WeRide Inc. (WRD) Earnings Call Transcript
November 24, 2025
Earnings Call Speaker Segments
Thanks. Hi, everybody. Appreciate you being with us today. I want to kick things off by walking through a few of the major milestones from this past quarter. For WeRide, Q3 brought remarkable headway. The standout: in Abu Dhabi, we secured the first city-level fully driverless robotaxi commercial permit anywhere outside the United States — a historic first. And we're about to start — in fact, we already kicked off driverless operations via Uber, and I'll cover that in more detail on the slides later this week. Thanks to our recent move into Belgium and our first driverless robotaxi license in Switzerland, WeRide is now the sole company holding autonomous driving permits across 8 countries. By October, we had rolled out L4 fleets across 11 nations and more than 30 municipalities, and over 1,600 vehicles capable of L4 autonomous driving are now operating worldwide. Now let's turn to our third-quarter achievements. Let's bring up the Abu Dhabi slides. Okay. Could we see the next slide, please? As I mentioned earlier, WeRide has obtained formal clearance to operate a completely driverless commercial robotaxi service in Abu Dhabi, the capital of the UAE; this milestone approval eliminates the need for an in-car safety officer and reflects regulators' deep trust in our technology. Once that approval came through, WeRide and Uber jointly launched the area's first robotaxi service using fully driverless wireless charging, which kicked off this week on Yas Island and is now being extended throughout the city. Moving to the next page: our commercial operations in Abu Dhabi got underway last December. At present, our coverage reaches about half of the city's central zone. During a 12-hour period, each of our vehicles is capable of finishing as many as 20 trips daily. To me, that's pretty exciting headway. During the third midterm — sorry, for the midterm, our goal is to stretch service availability to 24/7, push vehicle utilization past 25 trips each day, and bring the human-to-vehicle ratio down to 1:10; with those numbers, unit economics become quite healthy. Abu Dhabi, in our view, will establish a worldwide benchmark for robotaxi operation that is both large-scale and commercially viable. Based on those figures, I'd say each unit's economics are very strong and genuine profit potential exists. What I especially want to emphasize is how exciting this sort of leap is — an entire year of enormous effort went into getting fully driverless robotaxi service up and running in Abu Dhabi. That marks the first city-level operation of its kind beyond the United States — the first city-level robotaxi service outside America, in fact — and it's delivered via the Uber platform. Taking all those key elements together, nothing else compares, and we're thrilled to be part of making history. Now, let's turn to what we're doing in Dubai at present. During September, we obtained a permit for self-driving vehicle trials from Dubai's Roads and Transport Authority, and road testing has started for our driverless operation there. What we're aiming for is to kick off a supervised trial on Uber within this year, with the driverless commercial operation set for 2026. Therefore, over the next year, we will roll out a driverless robotaxi service in Dubai. Let's turn to the following page. Following that, I want to cover our current activities in Saudi Arabia. In Riyadh, we began providing robotaxi trips through Uber back in October. That achievement meant our service became the first and only robotaxi option in the Kingdom that anyone can access. Thanks to the strides we've made across the Middle East's 3 biggest cities — Riyadh, Abu Dhabi and Dubai — we currently operate more than 100 robotaxi vehicles throughout the Mid East region. Getting driverless service up and running in Abu Dhabi is paving the way for us to expand the fleet past 500 vehicles by next year and reach tens of thousands by 2030. So we're incredibly excited and incredibly confident — and incredibly confident and incredibly optimistic about our fully driverless robotaxi operation in the Middle East. However, those aren't the sole markets where we operate. Let me move on to discuss East Asia along with Europe. First off, in Singapore, working together with Grab, we secured approval from the Land Transport Authority covering both robotaxi and robobus services in what's called the Punggol District. Our goal is to quadruple the amount of AV testing we conduct before this year ends. Beyond that, we're integrating our technology into Grab's fleet management and routing systems, which sets us up to eventually provide driverless robotaxi rides via Grab in Singapore — precisely the same thing we've already pulled off with Uber in Abu Dhabi. In fact, this — every one of these initiatives will serve as the foundation for commercial service during the following phase. Then let's turn to Switzerland; in Europe, our expansion in Switzerland keeps leading our robotaxi deployment. We got the country's first driverless robotaxi license, which enables our autonomous operation in the Furttal region. A fully driverless public service is anticipated to launch in the first half of 2026. That covers our current operation in Europe and in East Asia and Singapore. Now let's discuss China; in our China market, we keep expanding and innovating. As we scale our commercial fleet, we also introduced a 24/7 driverless commercial