This marks the 1397th original article from New Energy Frontier.

This article solely reflects the insights of 'New Energy Frontier' and does not serve as investment advice. The author does not manage investment groups, charge for stock tips, or handle client funds.

--------

It's time to reevaluate the prospects in autonomous driving, particularly for independent, specialized autonomous driving companies!

01 Beyond Just Leading New Energy Vehicle Manufacturers

Long-time readers know that New Energy Frontier firmly believes in the future potential of autonomous driving. We've repeatedly stated, "Before autonomous driving technology is truly commercialized on a large scale, there are limited investment opportunities in the complete vehicle sector. Only when autonomous driving becomes widespread will the business model of new energy vehicles undergo a transformation, ushering in a new wave of investment opportunities. At that point, a company with the potential to be ten times larger than Tesla and truly unique is likely to emerge." We've also posited, "Theoretically, the leader in autonomous driving can only be a new energy vehicle manufacturer."

These conclusions still hold true today. However, recent reflections have led us to a new perspective on investment opportunities in this field. There may be even more opportunities than initially anticipated.

If we adhere strictly to our previous line of thinking, we'd essentially be viewing the historic opportunities brought by autonomous driving through the lens of current vehicle manufacturers. Yet, in today's fiercely competitive environment, most vehicle manufacturers offer poor value propositions, and even industry leaders have their issues. Tesla is expanding its reach, but its automotive business faces increasing competition from Chinese manufacturers, raising questions about future sales guarantees. BYD, while leading globally in shipments, has struggled to break into the high-end market, lags in autonomous driving technology, and faces escalating management costs due to its massive scale. Smaller and newer players also have their own problems, with many facing uncertainties about their long-term survival.

Given our firm belief that, in the long run, the industry leader can only be a new energy vehicle manufacturer, we previously overlooked specialized autonomous driving solution providers. However, a deeper analysis of industry characteristics suggests that this conclusion is worth revisiting. Perhaps some specialized autonomous driving solution providers are worth considering at certain stages.

02 Is There Differentiation in Autonomous Driving?

Technologically, there are differences. Currently, the main approaches are pure vision and multi-sensor fusion. However, these routes are likely to converge eventually. Theoretically, from a long-term perspective, the pure vision route should dominate. Regardless of how technology evolves, from a consumer standpoint, autonomous driving lacks significant differentiation, or the differentiation is minimal and will approach zero as technology matures. After all, for consumers, they just want a solution to the driving problem, making it difficult for manufacturers to gain a premium through differentiation.

Especially in China's fiercely competitive environment, where autonomous driving does not constitute significant differentiation, it's challenging for autonomous driving to become a paid option. It can only enhance a model's competitiveness or overall pricing through standard inclusion.

While consumers may not perceive significant differentiation, the differences for autonomous driving manufacturers can be substantial, whether in terms of technological route selection (which will likely converge as technology matures), business model choices, or the resulting differences in corporate operations and management. Let's delve into these aspects.

For instance, pure vision technology solutions are cost-effective but relatively weaker in safety. Therefore, companies focusing on this route prioritize algorithm research, mainly in computer vision, deep learning, end-to-end systems, large models, etc., with relatively fewer teams dedicated to sensors and functional safety. Conversely, the multi-sensor route places greater emphasis on systems engineering, hardware-software integration, verification, and calibration, resembling interdisciplinary engineering teams and posing completely different challenges to corporate culture and management.

Business models also vary among different manufacturers.

Currently, there are three main business models for autonomous driving: First, as a pure solution provider, where autonomous driving solution providers charge downstream vehicle manufacturers for their solutions. Many independent autonomous driving companies operate this way. Second, operating robotaxi fleets, providing driverless taxi services, which is also pursued by many companies. Third, expanding into broader areas, such as not only passenger transport but also cargo transport, with even more imaginative possibilities in the future.

The first business model competes on the quality of technological solutions and business development (BD) capabilities, with no significant differences. The second model, as a robotaxi operator, involves substantial differences, such as deployment scope and management optimization. Driverless fleets also require management, including cleaning, maintenance, cost-effective charging, and the placement of management centers, all of which present numerous challenges. Different solutions can lead to vastly different outcomes. For example, well-optimized companies can significantly outperform competitors in terms of charging costs.

In short, the company that can prove its leading position in intelligent driving technology solutions while also achieving cost leadership in robotaxi fleet management is likely to become the industry leader. It can leverage its advantages to engage in price wars, rapidly capturing market share, potentially replicating Didi's success in the autonomous driving industry.

03 Opportunities for Independent Autonomous Driving Manufacturers

Previously, New Energy Frontier was concerned that independent autonomous driving manufacturers lacked an advantage in driving data compared to leading new energy vehicle manufacturers. However, with the gradual maturation of end-to-end technology, especially the rapid progress of AI, this data disadvantage is gradually being mitigated. Additionally, while leading new energy vehicle manufacturers still have an advantage in vehicle manufacturing, given the intense industry competition, finding cost-effective suppliers is not a significant issue.

Conversely, fleet management requires substantial time, giving early movers a significant advantage. Leading companies have comprehensive advantages in subsequent operational costs compared to latecomers and also benefit from scale advantages, with earlier deployment leading to more fleets and users. Of course, referring to the shared bicycle war, determining the industry leader will also require a fierce battle.

Furthermore, the market space is another reason to pay attention to independent autonomous driving manufacturers.

Currently, the daily order volume of the national taxi market (including taxis and ride-hailing services) exceeds 60 million. However, the combined daily orders for autonomous driving across the country are within 50,000, indicating significant room for penetration growth.

This means that even after autonomous driving technology gradually matures and new energy vehicle giants enter the market, independent autonomous driving manufacturers will have sufficient survival space for a considerable period, without needing to worry excessively about competition. Similar to the initial explosion of new energy vehicles, everyone first competed for the fuel-powered vehicle (internal combustion engine vehicle) market. It wasn't until penetration exceeded 40%, especially 50%, that new energy vehicle manufacturers began to compete with each other for market share.

Moreover, unlike the new energy vehicle market, which was already relatively saturated, where new energy vehicles competed for the existing market of internal combustion engine vehicles, autonomous driving's competition for taxi orders is somewhat different.

Theoretically, once driverless technology is mature enough, the number of private cars may decrease significantly (if autonomous driving is readily available and cost-effective, the necessity of owning a private car will greatly diminish. In reality, many private cars have very low utilization rates, with annual mileage not even exceeding 10,000 kilometers, making car ownership less essential). This implies that the daily travel order volume in China could potentially exceed 80 million or even 100 million in the long run.

With such vast potential for penetration growth, there's no need to worry about industry competition. In simple terms, for a considerable period, the growth in autonomous driving taxi orders will primarily depend on the supply side, not the demand side.

Considering all these factors, the few listed independent autonomous driving manufacturers (such as Pony.ai, WeRide, etc.) are worth studying, and New Energy Frontier will provide coverage on them in subsequent articles.