Auto Driving Investment Is Coming Out Of The Trough?

Recently, the "cool" field of automatic driving has been frequent. Investment in the global market is hot. The giants have increased capital injection, and domestic policies have been launched continuously.

First of all, look at the support of policy. In March 9th, the official website of the Ministry of industry and Commerce publicized the recommendation of the recommended national standard for automatic driving classification, which is to be implemented in January 1, 2021. This is China's first automatic driving grading standard, and the industry has been using the SAE classification standard developed by the United States.

It is understood that the "automatic driving classification" started work in 2017, which shows that China has been observing and studying the automatic driving industry and giving guidance from policies and promoting mass production.

At the same time, the industry of artificial intelligence belongs to the bonus industry. In March 4th, artificial intelligence became an important item in the seven parts of the new infrastructure.

Earlier, in February 24th, the 11 ministries and commissions of the state promulgated the "intelligent vehicle innovation and development strategy". It was proposed that by 2025, the intelligent automobile with conditional automatic driving would reach the scale production, and the intelligent vehicle with highly automatic driving would be marketed in a specific environment.

In addition to the continuous output of policies and standards, the capital level is also on the move.

In March 3rd, Google's Waymo was financed 2 billion 250 million dollars. The investors of this round of financing included Silver Lake capital, Canadian pension fund and Abu Dhabi's sovereign wealth fund Mubalada. In February 26th, the domestic automatic driving start-up company Xiao Ma Zhixing received $400 million from the Toyota Auto Body Co. On the same day, Yu technology announced that it had won BOSCH strategic investment in the B round of financing. Endowment.

It is said that capital is the smartest. In the past two years, auto driving and artificial intelligence have been cold together, and the financing fever in related fields has also decreased. Now we can see everyone's expectation of the potential of automatic driving.

At the moment, it will take time for the industry to really break out. Once, the giants including Google, Tesla and TOYOTA announced that they would achieve automatic driving in 2020. As a result, we all had to vote in advance. Why is there such a forecast gap?

This has to mention artificial intelligence, because autopilot grows with AI. In recent years, thanks to the technological breakthrough of deep learning, the wave of artificial intelligence has sprung up again. From the underlying chip, to the frame model, to the algorithm, the application, the hardware and software have been changed. Autopilot, as a master of artificial intelligence, has benefited from this.

But the development of technology tends to be wave like progress. For AI, this is also the case from the outbreak stage to the platform stage, to the low and low trough period, and then to the outbreak again. At present, deep learning is confronted with theoretical bottlenecks and artificial intelligence is very difficult in commercial process, so autopilot has entered the silent period.

Comprehensive capital action, automatic driving is more like the end of the trough, waiting for the next round of explosion. Why is it located at the end of the trough? The commercialization of automatic driving technology is closely related to the cost. Although the challenges of commercialization are numerous, the trend of cost reduction has emerged.

At the algorithmic level, such as Google AI framework and some algorithms have been open source for many years, there are many start-ups at the algorithm level that can optimize and innovate on this basis. This also means that a lot of low cost, skilled algorithm engineers have mastered the relevant technologies, which also allows salaries to enter a more fair price, resulting in a relative reduction in labor costs.

The chip level of computing power is not limited by powerful NVIDIA. Tesla used the intelligent brain of up to tens of thousands of dollars. Now Tesla has switched to its own chip without having to bear such a high cost of hardware. Recently, Tesla's chip hybrid led to a wave of word-of-mouth, but using self research or proprietary low-cost automatic driving chip is the general trend. Apart from Tesla, more and more giants and emerging companies are entering the field of autopilot.

In addition, on the road of wisdom, there are enterprises in the use of low-cost Internet of things equipment for high-density construction, data collection and monitoring.

When algorithms, computing power and external infrastructure enter low-cost areas, autopilot will create an exclusive ecosystem and architecture with a new automatic driving hardware and software system. Now, the field of automatic driving has entered a stage of highly division of labor. Some companies focus on algorithms, and some enterprises concentrate on computing power chips. In their respective fields, they have the opportunity to produce products equivalent to Android's operating system and high pass chips, and do not need to migrate from the PC end or the mobile phone side to the computing power algorithm. When software and hardware stand firm, automatic driving will also rise from the bottom of the curve.

? ? ? It needs to be pointed out that autopilot cars and traditional cars are two completely different species. For example, the traditional automobile industry is equivalent to the computer entering the recession bottleneck period, and the self driving vehicle is similar to the emergence of mobile phones, bracelets and other products. This is a new computing device. Instead of replacing traditional cars, it is a new growth point for automobile companies. 。 We can even imagine that there may be a driveway for autopilot cars in the future.

Now, we don't know where auto driving will go. In analogy, the evolution of iPhone has undergone many generations. There will be many iterations in the form and business mode of autopilot, and eventually to more advanced automation.

?