,500, and then dropped to 2,500 in 2024, and then further dropped to around 1,200 in the second half of 2025. I would as but I can tell you I think this type of, I think, price deflations already close to a long tail, okay, close end.

Why I say so? Because I think I think if you look at this question, you have to divide, I think, the LiDARs into two categories.

One is, I think, the high-performance, high-definition high high-beam-count LiDARs. Another one is, I think, the blind spot LiDAR, the lower laser beam count LiDARs.

As what Johnson mentioned, I think, in his remarks, right now, I think, forward we will see, I think, most of the automakers they are now transi- transitioning, I think, to adopt high-performance and high-laser-count LiDARs. This type of, I think, trend will become the major mainstream, I like, I think, beside of default configuration, okay, for all those OEMs.

I think I think for high-laser-beam li- LiDARs specification, I think, going forward is a irreversible, I think, industry trend going forward. Of course, I think, this type of high-performance, high-laser-count LiDAR deserve for higher ASP or high higher higher price line.

However, I think I think I think this type of high-laser-beam LiDAR does not e- create to higher production cost because because our self in-house our self-developed in-house chips can offer, I think, I think, I think, really, I think, more significant cost advantages to reduce the production cost. Like, okay, all those high-laser-beam LiDAR, I think, for example, 200 7- 720 720 laser version has already, I think, going forward we we would as a minimum standard configuration for all those automakers.

Therefore, for all those high-laser-beam, we we believe, I think, the ASP can can be stabilized and also possible, I think, it was a slightly up uptick, I think, for all those high high high-laser-beam LiDARs. I think for all those lower-laser-beam LiDARs, like blind spot LiDAR, I think, going forward, I think, with increased in the installation rate and the penetration rate of this type of blind spot LiDAR, we can enjoy a higher economy of scale and further reduce the cost.

Then there will, I think, we forecast, I think, for this type of blind spot LiDAR going forward, the ASP will drop a little bit. Net-net, I would say, combined together, this the overall ASP for li- for overall LiDAR, I think, can be stabilized, I think, going forward.

Would not be happen again and again like we took a steep decrease and then there's a a slip drop in the in the ASP. Even even if it happen again, this is what we predict.

Yeah. Yeah.

Thank you so much. That's very clear.

Yeah.

Thanks for taking my question. This is Sherry Liu from J.P.

Morgan. Also want to congrats on the robust volume and revenue growth.

We're also glad to see a very more balanced product mix in terms of the volume. We do see the blended gross margin dropped on YoY basis.

I think Johnson already mentioned margin pressure from outsourced chips. Could you elaborate a bit more on the reasons behind the margin correction?

Thank you.

Thank you, Sherry, for your question. To elaborate, I think there are probably a few reasons, and externally sourced chips is only one of the few reasons.

You all know this year's auto market is slower than previous years because of the overdraft in subsidies. As a result of poor sales on the automotive side, that has actually led to lower utilization rate of our facilities, and that actually increased our fixed costs.

Second of all, as we are now upgrading all our lines to digital LiDAR, we've actually had to put in new pieces of machinery, and that actually increased our costs, and until they're fully ramped up, that will also impact our gross margins. Third, we've also seen increase in some of the raw material costs, as you are aware that there is a global commodity inflation that is happening only on, certain products.

Also our cost reductions with our upstream suppliers in terms of those negotiations have not yet fully been completed, and those negotiations will be completed throughout the year. Of course, the last reason was the externally sourced chip question, which is as we can migrate to our more of our in-house SPAD SOC chip, that will actually help improve the margin profile of our company in the, coming quarters and months once we fully migrate to our in-house products.

That's got it. Thank you, Johnson.

Thank you for taking my question. This is Kai Chi from Cinda Securities.

My question is about SPAD. We notice that compared with major suppliers, such as Sony, what are the key performance advantages of our in-house SPAD technology?

Also, compared with third-party solutions, how much impact do in-house chips have on product pricing and overall cost structure? In addition, how is the mass production ramp-up for products based on the company's in-house chips, such as the Phoenix and the Peacocks?

Thank you. Thank you for your question.