service in Huangpu District at Guangzhou. This covers an area of 150 square kilometers. As of October, we've deployed over 300 robotaxi in Guangzhou and more than 100 in Beijing. Through our WeRide Go app, you can summon a driverless robotaxi for any of these services in this area. Keeping user value at the center of everything we do, we recently introduced PU/DO service — China's first complimentary pickup and drop-off option for robotaxi rides — which allows the system to suggest the best boarding spots on its own. This greatly improved operational flexibility as well as the user experience, and our operational data reflects that clearly: in November, robotaxis in Guangzhou reached as many as 25 trips per day each, while those in Beijing hit 23, a clear sign that adoption is speeding up. That wraps up the exciting progress we've made on robotaxi. Now I'll move on to our other applications. Starting with robobus. Okay. Let's go to the next page. Belgium issued its first Level 4 test permit to our robobus, and we began running services in Leuven — that brings the total number of countries where our service operates to 11. Back in 2021, we started carrying passengers in Guangzhou, and we've now moved more than 1 million of them, which led to a request for an additional 100 midsized robobuses. It's a genuinely exciting achievement. The robobus itself, in fact, is a brand-new design; meanwhile, in Hong Kong, we signed a partnership with Kwoon Chung Bus Holding that will see over 500 Level 4 vehicles deployed during the next 3 years. Turning to our L2+ Level ADAS system, WeRide and Bosch reached a major milestone in November, when WePilot 3.0 entered production. This is a complete, integrated system, quite similar to the Full Self-Driving solution Tesla has built. Our WePilot stands on equal footing with Tesla's FSD. WePilot 3.0 is set to launch alongside the updated Chery EXEED ES and ET, and those who already own these vehicles will receive over-the-air updates. With this new capability, every Chery EXEED ES and ET owner can enjoy the same kind of experience Tesla's FSD provides.
Okay, so thanks to this really strong end-to-end setup, WePilot was picked by Guangzhou Automotive Group, GAC, to serve as the primary ADAS system supplier for a number of their passenger car models. That development is genuinely exciting, because it proves WeRide isn't just able to handle L4 robotaxis — we can also deliver L2++ ADAS for cars built at mass scale. And this page gives you a snapshot of where we stand around the world. Our approach at WeRide has always put balanced growth across global markets first. Let me walk you through this slide a bit. As you can see, across these 11 countries, our operations sit at varying levels. In some we've run tests, in others we can run fully driverless. The legend lays that out. What this shows is that offering multiple products has extracted the most value from our strategy, and it makes us the sole company whose technology is deployed across all 11 countries you see here. That gives us a broad range of applications and services for markets worldwide. Now I'd like to talk about what underpins our technology — the WeRide One Universal platform. In the early days it only supported L4 applications, but WeRide One has steadily evolved into something far more capable, powering everything from L2 through L4, all while continually generating new tools and systems. Among the most notable is our world model, WeRide Genesis. With Genesis, you're looking at a brand-new platform that lets autonomous vehicles be validated inside a digital twin of the physical world — safely, efficiently, and at large scale. It brings together the data loop, the algorithm loop, and the simulation validation loops, all of which are critical if autonomy is to scale. That's our world model, and it plugs seamlessly into our end-to-end system. This is a distinctive technological edge. Our world model integrates smoothly with the end-to-end ADAS system, and our L4 system is able to draw on data gathered at the L2 level. So Genesis functions as the central flywheel — really, what we describe as a double flywheel. We're able to take massive amounts of data from L2++ production cars to sharpen our robotaxi, and conversely, the redundant data from our robotaxis helps lift the performance of our ADAS system. I want to stress that in this world, WeRide is the only company capable of this. On one side, our large-scale robotaxi fleet lets WeRide deliver leading robotaxi service the way Waymo and others do in the U.S., which means we can capture every characteristic of fully driverless operation. On the other side, WeRide applies < > 3.0 and a highly advanced ADAS system on par with Tesla's FSD, and we can draw on mass production car data, gathering it broadly across every kind of scenario, to help raise robotaxi performance. Our view is that the strengths of L4 together with L2+ can be brought together. Adaptability, reliability, safety, and transparency all get a boost from this hybrid architecture, and that eventually makes robust commercial deployment possible. We're eager to tell you more about this advantage before long. To wrap up, Q3 was a quarter where execution was exceptional. Our global leadership grew, and we turned technology innovation into commercial reality; so with that, I'll pass things along to our CFO, Jennifer, so she can go over our financial performance. Jennifer, please go ahead and talk through the financial numbers.
