From Honda's release of the world's first L3, how can we see the competition among major autonomous driving countries?

In recent years, when delays in the implementation of autonomous driving (ADAS) have become the norm, Honda has become a breath of fresh air.

Honda promised to deploy L3 autonomous driving cars last year, and it was initially fulfilled on March 5 this year. Honda launched the "Legend" model equipped with the L3 autonomous driving system (Honda SENSING Elite) in China, priced at about RMB 660,000. The first batch was a bit small, with only 100 vehicles.

Different temperatures

The difference between L3 and L2 is not that the applicable scenarios have become wider, but that the control cannot be frequently and unconditionally handed over to humans. That is, even if humans do not respond, safety can still be guaranteed (the system will choose to pull over).

Before Honda, there were already dozens of OEMs claiming to have L3 or higher technology, but the regulations do not allow it. When saying these things, they also have to express the regret that "their skills cannot be fully utilized".

Japanese OEMs also follow the logic of "I want to do it but you don't let me", which is equivalent to kicking the ball to the government.

Officially, Honda is confident. It claims that the safety and reliability of this system has been simulated in about 10 million different states and has a test mileage of more than 1.3 million kilometers. It also adds a redundant system (also known as a backup system) to ensure that even if some parts have problems, the overall operation is still safe.

Unexpectedly, the Ministry of Land, Infrastructure, Transport and Tourism (equivalent to China's Ministry of Transport) directly approved it. According to the "Guidelines for Safety Technology of Autonomous Driving Vehicles" compiled by the ministry in 2018, the Honda Legend belongs to "conditional automated driving" and should be allowed to go on the road.

However, traffic laws, regulatory frameworks, insurance, etc. have not been implemented. When the Legend is on the road, it is actually still a "road test", except that there is no electronic fence set up.

Interestingly, Honda's statement emphasized that "the Ministry of Land, Infrastructure, Transport and Tourism of Japan has approved mass production." The Ministry of Land, Infrastructure, Transport and Tourism vaguely stated that Honda's products "may help" improve traffic conditions, and emphasized that these vehicles have obvious declaration signs on the rear.

Semiconductor dominance changes hands in five years

As the first major automobile country to approve L3 vehicles for use on the road, is Japan the result of a natural development of technology, or is it just a forced "overtaking" and a forced attempt to get a taste of the soup?

If it is the former, Japan's ADAS technology should be at the forefront of the world. If we flash back to five years ago, at least in terms of hardware, this conclusion will most likely be recognized.

In 1973, Toshiba creatively introduced semiconductors into automobiles and invented the first engine microcontroller unit (MCU), and other Japanese companies followed suit. In the automotive MCU circle, Japanese companies maintained their leading advantage until 2016, when the MCU king was Renesas Semiconductor, with a market share of 31%. If NXP is included, the two companies will occupy most of the market.

Although Renesas is still one of the "Five Elders", its status is obviously not as high as it was in the past. The market concentration has also dropped, and Toshiba, which was ranked in the top ten at that time, has fallen far behind. Sensors, IGBTs, and MCUs are all led by European and American companies, not to mention the newly emerging visual chips and computing platforms.

As a sign of Japan's loss of technological leadership, Japanese OEMs have purchased a large number of ADAS kits from Bosch and Continental. 30% of ADAS sensors and 25% of ECUs for Toyota, Honda and Nissan products come from these two Tier 1 companies. This ratio may not seem very high, but it is actually a humiliating ratio for Japanese companies that are used to keeping their own wealth within their own borders. Their exclusive and cross-shareholding supply chain is called "keiretsu". It is usually translated as "corporate alliance", but it obviously does not express the "uniqueness" of the Japanese.

For example, Toyota holds 24.7% and 22.2% of the shares of Denso and Aisin, respectively. If there are any good orders, they will first go to these two brothers. The latter two also reciprocate, and they kowtow to the elder brother when he says he is a temple.

There is no separate comparison method for software algorithms. However, if we follow the disengagement report released by California every year, Japanese companies are nowhere to be found, including Toyota TRI. Waymo and Cruise are in the lead. It is worth mentioning that Honda chose to cooperate with GM to develop ADAS systems, while Cruise is the only ADAS R&D company under GM.