Answering your question is that, in terms of, I think, the core advantages of our self developed in-house proprietary chips, I think the most important number one is, I think, our superior high-pass performance of our self-developed chips. I think our self-develop in-house proprietary chips can support higher laser beam lines specification, and also with optimized the cost structure.

Okay? It's number one.

Number two, I think, very importantly, I think, is the PDE, the photon detection efficiency. Okay.

Our in-house chips can, I think, the capacity can reach, I think, possible, I think, 45% of the PDE, which is well above, I think, the industrial average. It's very important.

I think, the number three, I think, another important point, I think, is the cost efficiency. If we compare with, I think, all those chips we the similar chips or the equivalent chips we purchased from third parties, in terms of cost, I think we can really cut the cost by about 50%, cut down by half.

Is a very obvious and very important, I think, cost efficiency factors we have to consider. Another, as what I say, okay, our chips can support, I think, ultra high laser beam.

Okay. Right now, our chips accept support, LiDAR with laser beam up to 2,000 2,160 laser beam.

Going forward, our next-generation products, chips, can, I think, can achieve the laser beam line reaching, I think, 4,000 more than 4,000 laser lines. Lastly, I think, chip, I think, of course, I think, can provide a more compact chips dimension, chip dimension.

This, I think, these are all the core advantages, I think, why we developed we still focus on developing our in-house proprietary chips. This is very important for the future competition this in this industry.

In terms of the mass production timeline, you mentioned this before, okay, Phoenix chips. For Phoenix chips, I think, we already obtained a design win from some leading OEMs, I think, in the market already, and we expect it to start the mass production of our Phoenix chips, I think, within 2026, mostly in the second half of this year.

Our Peacock chips, Peacock, I think, we already, I think, right now currently, we already achieved that the small-batch delivery, as what I mentioned in the in my remarks already. We will enter into, I think, mass production, I think, in Q3 of this year.

You will see, I think, this type of high-performance customized chips, you didn't find that there's so many, I think, right now in the market. Going forward, we believe, I think, we strongly believe that adopting, self-developed high-performance chips tailor-made to our OEM customer and also our robotics customer will become definitely, I think, going forward is a irreversible trend going forward.

Yeah.

Thank you. It's very helpful.

Thank you very much.

Hi management, this is Yu Ho from CICC. Thank you for taking my question and congrats on your strong results.

My question is about your RGB-D solution and SPAD technology. We see that peers also launched the 6D full color products.

Could you help us understand and share more details on the key difference between your RGB-D solution and peers' full-color approach? Also you have already established a leading position in SPAD SOC technology.

Going forward, what product or pricing strategy will you plan to adopt to strengthen and monetize your advantage? Thank you.

Thank you for the question. I think you have seen on the media that some of our competitors as our peers have actually launched similar RGB-D camera products.

For example, Ouster in the U.S. also launched a product recently.

I think the core difference of our product versus that of others is that, first of all, we have our own in-house SPAD SOC chip, right? That our line count has now reached 2,160 lines, and in our next-generation product, it's going to be double that, right, in terms of the line count.

Now if we think back to LiDAR, we always communicated our LiDAR beam count using lines, right? Now if I was to convert that into resolution or pixels, 128 lines is only 150,000 pixels.

It's actually very, very poor in terms of resolution. When we get to 2,160 lines, that's actually 4 megapixels.

When we get to 4,320 lines, it's going to be 8 megapixels. Now the human eye, the cornea of your eye, is actually 6 to 8 megapixels.

For the first time ever, we are able to create a sensor that is better, if not on par, with the human eye. That's why I refer to our LiDAR solution or LiDAR camera solution as a superhuman vision solution that many people have actually not thought about that LiDAR is capable of doing.

It is definitely going to replace a lot of the camera solutions that are in the market right now. I want to give you a couple of examples of whether, where we can deploy these products.

You can think about just an auto itself, right? A typical car can have, let's just say, 1 LiDAR, 12 cameras, 6 mm-wave ultrasonic radars.

What if I was to tell you that we can start replacing a lot of the cameras on the cars now by using our LiDAR camera? That in itself is a significant TAM.