Thanks, Tony. Hi, everybody. Before we get into the Q3 numbers, I'd like to point out that every figure is denominated in RMB, and any comparison is against the same period last year unless I say otherwise. So let's turn to how we performed financially in the third quarter. Total revenue came in at RMB 171 million, up 144% from a year earlier, which reflects our ongoing expansion of the fleet as well as greater penetration of services. That top-line increase also speaks to a major milestone reached in this quarter, made possible by our sophisticated technology along with strong deployment and operational capabilities. Our top line was split between product revenue and service revenue. Product revenue posted robust growth of 428%, reaching RMB 79 million in the quarter — an encouraging outcome fueled by higher sales of our robotaxi and robobuses. Service revenue rose 67% to RMB 92 million in Q3, helped by a RMB 29 million pickup from intelligent data service plus a RMB 8 million gain in autonomous driving-related operational and technical services. In this quarter, service revenue moved ahead of product revenue, which shows sustained growth momentum and a sound business structure. Within our product lines, the standout in Q3 — just as it was across the last 2 quarters — was our robotaxi businesses. Robotaxi revenue climbed 761% year-over-year to RMB 35 million in Q3, making up 21% of total revenue for the quarter. Thanks to our new federal permits in UAE, we are the first and only robotaxi company to have started fully driverless robotaxi operations in UAE. Taking the in-car safety officer out represents a key milestone financially, and it will let our robotaxi service reach unit economic breakeven. The quality of our growth is compelling as well. Gross profit at the group level rose 1,124% to RMB 56 million for the third quarter, with a group level gross margin of 33%, which shows our industry-leading gross margin as our business keeps expanding. Our aim is to keep delivering business value alongside our globalization strategy. Operating expense fell 51% to RMB 436 million, with R&D expense making up 73% of total operating expenses. Breaking that down further, R&D expense rose by 24% to RMB 316 million in the third quarter of 2025 versus the same period of 2024. If we exclude share-based compensation, R&D expense was up 39% to RMB 288 million, as we further bolstered our global data compliance and advanced R&D work for our pre-installed robotaxi. The rise in R&D expense mainly came from a RMB 31 million increase in service fee for R&D projects, a RMB 21 million increase in personnel-related expense due to headcount growth, and a RMB 23 million increase in material consumption and depreciation and amortization expenses. Administrative expense dropped by 84% to RMB 100 million in the third quarter of 2025 compared with the same period in 2024. Excluding share-based compensation, administrative expense rose by 23% to RMB 74 million. The rise came mostly from a RMB 6 million bump in professional service fee, largely tied to legal compliance service, plus RMB 4 million more in personnel costs, since we keep constructing the supporting function needed to expand our business. In the third quarter of 2025, selling expenses rose 23%, reaching RMB 19 million, versus the same period of 2024. If share-based compensation is left out, selling expense climbed 36% to RMB 19 million, a figure far under the growth in sales. R&D dedication forms the core of what we do. We will keep steering our resource in that direction so we can lead industry innovation and go on sharpening our competitive advantage. At the same time, we will deliberately expand our global team, concentrating clearly on region where adoption of L4 solutions has sped up. That guarantees we possess the world-class talent required to back our business expansion. Our net loss shrank 71%, landing at RMB 307 million in the third quarter of 2025. Under non-IFRS, adjusted net loss grew 15% to RMB 276 million, chiefly because of continuing R&D investment and the wider operational support our business expansion demands. At September 30, 2025, our holdings included RMB 4.5 billion in cash and cash equivalents and time deposits, RMB 926 million in investment in wealth management products, and RMB 18 million in restricted cash. Short-term bank borrowing stood at RMB 245 million. Thanks to our current liquidity reserve and the proceeds from our recent Hong Kong due primary listing in November, we now hold a resilient position for our R&D-focused strategy and our globalization deployment process. Securing our fully driverless robotaxi commercial permitting in Abu Dhabi counts as more than a local milestone; it is a scalable blueprint for the global industry. It shows a viable path toward city level full driverless cooperation outside the U.S., as well as the potential for profitable unit economics in major international markets. Scaling this model globally is our strategy. The complete package is ours: the technology, the operational experience, a proven safety record and regulatory trust. Over the next 5 years, we will bring about large-scale L4 deployment, building a sustainable business and handing tremendous value of autonomous driving to the shareholders. With that, operator, we are now ready to take on some questions.