Bundled vs Standalone

What has happened in the past few years? The Goose Formation model, which is bound by interests, is a double whammy. What if the big brother takes the wrong path? Japanese companies, led by Toyota, still hold on to the magic weapon that occupied the US market 40 years ago - fuel saving - as the criterion. Although they have seen the new four modernizations and shouted about embracing the future, the product advantages caused by fuel saving advantages make it too easy to make money and it is difficult to give up. If they make electric cars, wouldn’t it be a self-destruction? This has created a split between idealism and reality.

In fact, the same is true for Volkswagen. Volkswagen is a Chinese fan. Until 2019, the tough guy Diess turned to electrification at all costs, and he is still in the process of turning around today. He gambled his career for this.

However, even if OEMs in Europe and the United States go the wrong way, Tier 1 and 2 are still inclined to choose their own routes due to their high degree of independence. They do not confuse major customers with their biological fathers.

This led to the reorganization and rise of Infineon, STMicroelectronics, NXP, TI, ON Semiconductor, etc. When the OEMs established their electrification strategies, they were pleasantly surprised to find that the suppliers were already prepared in advance. Of course, the battery field is an exception, and East Asian suppliers are invincible all over the world. In this regard, the United States basically does not care, relying on holding shares in Samsung SKI and LG to bind each other. Europeans are willing to spend tens of billions of dollars to cultivate EU battery suppliers. The EU rarely reaches a consensus on spending big money.

Today, the Chinese are catching up in the fields of IGBT, sensors, MCU, and computing chips, relying on national policies and scale advantages. The only semiconductor hardware advantage Japan has over China is MCU. Of course, Japan still has a big advantage in semiconductor equipment and production processes (photoresist, hydrofluoric acid, wafer manufacturing), but they are separated from ADAS by several layers and cannot directly feed the application layer.

Of course, it is unfair to say that the Japanese have done nothing in the field of ADAS. In 2016, Denso also established the "ADAS Business Technology Development Department", but it only had more than 100 engineers, while Bosch's assisted driving department had more than 2,000 R&D personnel.

As an investor, SoftBank integrates the strength of Japanese companies at the final application level. Last year, SoftBank and Toyota took the lead in establishing the mobile service company MONET. Honda, Subaru, Mazda, Renesas and 276 other Japanese companies have invested in it, becoming a veritable "family fun" of Japanese companies. However, this is the backend of the application, and the front-end technology still needs to be solved by the companies.

Japanese companies are returning to their normal position

The deeper reason is that Japan no longer has the material and human resources to compete in the global industrial sector. Coincidentally, ADAS reflects a country's all-round industrial strength in IT technology, communication technology, semiconductors, fine chemicals, precision instruments, metal processing, etc.

Apart from other things, Japan had 250,000 college graduates in 2020, while China had 8.74 million college graduates in the same year. The differences in human resources, market and industrial sector size determine that Japan can only adopt a strategy of small-scale high-tech and large-scale follow-up. If Japanese companies choose to "go it alone", it may be manifested as "clicking the wrong technology tree". This is not about luck, but about destiny.

Sony has the world's leading image sensor technology, Panasonic has power battery technology, Mitsubishi Electric has world-class power electronic modules, Murata and Rohm have advanced automotive sensors and other electronic components. This requires Japanese companies to integrate into the global automotive industry chain, rather than just follow Toyota.

If they advance their own projects in isolation, they will not be able to provide OEMs with integrated ADAS solutions, so the OEMs will have to seek orders from Bosch, Continental, etc. This is exactly what has actually happened in the past few years.

The cooperation between Honda and GM also reflects this. To this day, Japanese companies still have reliable, low-cost automotive hardware manufacturing technology, which is still unique in the world. However, their backwardness in artificial intelligence, 5G communication technology, and perception algorithms prevents them from taking the lead in ADAS. They cannot compete in every track like China and the United States.

Now, do you still think that Honda's L3 product "landing" is the result of technological leadership? Japanese companies' early deployment is not only a stepping stone for big countries, but also a show of strength and a desire for favor. In the long run, the best fate for Japanese companies is to provide supporting services for the entire industrial chain of big countries, while the latter cannot exclude them because of their technology in some areas.