Second of all, you'll also see that there are a lot of scenarios on the industrial side, such as industrial inspection, that the LiDAR solution is both way cheaper and more precise versus that of some of the 3D vision solutions that are out there. This RGB-D camera or LiDAR camera that we're going to be launching very soon, it's going to be game-changing.

I don't want to disclose too many details about our product until it's officially launched, but please be on the lookout for this product.

Got it. Looking forward to that.

Thank you.

Thanks for taking my question. Here is Sarah from China Galaxy International.

My first question is about our chip business. We established the in-house chip strategy back in 2017.

Could you please share more what are the key advantages we have observed now? My second question is about the market share.

What are the key factors that will help the company continue gaining market share going forward? How will the company further strengthen its leading position and capture a bigger market share in the future?

Thanks.

Good question. Thank you so much.

I think yes, again, I think quite a lot of investors keep on asking us why we established our in-house chip development team, I think, back to 2017. I think starting with back in 2017, we already realized that, okay, going forward, if we want to capture a larger market share, okay, in the LiDAR industry, we have to have our own self developed in-house proprietary chips.

Okay. Why?

Okay. In-house, what is in-house-developed semiconductor chips?

Definitely, number one is we can use our self developed chips to strengthen our core competitive advantages. Okay?

You can see that going forward, okay, in terms of LiDAR competition, I think all these the competition will be focused on the semiconductor chips competition. LiDAR already right now is already become a very sophisticated electronic device already.

Going forward, I think the competition will be focused on the chips development and the advancement of the chips. Moreover, okay, our G&D chips will also constitute, I think, another core technological barrier, I think, for all those dual cameras in the future.

Okay. Second reason is that, okay, in-house developed chips can facilitate, I think, the LiDAR performance upgrade, upgraded.

You can see right now based on our Peacock and also our Phoenix chip, we can achieve, I think, LiDAR up to laser beam up to 2,160, and in future will go up to more than 4,000 laser beam. Okay.

A higher laser beam means, I think, the detection range, the accuracy of the detection of the tiny objects, I think, up to 150 m, can be very, I think, can be very accurate. This is very important going forward, I think, for all those L3 and L4 autonomous driving.

Another one, I think, the third one, I think the third reason is, I think, the cost efficiency that I keep on saying that, okay, in my in the in my last question is that I also mentioned that if you compare the cost of, I think, purchasing all those chips from other third parties, and Sony, compare that with our in-house cost incurred, I think it's about 50% reduction in terms of cost. Going forward, in terms of, I think, cost efficiency, I think we can get more advantages on the cost efficiency.

Another one is, I think, in terms of, I think, the long-term strategic value. You everybody know, okay, demand as as we can see that all those OEM right now, they are demanding, I think, for more high-laser-beam-count products going forward.

I think not only, I think, like BYD are asking for, at least, I think, 720 laser beam for the core, and some of their models also be equipped with higher, I think, 2,160 laser beam. Also when all those vehicles, I think, go into L2+ or L3 or L4 autonomous driving, okay, they will demand for, I think, even higher, I think, laser beam.

With our own self-developed chip which can develop a very high performance and also more cost-efficient products, we can secure a higher market share in the ADAS we believe, I think, in the long run. Moreover, okay, our self-developed chips can develop more customized sensor products, I think, for our robotic customers, like lawn mower.

You will see in the past years, I think, we already developed some new products, tailor-made products, I think, for our lawn mower and some others robotics customers, all because of our in-house-developed chips. In terms, I think, in terms of the technological leadership, we strongly believe that, okay, in terms of the progress of developing this type of SPAD SOC chips, we are leading, I think, we are holding a leading position in the industry compared with our peers.

If you look at all those, okay, some of the people, some of the investors say, okay, why all those pure chip manufacturers they cannot provide this type of chips, I think, for LiDAR company like RoboSense? What I can tell you is that, okay, for all those pure chip manufacturers, they don't know how to do LiDAR, okay.

They cannot come up with all those products which 100% fulfill the requirements, okay, raised by LiDAR company. I think, all this, I think, you've got to be relied on our in-house, I think, like chip-development chips to support our more and more sophisticated products going forward.