Tim here, from Morgan Stanley. Nice work on the impressive numbers and the ongoing global growth of robotaxi operations. Two things I'd like to ask. First, we saw that WeRide has formally kicked off commercial driverless robotaxi service in Abu Dhabi, in the UAE. Beyond the added volume for revised fleet sales that Tony referred to earlier, how do we think about sizing the long-term revenue potential from vehicle sales, revenue charges, and profit sharing? That's question one.
All right, thanks Tim. Good question. I'll handle the first one. For everyone listening, let me quickly walk through how our robotaxi business model works; inside China, we mostly own and run the vehicles ourselves, on our own ride-hailing platform, WeRide Go. So once the UE reaches breakeven in China over the coming years, we'll bring in third-party owners step by step and form partnerships with them. As things stand, we own essentially all of the vehicles ourselves. The international market works differently. Starting on day 1, we work alongside platform partners such as Uber, Grab, SBB, TXAI, and money comes in through 3 main streams — 3 streams. Stream one is a share of the ride fare, stream two is the yearly licensing, and stream three is selling the vehicle. So the vehicle sale counts as product revenue. WeRide is able to grow the robotaxi fleet far faster under a lighter business model — more asset-light, since the operating robotaxi fleet isn't held on our own balance sheet. We simply sell it to our partner. And the revenue share plus annual licensing make up recurring service revenue across the vehicle's entire lifespan, typically 5 to 7 years. In particular, as fleet size expands, revenue share turns into a major multiplier. Take the United States as an example. A robotaxi running at human-level utilization, meaning about 25 orders completed each day, can bring in yearly revenue of more than USD 90,000 on the platform. Should WeRide capture 30% of the revenue share, that gives us USD 30,000 per car each year as service — as a revenue share. Should we capture 70% of the revenue share, that gives us USD 60,000 per car each year. So if we move one step closer to that goal — and look at what we already have, say, in Abu Dhabi, at present we hold a significant presence with close to 100 robotaxi in Abu Dhabi; we now already cover 50% of the city core area. And the commercial model has us integrating on a platform like Uber. At the moment, we charge at the same price level as UberX and Uber Comfort. Actually, suppose you open the Uber app and request a robo — or simply hail a standard ride-hailing vehicle across 50% of the city; whether you choose UberX, Uber Comfort, or the autonomous choice, a car will arrive for you. That proves from day 1 our offering matches mainstream ride-hailing on price competitiveness. Turning to unit economics, utilization stands as the single most important measure. At present we've hit an average of 12 orders per vehicle daily during a 12-hour shift. On favorable days, we occasionally exceed 20 orders per vehicle each day within that same 12-hour shift, which already signals robust user preference and stickiness. Just so you know, at 12 orders per vehicle per day, we've already crossed the breakeven point in this market, meaning enormous profit potential exists. Given our present driverless cost structure and our plan to expand operating hours to 24 hours next year, we forecast average daily orders will hit 25 per vehicle per day. Such utilization levels should produce very strong profitability potential next year. As for the exact revenue-share percentage, that stays confidential among the various partners — yet this model can foster a lasting win-win partnership for all participants in the ecosystem.
For my follow-up, I'd like to ask about WeRide's business around the world. Looking ahead, apart from the operations in Abu Dhabi and Switzerland that were just announced, which markets might emerge as major volume drivers for WeRide? Also, to fund the company's strong overseas expansion, does WeRide need to speed up R&D and sales efforts and spend more aggressively as we head into 2026? That's my second question.
