Translation

NIO, WM Motor, XPeng, and Ideal… The first threshold for competition among new car manufacturers is high topic relevance and cannot fall behind in popularity. From this perspective, the pressure on Hozon is not small.

Why do we pay attention to Hozon? The first step for new car manufacturers to produce cars is to be “legal”. Today, it seems that various companies have obtained production qualifications through various means such as cooperation and outsourcing, and have obtained legal identities for car manufacturing.

On May 17, 2018, the Ministry of Industry and Information Technology announced the 308th batch of “Road Motor Vehicle Production Enterprises and Product Announcements”, in which Zhejiang Hozon New Energy was listed. In addition to the 13th new energy vehicle production license issued by the National Development and Reform Commission in April 2017, Hozon has become one of the few new car manufacturers to have obtained both application approvals from the National Development and Reform Commission and the Ministry of Industry and Information Technology for production qualifications.

Well, paying attention to Hozon is because Hozon is serious.

However, competition in the automotive industry is extremely cut-throat, and just being “serious” is far from enough. In the current era of intelligent electric vehicles, can Hozon secure its own place?

Will software “Parts” appear independently in the BOM (bill of materials) table?

How to judge the sincerity of a new car company? How to judge if Hozon is really “serious”? Intelligent driving is a very good entry point.

Because the core of intelligent cars is the ability to grow and be lifelike, supporting iterative updates, as demonstrated very clearly on Tesla’s Autopilot. However, intelligent driving is a comprehensive technology that spans many technical disciplines and involves the upstream and downstream industry chains, making the research and development threshold particularly high.

If a supplier solution is adopted, Hozon will not have the know-how for the system, and the subsequent iterative updates will become an empty talk. How does Hozon view intelligent driving?

Let’s first list the intelligent driving system configuration of Hozon U:

8 cameras (front dual-lens + 4 panoramic + 2 side perception)
1 millimeter-wave radar (front)
12 ultrasonic sensors
Mobileye EyeQ4 visual perception chip
Intel 14-core X86 central computing chip

From the architecture of the sensors and chips, Hozon U is fully equipped with first-tier configurations. A pile of hardware does not necessarily mean excellent real-life experience. But for many cars, if they cannot even achieve “hardware programming”, they can forget about the next step.# Software as a Part in BOM

As a startup in the automobile industry, Hezhong, faces great challenges in matching hardware solutions with large-scale software development, considering the difficulties in funding and talent reserve compared with traditional automakers. However, Hezhong Director Wu Junjie’s PPT at the Intelligent Driving Research Institute of Hezhong contains a sentence that states:

Software is also a “part”, and software “parts” will appear independently in the BOM table.

The fact that software has become an independent part of the BOM indicates that the importance of software to Hezhong’s intelligent vehicles is indeed significant. But, we all know that in this field, Tesla has the most thorough software system, which is actually self-developed, and Volkswagen, which is transforming recently, announced that it will acquire its previous software supplier and independently establish software capability.

The term “Hezhong” seems to have a certain difference from Tesla and Volkswagen. I raised this issue to Wu Junjie, and he explained why it is more reasonable for Hezhong to first establish a software capability management mechanism for current Hezhong.

In fact, we, Tesla, and Volkswagen are doing the same thing, which is to support self-developed hardware and software in the system. How to manage the products with independent intellectual property rights we have self-developed? We must manage them according to the existing screw and nut system now. We must make the invisible software into parts and put it in my supply chain system. At the time of actual production, there will be no basket catching software to throw it in the production line. It needs a set of IT mechanisms to ensure the company as a whole.

For example, there is a function of offline configuration on the current car. The software of each controller will only be flashed onto the vehicle bus when the car is finally assembled. Because my car only knows what configuration it is in the last step, and the software package of the lowest and highest configurations are different. When I flash it, I can send it in according to different configurations.

The software supply chain system can support self-developed of all products in the future with more significant volume. At present, our core product insists on self-development, and we still need to rely more on partners for standardized things.

When the R&D force is insufficient to support complete vertically integrated self-development, Hezhong proposes selecting the core software capability to be controlled in its hands and handing over the underlying standardized capabilities to suppliers, which is the strategy proposed by Hezhong.

