The commercialization of autonomous driving is entering a phase of divergence: what do Tesla’s controversy, Waymo’s expansion, and WeRide’s overseas deployments mean?

Title The commercialization of autonomous driving is entering a phase of divergence: what do Tesla’s controversy, Waymo’s expansion, and WeRide’s overseas rollout mean?

Introduction In May 2026, the global autonomous driving industry once again showed clear divergence. On one hand, Reuters raised new questions regarding the safety data and internal training processes behind Tesla Full Self-Driving (FSD); on the other hand, Waymo is moving its sixth-generation driverless system from testing toward broader citywide deployment. At the same time, WeRide continues to advance operational validation of its L4 Robobus in highly visible scenarios in Europe.

Together, these developments point to a larger question: competition in autonomous driving is shifting from “who can tell the most aggressive story first” to “who can find the balance between real-world operations, regulatory constraints, and commercial replicability.” For the global EV Industry, this is not only progress in smart vehicle technology, but also a restructuring of the electric vehicle platform, charging infrastructure, map and sensor supply chains, and fleet operation models.

Industry Context Autonomous driving has long been regarded as a core entry point for the deep integration of Electric Vehicles and Smart Mobility. Unlike traditional gasoline vehicles, electric vehicle platforms are naturally better suited to support advanced autonomous driving systems in terms of electric drive control, electrical/electronic architecture, software updates, and fleet dispatch. As a result, competition in the electric vehicle industry is no longer simply about range or battery capacity; it is gradually shifting toward industrial competition centered on “software-defined vehicles + data closed loops + scaled operations.”

From a global industrial chain perspective, the commercialization of autonomous driving affects at least five key areas:

• Vehicle platform: whether the architecture can accommodate sensors, computing units, and redundant systems;
• Battery and thermal management: energy consumption, stability, and durability during long-duration operation;
• Sensors and computing power: cameras, LiDAR, radar, and onboard computing platforms;
• Maps and data: HD maps, localized operational domains, and data training closed loops;
• Regulation and operations: road testing permits, liability boundaries, remote assistance, and safety certification.

Against this backdrop, Tesla, Waymo, and WeRide represent three different paths: a consumer-oriented route centered on software and data, a robotaxi route centered on closed operational domains and high-cost hardware, and a regional rollout route centered on commercial L4 public shuttle services.# Key Developments ## Tesla: Autonomous Driving Safety Narrative Comes Under Scrutiny Again Reuters’ investigation points out that Tesla FSD’s safety statistics methodology is controversial, with the core issue being that the company’s internal data and federal crash statistics use standards that are not fully comparable. The report also notes that some former data labelers expressed distrust in the system’s training and real-world performance, especially when dealing with emergency vehicles, construction zones, pedestrians, and complex road scenarios, where the system still shows insufficient handling of edge cases.

What is even more worth industry attention is that the report also shows Tesla carried out relatively concentrated high-definition map preparation and route-level localization work before some robotaxi demonstrations. This suggests that even a technology path emphasizing “pure vision” and “generalization capability” may still rely on region-specific data preparation and operational constraints when deployed.

For the industry, such controversies will not only affect one company’s public image, but also influence regulators’, partners’, and consumers’ judgment of the feasibility of unsupervised autonomous driving.

Waymo: Sixth-Generation System Continues to Expand Deployment Waymo has begun deploying its sixth-generation driverless system to its San Francisco and Los Angeles fleets, initially for employee and invited-user testing. The focus of this upgrade is not simply increasing the number of sensors, but improving system stability and scalability potential through more mature hardware integration, cost optimization, and weather adaptability.

Referenced reports show that Waymo’s sixth-generation system is based on Geely’s Zeekr EV platform, and it also plans to expand to the Hyundai Ioniq 5 platform. For the industrial chain, this indicates that robotaxi competition is no longer limited to software companies alone, but is gradually evolving into a systems-level competition involving vehicle platforms, sensor suppliers, cloud computing, and local regulatory coordination.

Of particular note, Waymo’s strengthening of sensor cleaning and perception capabilities in adverse weather reflects that once autonomous driving moves from the lab to city streets, the real challenge is often not “whether it can run,” but “whether it can keep running stably in complex weather, dirty environments, and high-frequency operations.”

WeRide: L4 Public Shuttle in Europe Continues to Validate the Business Model For the third consecutive year, WeRide has deployed its L4 Robobus at the Roland-Garros French Open in Paris as an autonomous shuttle service during the event. The project is managed with beti participating as the ground fleet operator, with fixed routes and clearly defined stops, making it a typical controlled commercialization scenario.

Although such use cases are not equivalent to large-scale public-road robotaxis, they are very important in business terms. For autonomous driving companies, events, campuses, airports, gated communities, and fixed-route buses are the best entry points for validating operational efficiency, passenger acceptance, fleet dispatching, and remote maintenance capabilities.From a global market perspective, WeRide’s continued expansion of pilot programs in France, Spain, Belgium, Switzerland, Slovakia, and other locations shows that the European market still prefers a “low-risk, gradual” approach to autonomous driving deployment, rather than opening up full-scale commercialization all at once.