All right. I'll handle that one. First off, aside from Abu Dhabi and Switzerland, which other markets could turn into big sources of volume? For our part, our approach starts in the Middle East, where we have two key cities in the UAE — Dubai and Abu Dhabi. At this point we've already secured the permit, and we're doing a lot of road testing. The UAE is definitely a country we're putting a lot of emphasis on. It's a really significant market. Saudi Arabia is in the picture as well, so those are the main Middle Eastern nations you can point to. That puts them among the possible countries. Europe too, plus other developed Asian nations like Japan, Singapore, and Korea — all of those are possible markets that could boost our volume. But here's the thing I want to highlight: your question is really good, and honestly it's something I've kept pondering for years — namely, where should we aim, and in which markets can our products and services turn a solid profit? Based on what we've tried, based on how we operate in Abu Dhabi, we've identified what we call the Abu Dhabi model, right? Working alongside Uber, what we discovered is that the per-ride economics hold up well, and their forecast for us looks quite encouraging — suggesting that before long, in this market, we'll be able to run a service that performs strongly and delivers solid margins. The approach we used in Abu Dhabi effectively laid out a blueprint that other cities can follow. So together with Uber, my expectation is that we'll attempt to replicate this style of model in comparable cities. This blends what we already have technology-wise with our overall approach, and it ties in the work we're doing alongside our strategic partner. Using that framework, we plan to replicate the setup in Singapore through one more partnership with Grab, a strategic collaborator of real importance to us there and possibly throughout East Asia. As for China, our efforts center on building out a robo testing service that runs on our own application. On the expansion front, we're putting more into R&D so we can build stronger technology platforms, and we're also going to work hard at bringing in top talent. That said, we anticipate related expansion will grow only moderately, since our plan is to use a satellite model that balances scaling against investment. One of our near-term goals is to reach profitability while holding a solid market share. Past that, innovation stays a priority for us, which means we have to balance spending, investment, and development. Looking at how things stand today, I'm fairly optimistic, because the Abu Dhabi model has already shown it works. The next step involves identifying every location where that Abu Dhabi approach can be replicated with ease, and by pooling the strengths of those prospective markets, our aim is to reach profitability soon.
Thanks a lot for providing such valuable perspective, congratulations, and I'm eager to see additional thrilling initiatives worldwide.
Two things I'd like to ask. First, how do you view the robotaxi sector in China? In what ways do you expect the economics to evolve going forward, separately for China and for markets abroad? My second inquiry concerns how soon the driverless achievement from the Abu Dhabi operation could be duplicated elsewhere. What should we anticipate regarding your worldwide fleet growth strategy? Also, which catalysts or obstacles stand in the way of that strategy?
Alright. I'll attempt to address them sequentially. Even though you framed it as 2 questions, it's really 2 sets of questions. Regarding the first one, if my memory serves, it concerns roughly our — your perspective on the Chinese market, what we intend to do there, and our views on its economics. So to begin, I've discussed extensively the worldwide market, Mid East, Asian market — Mid East, Asia market, East Asian market, European market. And — yet what about China, right? China, where our headquarters sits. China is unquestionably a significant market; and we also — certainly China ranks among the most vital markets we are aiming at, correct? Now allow me to explain the advantages and disadvantages of allocating resources to the Chinese market and what our approach looks like. So first, China stands out as a distinctive market boasting the biggest user base along with dynamic economics. Additionally, it serves as an excellent testing environment. It has functioned as both our technology proving ground and an ideal operation — sandbox for our highly inventive concepts. Naturally, though, that doesn't imply we regard it merely as a lab, alright? We aim to — as we conduct robotaxi operation or trial operation, we maintain safety as our foremost concern. Yet despite that, with every kind of scenario, every kind of different climate, varying weather conditions, China is an enormous country. And we actually — we examined so many distinct methods, distinct algorithms. So in that regard, actually, China is distinctive and a very sizable market. However, we believe — yet China is also quite — when it comes to development, it isn't especially balanced; there exist Tier 1 cities resembling Paris and New York. There are Tier 4 cities resembling rural areas. Thus we believe profitability within Tier 1 cities may be attained through a mix of 3 components. First, city level drivers per meet; second, average daily order of high double digits. That means what we strive to accomplish in Mid East ought to be accomplished in China as well, exceeding 20 orders each day. Third, a somewhat healthy price, alright? Even though currently, taxi fares in China remain comparatively