Automakers are at the gates: vying for turnkey autonomous driving systems

The relationship in the autonomous driving supply chain is becoming increasingly complex. The core problem is that the cost of investing in technology research and development, verification testing, and mass production remains high and seems to be a "bottomless pit."

Tesla CEO Elon Musk has promised that by the end of 2020, there will be "a million self-driving taxis" on the road, taking advantage of Tesla's self-driving technology.

However, Musk soon admitted that his plans were sometimes wrong, and Tesla once again raised the price of the full self-driving feature by $1,000 starting November 1 this year.

Prior to this, Cruise, a subsidiary of General Motors, also announced the postponement of the Robotaxi commercial operation service originally scheduled for the end of this year. For Cruise, the focus now is to expand the mileage of its test fleet.

Although Waymo has started small-scale Robotaxi demonstration operations and launched driverless fleet services without safety officers, various aspects such as the driving area and the number of fleets are still in the early stages.

1. Robotaxi’s profit model is unknown

Public data shows that taking the taxi market in San Francisco, the United States as an example, the price of Robotaxi will be between US$1.58 and US$6.01 per mile, while the cost of privately owning a car is US$0.72.

Some staunch Robotaxi supporters believe that the cost can be calculated if driver costs are excluded and based on the longer operating time of the driverless fleet.

However, they overlooked the biggest question: where do the orders come from? Are the orders sufficient to support the operation of the fleet?

For most startups, it is not easy to be the "Didi" of the driverless era. You know, the current online car-hailing market is already oversupplied.

According to the latest released data on the average daily order volume of online ride-hailing services in Shenzhen from July 1 to September 30, each vehicle received approximately 7.2 orders, a month-on-month decrease of 22.5%.

Data shows that the average number of vehicles in Shenzhen that receive less than 10 orders (not including 10 orders) per day is 38,934, accounting for 55.4% of the total number of vehicles receiving orders.

The average daily order amount is 244.43 yuan. Such operating data is difficult to make a profit in second- and third-tier cities, let alone in first-tier cities like Shenzhen. (This means that the investment payback period of a car is as long as 3-5 years. If the cost of the autonomous driving system and daily maintenance costs are added, the time will be even longer.)

According to the data previously disclosed by Ford, the scrapping period of an autonomous vehicle is 4 years, which means that Robotaxi has almost no profit in the life cycle of a single vehicle. At present, the price of an autonomous vehicle that can be tested on the road is about 1.5-2.5 million yuan.

For automakers, except for Robotaxi, the launch of private self-driving cars will take several years, and the initial prices are unlikely to lead to large-scale sales.

This means that the current investment in the R&D of autonomous driving technology has a huge time cost risk. In the short term, "sharing the R&D costs" has become an inevitable choice.

2. Sharing R&D costs

Recently, a senior executive of Volkswagen said that the company is willing to share future self-driving car systems with other manufacturers to share the cost-intensive technology research and development investment.

This is the third step of Volkswagen's strategy proposed in the context of reaching an autonomous driving alliance with Ford (investing in the latter's autonomous driving subsidiary Argo AI) and restructuring the group's internal autonomous driving R&D system.

Volkswagen is not the first company to propose external sharing (selling technology), and this is also a development model that can be traced in the automotive industry. Back then, a group of Tier 1 companies were separated from the main manufacturers in order to reduce costs.

Prior to this, Volkswagen had already begun sharing its pure electric MEB platform with other automakers for vehicle development and mass production.

The development costs and huge investments of new platforms require other partners to join in to form economies of scale, thereby expanding the economic benefits of the platform and reducing production costs.

In addition, for other partners, Volkswagen's already established MEB platform will undoubtedly help them quickly launch electrified products, creating a win-win situation.

For example, GM has invested heavily in self-driving car development, particularly through its Cruise unit, and Honda's investment in Cruise is part of the two automakers' plan to collaborate on developing and manufacturing self-driving cars.

In recent years, almost every automaker has chosen to partner with one or more competitors to share costs and avoid technological uncertainties.

In February this year, BMW and Daimler announced for the first time that they would cooperate in the development of self-driving cars, and said that the two sides were studying the possibility of expanding cooperation.

Just five months ago, the two automakers issued a joint statement announcing that they would expand their cooperation and pool their development resources. The two companies hope to bring Level 4 autonomous passenger cars to the market in 2024, while also opening up to other OEMs and technology partners.

Prior to this, FCA became the first automaker to join the autonomous vehicle partnership formed by BMW, Intel and Mobileye. The goal of this alliance is to create a brand-agnostic autonomous driving platform that can be used by multiple car companies.

"Through cooperation, they have the potential to share costs, and even if they cannot fulfill their original plans in the short term, they can at least try to keep the losses at a low level," said an industry insider.

3. Fight for “turnkey solutions”

As this change becomes more and more obvious, the traditional upstream and downstream relationships in the supply chain are also adjusted and becoming more and more delicate. Among them, the startups that are preparing to sell autonomous driving systems to automakers are hit the hardest.

For example, three years ago, Volvo formed a joint venture with Swedish Tier 1 automotive parts supplier Autoliv to develop autonomous driving software and sell it to other automakers.

This is the first time an automaker has worked with a Tier 1 supplier to develop autonomous driving technology. Both parties will transfer the intellectual property rights of autonomous vehicles to the new joint venture.

This year's joint venture between Hyundai Motor and Aptiv is the second such case in the industry. The joint venture developed technology products such as a common platform for self-driving cars and sold them to other manufacturer customers.

According to the plan, the two parties will start testing of fully unmanned driving systems in 2020 and provide mass-producible autonomous driving systems to self-driving taxi suppliers, fleet operators and automakers in 2022.

According to Volkswagen's expected plan to establish a subsidiary, VWAT GmbH, to develop an L4 autonomous driving system (SDS) module, in addition to installing it in its own vehicles, Volkswagen hopes that the SDS module can be installed in vehicles of more brands.

This is also reflected in the changes in the roles of the industry chain in the development of autonomous driving systems (non-ADAS).

As more automakers deal directly with software development, hardware and semiconductor suppliers, the past strict Tier2-Tier1-OEM chain is changing.

Among them, some Tier 2 will transform into Tier 0.5, while the roles of OEM and Tier 1 are becoming more and more subtle. This changing environment is forcing all participants to rethink their roles in the supply chain.

In addition, technology giants (the most typical example is Waymo) are new members of the supply chain, and they will weaken the traditional supply chain by transferring their autonomous driving systems to other companies.

Automakers are increasingly working directly with traditional Tier 2 suppliers, such as semiconductor companies, which may squeeze Tier 1 suppliers out of the market (for example, in terms of electronic and electrical architecture, domain controllers, etc.) or squeeze out the value of the industry chain.

Now, the goal of automakers and alliances that independently develop autonomous driving systems, as well as powerful Tier 1 companies, is to deliver "turnkey solutions" for autonomous driving systems to automakers and travel platforms with weaker R&D capabilities.
This means that the market competition will become increasingly subtle. The "friend-enemy" relationship between upstream and downstream of the industry chain will become the main theme of the industry in the next few years.