Another very important point is that, okay, you got to remember that for all those OEMs, okay, if you want to shift, I think, from one LiDAR supplier to another supplier to another LiDAR supplier, they have to incur additional shipping costs, I think, or machine cost, I think, on this type of other suppliers. At this stage, I think, if we got a leading position in terms of the higher laser beam LiDAR, we can ensure, I think, the stickiness of the customer relationships with all those OEMs.

Going forward, net-net is, I think, what I would say, firmly hold, on this type of in-house chips development technology. We are confident that, going forward, okay, we can capture a larger market share, I think, in the ADAS and also in the robotics sectors, yeah.

Thanks, Kelvin. Very clear.

Appreciate that.

Thank you.

Good evening, management. This is Shuwan Li from CICC.

Congratulations on the consecutive and strong YoY growth on our shipment and also revenue. I have two questions.

The first one is we found that LiDAR players are all pushing for higher beam count, and the RoboSense has already reached 2,160 beams. The question for me is how much real-world improvement do you actually get from higher resolution for level 2+, level third level 3, and level 4 robotaxi applications?

What would you say is the rough upper limit of beams that is really needed? My second question is looking ahead over the next 2 to 3 years, where do you see the main competition in ADAS and robotics sensors?

Is it higher resolution, sensor fusion, or cost? How is RoboSense positioning itself in this direction?

Thank you. I'll take the question.

I think first of all, if you look at safety, safety for autonomous vehicle, I don't think there is a limit in terms of safety, right? Obviously with a higher beam count to your LiDAR, you are enabling the vehicle to drive safer.

For example, our 2,160-line beam LiDAR can see objects that are 600 m away. For the same LiDAR that was only 128 lines, you can only see the object that's only 100 or so meters away.

In which scenario is it safer? It's the 2,160 scenario or the 128 beam scenario, right?

Currently, if you look at a lot of the L2+ vehicles in this year, they're now adopting our EMx, which is the 192-line solution. We're already in talks with a lot of the OEMs for next year's product, and I don't want to disclose too many details, but the jump is going to be a lot higher, and it's not going to be 300 lines, right?

Now L3, I think currently will this year will require 500 lines, because the L3 line and L4 line is actually quite blurred. Our EM4 has typically about 520 lines, and a lot of our robotaxi players right now are equipping that their main LiDAR with our EM4 LiDAR.

The 500+ line LiDAR for L4 is likely going to go to more than 1,000 lines next year. Now I think the breakthrough is when we start producing this 2,160-line, 4,320-line, because, like I said, we're creating a sensor that is better than the human eye.

If you think about it, because LiDAR gives you the depth information, right? In the past, we've always been training a lot of the autonomous driving by 2D data that we're getting from the cameras.

We're not using the 3D depth information that you actually get from a LiDAR camera. In the future, I actually think a lot more OEMs will adopt our LiDAR camera, because it can help them save on the compute capacity, right, going forward.

Basically, we're at the point where the competition, right, for LiDAR is now all going to be digital-based, and if the competition therefore on a digital-based LiDAR is on the chip. I think we have a head start from a lot of our peers, since we've spent the last 8 years doing R&D on our SPAD SOC.

Yeah. Thank you so much.

Hello, and this is Ben Wang from Deutsche Bank. I have a question about BYD, your likely most important client.

Could management share more detail about current cooperation with BYD? Over the past year or so, we've gathered City Cruise in all those BYD models, are there any new developments in the additional opportunity this year?

If BYD seems to be having an event tomorrow evening, is there any good news for us for current, tomorrow? Thank you.

Thank you, Wang Bin. Thank you for the question.

Yes. Yes, BYD is one of our major, major good customers, in our ADAS business.

If you look at if you look at the number last year, in 2025, in the LiDAR installation penetration rate of all those BYD models, I think, is around only 10%. Because last year, I think, their total shipment is about, you wait for it, it's about 3.5 million to the best quality.

Their total LiDAR purchase from RoboSense and our competitors, it's all it's all around 400,000 to 500,000. The penetration rate is only around 10%, something like this.