low, we anticipate the fare will increase somewhat; so it stays somewhat healthy. Given those 3 elements, our view is that we still must grow our presence in China, especially across Tier 1 cities. To this point, as I noted earlier, we've reached 300 robotaxi in Guangzhou, plus another 100 robotaxi — over 100 robotaxi in Beijing, covering an area of 150 kilometers — and we keep expanding that footprint. We've also rolled out PU/DO, meaning pickup and drop-off capabilities, which assist us in enhancing what users experience. Our goal, I'd say, is to deliver a robotaxi service superior to the conventional taxi service provided by a human driver. That way, we can secure additional orders and deliver a better user experience, allowing us to be ordered more often. We anticipate that the economics across all these markets will get better with time. We also want to genuinely take what we have in China and expand it — taking what we've learned in China and applying it as our competitive edge worldwide. So I'd say we continue to regard China as among the most important markets, and we'll keep investing and putting resources into this market. Alright. The second group question concerns how fast we can replicate the Abu Dhabi model across the rest of the world, alright? To begin with, I appreciate this question. We think we've identified this sort of Abu Dhabi model. It resembles a Matthew Effect. WeRide holds a somewhat unique first-mover position since we've just obtained a city level driver permit. Outside the United States, it's the only one so far where you can hold a citywide driver permit and deliver the service via Uber. That makes the service model < > available. So I believe we can rapidly replicate it in comparable markets such as Dubai in UAE, the same country, Riyadh in Saudi and the same region, and Singapore with backing from Grab; thus we're working to replicate in these sorts of countries. I think the regulatory situation and all other elements are fairly comparable. We'd also like to stress that Europe is a highly important market, and we're exploring whether we can replicate into Europe. As for the catalyst hurdles, the catalyst stems from this Matthew effect, since other countries are more cautious about permitting our operation, more cautious about granting us a permit. Yet simultaneously, the hurdles remain the regulatory matters; we want to ensure we use our — leverage our existing successful experience to obtain more driverless permit so we can deploy the service. That's everything I wish to say regarding this question.
I've also got two questions. The first one is this: we keep hearing about additional OEMs plus ride-hailing firms saying they intend to get into the robotaxi space. What makes WeRide stand out, and how ought we to view the way competition will shape up down the road? That's my first one.
All right. What I'd say is that right now, given how much more robotaxi gets talked about and how much more mature it keeps becoming, you've got a large number of car OEMs and platforms — car hailing platforms — beginning to discuss robotaxi or beginning to put out their robotaxi strategy. Still, there's one point I'd like to make: doing robotaxi is not simple, all right? It demands many years of effort, of building up technology, of exploring regulation. That explains why only so few mature robotaxi companies exist worldwide. If you tally mature robotaxi service providers, meaning those that have opened driverless robotaxi operation to the public, I believe you could count most of 3 or 4, all right? So there aren't that many. So it isn't the case that when you notice a handful of companies reaching maturity, you can then put out your strategy — rather, you need to demonstrate whether you possess sufficient technology accumulation and sufficient experience to run robotaxi. So our — WeRide's edge remains in a few areas. To start with, technology, correct? At WeRide, the capacity to deploy at scale both L4 and L2++ level mass production vehicles assists us in actually, first, obtaining data — gathering data with greater efficiency and rendering our algorithm more generalizable. That, in turn, genuinely reinforces our technology. Next, our ability for quick iteration is present. Consider that for < > OEMs, they typically maintain a comparatively small ADAS system development team. Whatever you can accomplish for L2++ FSD — an L2++ ADAS system — falls far short of what's achievable at L4, since L4 means driverless operation. L2 is merely something like an assisted driving system. With an L2 system, you aren't required to bear final responsibility. With an L4 system, though, you must be redundant and must bear that responsibility. Across the past 9 years, we've built up a great deal of experience. We — that's the reason we're able to roll out robotaxi service. And I have yet to see any other car OEMs or car hailing platform manage it. So that's one of the advantages we hold. And one more thing — the final point I'd stress concerns what sits at our company's core. What sits at WeRide's core is truly AI technology. From day 1, WeRide has been an AI company, and we bring on so many top talents and build our company around fast iteration in AI algorithms. I don't believe traditional car OEMs or traditional car hailing platforms can manage this sort of quicker iteration. So let's wait and see. Up to this point, though, I have not witnessed any significant automaker or ride-hailing platform successfully launch a robotaxi service — a driverless robotaxi service — to the general public, all right? That is how I would respond to those 2 questions.