Helm.ai and FORT Robotics: Infrastructure-layer gaps are being filled In addition to vehicle OEMs and operating companies, the foundational autonomous driving software and safety control layers are also evolving rapidly. Helm.ai has launched a next-generation generative simulation system, emphasizing higher-resolution multi-camera simulation to narrow the gap between training data and real-world sensors. This reflects the industry’s growing emphasis on simulation data quality rather than relying solely on road collection.

Meanwhile, FORT Robotics has acquired Mapless AI, strengthening its remote takeover and active safety platform capabilities. This move shows that, as autonomous driving shifts from proof of concept to real-world operations, remote monitoring, teleoperation, and safety fallback mechanisms are becoming important components of commercialization.

Industry Impact ## 1. The competitive focus in autonomous driving is shifting from “demos” to “verifiable operations” The moves by Tesla, Waymo, and WeRide show that the industry increasingly values evidence that is repeatable, auditable, and operational, rather than one-off demonstrations or promotional narratives. For regulators and city administrators, the key question is no longer “Can the system operate on a certain road segment?” but “Can the system remain stable, safe, and accountable over a longer period?”

2. The role of EV platforms is being redefined Electric vehicles are no longer just carriers of new energy; they are the physical infrastructure of autonomous driving systems. The appearance of Geely/Zeekr, Hyundai, and other platforms in Waymo’s deployment ecosystem shows that automakers still play a critical role in the autonomous driving ecosystem. In the future, whoever can provide EV platforms better suited for sensor placement, redundant braking, thermal management, and computing integration is more likely to gain the upper hand in the smart mobility market.

3. The sensor, compute, and data-training chain continues to benefit Waymo’s hardware upgrades, Helm.ai’s improved simulation capabilities, and the data labeling and training controversies exposed in the Tesla incident all show that autonomous driving still depends heavily on data quality and model training workflows. Suppliers of lidar, cameras, radar, onboard computing chips, data platforms, and simulation tools will remain key beneficiaries of the autonomous driving industry chain.## 4. The Importance of Fleet Operations and Remote Support Is Rising WeRide’s shuttle services for events and FORT Robotics’ remote takeover solutions show that the industry is increasingly relying on an “human-machine collaboration” operating model. Even if the goal is full autonomy, real-world commercial deployment still depends on remote supervision, emergency intervention, and scenario-based dispatch. This will drive the emergence of new service-based business models, including fleet management, remote operations centers, map updates, and safety review services.

Challenges And Risks First, regulatory uncertainty remains the biggest external variable in the commercialization of autonomous driving. Tesla’s dispute over safety data shows that if companies use incomparable metrics to explain safety externally, it may further trigger regulatory and public scrutiny.

Second, the cost of scaling autonomous driving remains far from low. Waymo’s technology upgrades emphasize lower-cost hardware, but actual deployment still involves expensive sensors, map preparation, fleet maintenance, and operational support. Even as system performance improves, the commercial return per mile still needs time to be proven.

Third, cross-regional deployment faces data and sovereignty issues. The Geely/Zeekr platform used by Waymo has already sparked discussions about data security in the U.S., indicating that autonomous driving is not just a technical issue, but also involves geopolitics, industrial security, and supply chain trust.

Fourth, the commercialization pace of the L4 path remains slow. WeRide’s expansion into Europe is highly representative, but current deployments are still largely concentrated in fixed routes, event scenarios, and localized operational domains. Truly large-scale expansion for urban public roads still requires clearer regulations, insurance, and accident liability frameworks.

Future Outlook Over the next period, the global development of autonomous driving and Smart Mobility will likely proceed along three parallel paths:

• Consumer-facing L2/L2+ will continue to improve, but safety responsibility and capability boundaries will receive greater attention;
• Robotaxi and L4 commercial operations will continue expanding in limited areas, prioritizing cities with more controllable weather, road conditions, and regulations;
• Scenarios such as public transportation, campus shuttles, logistics, and special-purpose vehicles will be the first to form clearer commercial loops.

For Battery Technology and Charging Infrastructure, this trend is equally important. High-frequency autonomous electric fleets will place higher demands on charging scheduling, fast-charging efficiency, station utilization, and battery durability than ordinary private cars. In other words, the commercialization of autonomous driving is not just a software issue; it will further drive systemic upgrades in charging networks, fleet energy management, and the battery supply chain.From a longer-term perspective, the industrial significance of autonomous driving does not lie in whether a single company wins the public narrative, but in whether it can truly drive Electric Mobility to shift from “selling cars” to “selling services,” from “single-vehicle intelligence” to “fleet intelligence,” and ultimately integrate more deeply with the energy transition, urban transportation governance, and a low-carbon transport system.

Conclusion The controversies surrounding Tesla, Waymo’s expansion, and WeRide’s rollout in Europe show that global autonomous driving has entered a more realistic and more complex stage: the technological narrative is still advancing, but commercialization must stand the test of operational data, regulatory scrutiny, and supply-chain constraints. For the EV Industry as a whole, this change will continue to reshape electric vehicle platforms, the sensor industry, charging networks, and fleet operations models, while also pushing the future development of global transportation electrification, industrial chain restructuring, infrastructure construction, and smart mobility toward a more mature direction.