As what we mentioned before, okay, right now we captured, I think, the largest chunk, of the the share from BYD already because of our because of our new product, EMx. In terms of functionality, in terms of specification, our our EMx, models have more superior than our previous products, we capture more market market shares, I think, in BYD.

This year, we forecast, I think, the shipment of LiDARs going to BYD is around from from RoboSense side is around 450,000 to 600,000. If you look at, I think, the total number of the other domestic vehicle shipments, this year is 3.5 million, okay, still the same as last year, so that means, I think, the pen- penetration rate right now we estimate is only about equal to about 12% to 7- 17% only.

This I would say, I think, this is a relatively, small number, only 12% to 17%. Let us see, okay, if there what they are going to announce, I think, tomorrow in their tomorrow's summit event, summit event.

Also, you if you okay, one thing, you I think I think you went to the, I think, the Beijing Auto Show, you would see that, okay, for BYD vehicles, all those lower-priced vehicles, like Hai'ou, Haishi, you would see, I think, they started to install LiDARs in this lower-priced vehicles. We really hope to see, I think, going forward, especially this year, I think, we'll see, I think, the LiDAR installation rate, I think, in their vehicles this year, this year, Wang Bin.

Thank you. I want to add to So

I want to just add a quick comment also. I think not only do we have BYD and Geely, I think that have effectively given us the nod as a primary supplier to LiDAR.

If you look at the penetration rate of these two companies combined, they're only their penetration rate for LiDAR stands at about 10% to 15%. Over the next few years, we're going to see a lot of growth from these two customers alone, and that will actually help us regain our market share in the LiDAR market.

It's safe to say that the market share for RoboSense in terms of the in the ADAS market will bottom out. We will probably regain a leading position, likely in 2027, if not 2028, just by the sheer size of these two customers.

That's why I started the call by telling our friends here that, market share is a strategic priority for RoboSense in this year, and getting them on board, switching them to our digital LiDAR solutions, is key to this. Also

starting from this year, you can see that, okay, more and more lower-priced vehicles started to install LiDARs, like Leapmotor, like BYD. We are Net-net, I think, we are quite confident and optimistic, in terms of the LiDAR penetration rate, I think, going forward.

Great.

Great. Thank you.

Hi, this is Xiaoyi from Jefferies, and thanks for taking my question. I've got a couple on the lawn mower business.

First on the Roborock partnership you just announced, any color you can share on the incremental shipment contribution we should expect? Just to get a better sense of the landscape, where does RoboSense sit in terms of market share in the lawn mowers today, and are most of your key customers sourced with you?

My second question is regarding the shipments trends from here. Looking into the second half, what does order cycles typically look like with these customers?

I'm also curious how you are factoring the potential EU tariff changes at the end of June, and whether you've already seen any pull-forward demand earlier this year that might impact the second half runway? Thank you.

Thank you, Xiaoyi, for your question. I want to first update our friends here on our 2026 full-year volume guidance.

We do expect annual shipments to grow 2 to 3 times this year. Last year our total shipments was about 900,000, among which 600,000 was ADAS and 300,000 was robotics.

We're going to build on that and grow our volume 2 to 3x this year. Now by segment, the robotics LiDAR segment is probably going to be on the higher end of that range, closer to the 3x, the ADAS LiDAR growth segment will be closer to the lower end of that guidance.

Now if I was to look at the, to answer your question in terms of the lawn mower business, yes, we just signed a huge contract, strategic partnership with Roborock, so that is going to increase our volume. On the lawn mower side, we think the volumes are going to be going to closer to 600,000+.

Last year, we were about 200,000 in lawn mowers, and 2 years ago, we actually had almost no lawn mower volume. You can actually see how fast the robotics segment is growing.

They just come from nowhere, right? Our market share in the robotics segment right now is about in the lawn mower segment right now is about 70%.

A lot of the key customers like Jiuhua, Positec, right, Roborock, basically, they use our company's products. Now there is seasonality, though, right, with lawn mowers, and a lot of the shipments from lawn mowers happen in Q1 and Q4 of each year, because they're used in Q2 and Q3 typically.

Therefore, what you'll see in the Q2 is that we're going to have very little contribution from lawn mowers, but higher contribution from the ADAS side. Now in terms of the tariffs, so far, in terms of our shipments, we've actually not seen an impact.