In addition, there is another sentence in that first PPT slide, “Become China’s Audi – Benchmarking Audi zFAS controller”.

According to Wu Junjie, Hezhong has invested a lot of human and material resources to develop the intelligent driving domain controller.

In the next generation computing platform, Hezhong will adopt NVIDIA computing chips, and in the research and development iteration of software algorithms, Hezhong will jointly develop AI vision algorithms with its partners.Wu Wenjun focused on the development of domain controllers for intelligent driving:

The domain controller is a very complex hardware. Tesla Model 3, for example, only has three domain controllers, while Future Mobility Corporation will be divided into four domain controllers. This means that dozens of electronic control units in the entire vehicle will be replaced by one or four electronic control units. The software and hardware inside will not decrease but will only increase. Therefore, the first step is to separate software and hardware, split the software and hardware supply chain, and let professionals do professional work.

The second step is how to break the bottleneck of intelligent driving domain controllers. It is difficult for a new car company like Future Mobility Corporation to promote the strong control chain. What we need to do is to move market component suppliers from the “turnkey” mode to the “component supply” mode. Companies like TTTech provide us with a complete set of systems as soon as we collaborate with them, but our OEMs need to have their own vehicle system integration and forward development capabilities and decompose the system into individual components. Once the OEM has such integration and development capabilities, cameras can be handed over to one supplier, radar can be handed over to another supplier, and controllers can be handed over to another supplier. We will break the monopoly of international component companies on us.

The reason why Audi’s zFAS can become a benchmark in the industry is not only because of its supply chain support, but also because Audi has strong vehicle and forward development capabilities. We need to push Future Mobility Corporation’s vehicle system integration and forward development capabilities to the world’s mainstream levels. This is our goal.

It seems that Wu Junjie’s team is challenging some technologies that touch on the essence of the industry’s new trends, which makes us full of expectations for Future Mobility Corporation’s intelligent assisted driving system.

The Challenge of Transparent A-pillars

In the intelligent cockpit part, Future Mobility Corporation has a differentiated highlight: transparent A-pillars.

From a safety perspective, considering the potential collision probability, A, B, C, and D pillars cannot be omitted. However, due to the physical properties, it is inevitable to bring visual blind spots to the driver during driving.

The pain point of the A-pillar blind spot is universal. For example, when turning, the driver’s view may be limited, which may cause a collision with a pedestrian who suddenly appears. In fact, long before Future Mobility Corporation, Toyota and Jaguar Land Rover had released concept cars of similar shapes to try to solve the problem of A-pillar blind spots.

The reason why it has not been mass-produced so far is that the structure strength of the transparent A-pillar and the car-grade verification of the screen have huge challenges.

The director of Hezhong Intelligent Cockpit Research Institute, Zhang Shi, introduced that the two cameras and a screen for the transparent A-pillar function hardware are arranged on the premise of not affecting the structural load-bearing strength of the vehicle. As shown in the figure, there is still a strong A-pillar below the screen for support, which visually resembles the projection of two Samsung curved screen phones.

As for the screen material, it is a 12-inch OLED curved screen with a resolution of 1080 × 4320 and a contrast ratio of 10000:1, far exceeding the industry average. Hezhong independently developed an algorithm called “forehead tracking”, which allows the screen angle to change as the driver’s head moves forward, backward, left or right, to avoid distortion caused by a fixed depth of field.

Whether it is intelligent driving or intelligent cockpit, Hezhong seems to be challenging some underlying technologies or unresolved industry issues while doing the basic work well, and such an attitude is worth appreciating. Of course, solving these problems is not easy, and there are still many challenges ahead for Hezhong.

Zooming out the perspective, the law of survival becomes simple. As a new car manufacturer, it is difficult for you to compete head-on with traditional automobile giants in terms of scale, funding, and talent pool. If you cannot win by excellence in cutting-edge technology and product innovation, the pressure from the industry can easily deform your actions.

This is the challenge of Hezhong, as well as the challenge of all new car manufacturers.