I apologize, but I have an additional inquiry. Building on what was just asked, do you believe that the volume of data along with the progress of AI models has granted OEMs a particular opportunity to get into the robotaxi space and compete? Could a transition from L2 all the way to L4 actually happen?
That's a great question. To start, let's consider something. Who holds the title of the top L2++ or ADAS system company globally? Most likely Tesla. Within China, a handful come to mind, perhaps XPeng and Li Auto. Yet, examining their approach, they're pursuing L2++ ADAS systems and mention robotaxi, but once robotaxi is the focus, they consistently attempt to tackle this challenge or pursue this endeavor straight from the L4 tier. Why is there no L3 approach? Where's the L3 approach? Tesla has no L3 approach. There's no — XPeng and Li Auto, they bypassed L3 entirely. What's the reason? Since if you move straight from L2++ up to L3 and afterward to L4, they discovered it's extremely, extremely hard. It's comparable to scaling a sheer cliff. Alternatively, you might just address the issue head-on. That means employing your expertise to directly resolve the L4 system-level challenge; that's exactly what we've already accomplished over the last 8 or 9 years. The tech has existed; we've employed deep learning algorithms grounded in large language models, massive amounts of data. However, all this, I'd like to note, rests on our experience from the previous 8 years. At present, actual car OEMs can tap into the most advanced large language model technologies, but numerous infrastructures are involved, such as data simulation and the cloud computing platform. All this, I don't believe car OEMs possess sufficient accumulated expertise. It will still require many years for them to genuinely introduce a simulation platform, to introduce the protocol, to introduce the pipeline for testing the driverless robotaxi. One point I'd like to stress, possessing a fairly decent ADAS system that enables you to travel for 100 miles without takeover is nowhere near enough to launch a driverless robotaxi. To launch a driverless robotaxi, you must be able to ensure that vehicle drives itself beyond 10,000 miles. So that represents the scale of the challenges. It's similar to swimming in a pool and then wanting to swim across the English Channel, that's a different matter. So I think I'm not saying — I won't claim it's completely impossible to progressively advance from L2++ to L4, but it will require a long, long time. Prior to that, I believe first class, the Tier 1 robotaxi company like WeRide will have already captured the global market and already become very profitable. So the time remaining for these major car OEMs to progressively advance from L2++ to L4 is very, very limited.
To start, I'd like to build on the technical question that came before. This one's directed at Tony, since your confidence about keeping the industry lead is notable. Which tools and technological methods keep you ahead? Could you give us additional detail on the technical side? After that, I have a follow-up.
All right. To begin, as I mentioned earlier, to the best of my knowledge there is just a single firm globally that can do robotaxi, has already brought its operation to open to public status, and simultaneously supplies ADAS systems to a mass production car company — and that's WeRide. Thus we follow what we call a dual flywheel strategy: we take the data gathered by our robotaxi fleet, every sort of corner case, and apply it to speed up our L2++ work. In parallel, whatever we collect from the L2++ system, such as AI drive based on navigation, we feed back so our L4 system becomes more stable and more generalizable. By merging these 2 data sources and 2 problem sources, we steadily grow into a super platform able to handle robotaxi while running on limited hardware and limited HD map and navigation map, which lets us reach the whole global market. These 2 sides reinforce one another, helping us enhance and iterate our algorithm at a pace no single-strategy company could match, all right? That's one of our advantages. Another point is globalization. Because we've put fleets into the global market, we're able to gather data and obtain heterogeneous data sources from everywhere on earth. We can have data gathered in the extremely dry climate of Middle East. At times we get data from the highly humid tropical region of Singapore and from very cold places in Japan and in China. I cannot picture any other company holding this sort of wide spectrum of application area. With all this data plus the talented engineers we've brought on, we in fact progress very quickly. One more thing I'd stress is our Genesis platform. It relies on a physical AI model and takes into account a great deal regarding a real world model, physical AI world. It also integrates seamlessly with our end-to-end system; a simulation platform like this hands us a major edge in development. And once all these algorithms and data come together, I believe this massive effect is present. We hold many, many driverless permits across the world in many countries, and many car OEMs work with us. We have heterogeneous data collected from different country, different level of autonomy. So when we put all these pieces together, my response to what you asked is that I believe we're able to reach a quicker pace of development—one that's significantly faster than what our competitors manage.