We've actually seen customers coming in and placing orders for the back half of this year, and that would obviously be already past June. I think a lot of our partners, they also have capacity overseas also.

The tariff impact is actually minimized, or not as big as what the market thinks it is. We are still seeing extremely strong growth on the lawn mower side.

Thanks, Johnson. Appreciate the color.

Hi management, this is Xiangyu from Haitong Securities. Thanks for taking my questions.

I'd like to ask about the robotaxi business. How is the robotaxi business trending as a share of RoboSense overall business?

Which clients are driving the incremental growth? Could you also provide some guidance on this year's shipment volume and overall business progress?

Lastly, have there been any new breakthroughs with overseas robotaxi clients? Thank you.

Thank you. Thank you for the question.

I really love to talk about robotaxi sector. Because right now, I would say RoboSense already holding dominant, I think, position in the robotaxi sector.

Everybody know, okay, robotaxi is going to grow very, very fast. Not this year, but, in 2 to 3 years' time, I think I think in terms I think in terms of the this type of robotaxi numbers, I think, will increase, I think, significantly in 2 to 3 years' time.

Why I say we are we are dominant right now in the robotaxi sector is that, okay, you can see last year, I think, in terms of other shipments in this moving in robotaxi sector, I think, the total shipment last year is only about 15,000 only, okay? This year, we estimate, I think, the growth rate will be, I think, I think close to, I think, 5 times, 5 times.

Even though this is really a small number, but, I will start saying, okay, this this type of, robotaxi sector is going to have a robust move robust growth, in 2 to 3 years' time. If you look at our our customers, this robot- robotaxi sector, okay, all those four big big guys big robotaxi company are our customers right now, including DiDi, Baidu, Pony, and WeRide.

Let me go go through all this one by one with you. For DiDi, okay, all the LiDARs right now installed in DiDi's robotaxi company, including the 4 main LiDAR and also the 6 blind spot LiDAR, all supplied by RoboSense.

Okay? Also, recently, we already get obtained the design win from Baidu.

In Baidu's next-generation robotaxi, they will adopt our main LiDAR, our EM4 LiDAR as their main LiDAR. You you I think you remember you managed to remember to mention it before, EM4 is a very very powerful LiDAR, long-range LiDAR, ultra-long-range LiDAR with a detection range of 3- 300 m, and also, in terms of number of laser beam, up to 520 laser beam.

Another big client is our is WeRide. WeRide, in their next-generation robotaxi G8R, they are still a high possibility that they will adopt our EM4 as their main LiDAR as well.

Also, they are using our E1 LiDAR as their blind spot LiDAR. Last, but not the least, I think, the Pony.

Pony, their 7.5th and also the 8th-generation robotaxi, they are also seriously considering, okay, to shift to our EM4 LiDAR as their main LiDAR as as as as the main LiDAR in their next-generation robotaxi vehicles. You will see, that's why I say, okay, we are we're right now in a very dominant, I think, position in the robotaxi sectors, all because of our in-house, self-developed chips, and also our more powerful LiDARs.

In terms of overseas robotaxi business expansion, I think, this is very sensitive, but what I can disclose to you is that, activities are progressing smoothly. We are talking to all those big robotaxi guys, okay, in in in overseas market.

I think, going forward, we will disclose more, I think, I think when we when we have a more appropriate, more method to disclose all this plan. I think, in terms of here, I don't think I can disclose to you, I think, right now.

What another point I want to mention to I also clarify to all the investors is that, okay, right now in my book, okay, all the robotaxi business, I I grouped them in the robotics and others category, okay? There's no separate line for the robotaxi, because right now the volume is still in the very lower level.

In 2 to 3 years' time, and I think the market and also the revenue contribution become a more meaningful level, I will come up with a separate line to to disclose more clearly, I think, the the revenue contribution from the robotaxi sector. Thank you.

[unclear] www.ir.robosense.ai. If you have any questions or would like to request for additional information, we encourage you to visit our Investor Relations website at www.ir.robosense.ai.

Thank you so much for everyone.

Thank you.