That's quite useful. My next question goes to Jennifer. In May of this year, the company's Board authorized a share repurchase program of USD 1 million. Could you give us an update on where that program stands?
Thanks, Liping. About the USD 100 million share repurchase program that the Board authorized in May — so far, no purchases have been initiated. The reason is that the preparatory work for our Hong Kong IPO put us in a close period under securities regulation, and trading was restricted during that time. As a listed company, we also need to secure specific shareholder approval to ratify this program. Right now, we're getting ready to call an extraordinary general meeting to seek that approval, which will let the program move forward. Thank you.
Paul Gong with UBS here, and I've got two things to ask. First, on the revenue that robotaxis bring in: we've seen roughly 7x growth compared with a year earlier, yet next to Q2 there appears to be some fluctuation. Would you walk us through that a bit more? Second, about the approach in Europe — congratulations on the recent Swiss permit. Could you tell us what comes next in the company's plans for growing across Europe?
Alright. I'll handle the first one, and I believe Tony will likely field the second. On the robotaxi revenue swinging around, here's what I'd say: across the last 3 quarters, robotaxi revenue contributed approximately 20% — 22% in Q1, then 36%, and 21% of revenue over the two most recent quarters, which demonstrates quite a steady momentum already. Some fluctuation was to be expected, since our delivery timetable moves in tandem with permit upgrades and the matching enlargement of where we operate. Securing the city level driverless operation permit in Abu Dhabi marked a major step forward, and it will open the door to faster expansion throughout the Middle East from here on. Thank you. Tony, would you like to field the second one?
Yes, I'll address the second question. That one concerns our — well, actually our — given that we've already secured the driverless robotaxi < > Switzerland, what comes next for Europe. Alright. So when it comes to robotaxi, Europe is an excellent market. In my view, taxi fares across Europe run high, and moreover, throughout most European countries there's a labor shortage; that situation genuinely makes for a very, very favorable setup or opening for a robotaxi operator to roll out robotaxi service and bridge the gap between too few drivers and rising taxi demand. As we know from recent figures, people — particularly post-coronavirus — prefer private transportation whenever they can. So I'd say demand for our taxi is truly climbing. That means if robotaxi can be rolled out in a highly cost-effective way, customers will embrace this type of offering. So because we've already achieved this in Switzerland, right now we're looking at additional countries — for instance, we opened an office in Stuttgart and are also working to explore Paris. So over the coming 12 months, we'll actually strengthen our base through our trial operation in France, in Belgium, along with our existing operation in Switzerland, and hold talks with every country we possibly can. We've further built a strategic partnership with Uber, with Renault, with SBB, STL, and so on. Beyond that, plenty of airports across European nations are in discussions with us about exploring what's possible with our robobus; so our aim is to leverage all our applications, such as robobus and Robosweeper, to genuinely assist us in exploring the potential of extending into particular countries, since in certain European countries, the preference tends to be adopting robobus or Robosweeper first, and afterward attempting robotaxi. Step by step, we intend to reach Switzerland, Belgium, Germany, France, Spain and Norway — these sorts of countries — before extending to even more. Then, naturally, I'd like to stress that our approach is dynamic. It hinges on whether a strong local partner is available, and it also hinges on regulatory policy and the local — shortage of labours; every one of these factors must be weighed; yet certainly, we will step by step expand into the countries named above and every other potential country throughout Europe.
Thank you very much. Thank you. Bye.
Read the full transcript via the API
You're viewing the first half of this call. Get the complete WeRide Inc. transcript - plus 132,000+ transcripts from 7,000+ companies, speaker segments and full-text search - through the EarningsAPI REST API or hosted MCP server.
Get an API key View API docs →For developers and AI pipelines
Programmatic access to WeRide Inc. earnings transcripts and 132,000+ others is available through the
EarningsAPI REST API and the hosted MCP server.
Quarterly plans from $162 - full transcripts, speaker segments, full-text search,
and the /api/v1/transcripts/recent polling endpoint for ETL pipelines.