Huawei also learned the “One More Thing” formula when it launched its much-anticipated Mate 40 phone on Oct 30. After the mobile phone conference, HUAWEI Intelligent Automotive Solution (HI, HUAWEI Intelligent Automotive Solution) was also announced.
According to Huawei, HI smart car solutions include: a new computing and communications architecture and five smart systems; Namely, intelligent driving, intelligent cockpit, intelligent electric power, intelligent network connection and intelligent vehicle cloud, as well as lidar, AR-HUD and other full set of intelligent components.
To be honest, if I hadn’t confirmed again and again that this was Huawei’s launch event, I would have thought it was just another promotion by some new company to “draw doughnut money”. After all, these “bright spots” above are all the hot fields in the auto industry in the past two years without exception. Almost every auto release conference and auto show will have manufacturers mentioning the implementation plan of the above functions. However, few of them can really be better realized.
This aroused the curiosity of the author, so I carefully studied this HI solution of Huawei.
What is HI trying to achieve?
Relying on its own technical and capital strength, Huawei wants to provide auto OEM factories with a complete set of intelligent software and hardware solutions.
Based on Huawei’s chip experience, HI can provide powerful on-board computing power and build three computing platforms together with software: intelligent driving computing platform, intelligent cockpit computing platform and intelligent vehicle control computing platform. It is equipped with AOS (smart driving operating system), HOS (Smart cockpit operating system) and VOS (Smart car control operating system) respectively.
(MDC’s intelligent driving computing platform, which is based on Huawei’s Rising series OF AI chips, is no weaker than Mobileye and Nvidia in terms of computing power;
As obscure as it may sound, huawei wants to create a complete supply of hardware and software for autonomous driving, in-car entertainment and vehicle driving.
First for the autopilot system, huawei said HI to building China’s best road environment autopilot system, design thought based on “to the end of the beginning”, the level of L4 autopilot architecture as the foundation, to provide full level of L4 — L2 + autopilot stack solution, with all kinds of sensors (including the research of laser radar), the advantages of the central processing unit (CPU), algorithm, and the accumulation of huawei in the field of AI, the system can learn by machine itself, continuously optimized.
Second for electric system, will be based on huawei said HI PDDP (electric digital simulation development platform more room field coupling AI optimization algorithm) to create the industry’s leading intelligent oil cooling cooling technology, the core power components with immersion oil cold heat, heat dissipation effect signific
antly increased when the high speed, thus providing more power output, make cars more than three seconds to enter 000 level; At the same time, Huawei will also offer a full-line 800V high-voltage fast charging solution, which can increase the battery life by 200 kilometers after 10 minutes of charging.
Then there’s the smart cockpit, which Huawei envisions more like a scenario for fully autonomous driving. For example, AR-HUD can turn the ordinary windshield into a large hd screen. Combined with the 7.1 surround sound system, users can watch movies, play games and hold video conferences. And with visual recognition and speech ability, the interaction can be carried out through speech, gesture and expression.
Finally, Huawei said the HI drive will have two systems linked to achieve redundant configuration to ensure no loss of power at any time to ensure driving safety. In addition, HI will ensure the security of users’ personal information and privacy through AI and big data analysis to give early warning of battery abnormalities and huawei’s long-term information security accumulation.
Why does Huawei do HI?
At first glance, what Huawei HI has promised is quite good. At least, Huawei has strong technical strength in chips, electronics and software, and abundant capital as support, so it always has credibility.
The problem is that even with huawei’s huge capital investment and research and development team of around 3,000 people, they are trying to do too much. Each classification goal is achieved enough to compete with today’s traditional Tier1 tier vendors, which is no longer as “simple” as developing a “Hongmeng” system.
Such as automatic driving, although huawei has released last year applied to automated driving a few MDC data processing platform, but it is not just count than visible force parameters, the different route of image recognition, different technical solutions focus will lead to a big gap, such as nvidia chip series focused on recognition, computer vision and tesla FSD chip is focused on machine learning. In addition, hardware is only one aspect, and the development and optimization of algorithm software is the most important priority. In addition to the training of simulation platform, massive amount of actual road driving data is also needed to truly solve the problem of “Corner Case” and realize l4-level driving.
Another example is the car control system. Although it seems to be only about electronics and software, in fact, it requires a lot of traditional car tuning experience to make the driving experience better. And this is not what Huawei’s current technical staff and experience can do.
Therefore, no matter from which perspective, Huawei’s HI scheme is faced with many challenges. In particular, Huawei needs a lot of “help” from OEM factories in the early stage to realize the initial implementation of functions. Of course, Huawei is also very clear: “Huawei does not build cars, but focuses on ICT technology and provides incremental parts for intelligent connected cars to help car companies build good cars.”
As a matter of fact, the long-term goal behind Huawei is clear.
First of all, it is necessary to understand the operation logic of the capital market. In addition to the relatively mature fields, capital likes some high-tech fields very much, and the return on capital of these fields will be amazing once the technology can be realized. In recent years, there have been a lot of technological development caused by the wind, such as AI, 5G, automatic driving, etc. If you put these technologies together, you will find that the field of new energy vehicles includes all of the above.
Huawei is trying to grab the cake.
Due to the market changes brought by new energy vehicles, all the automobile companies suddenly have some “unadaptability”. The fundamental reason is that the original production model has been mature and stable for many years, and most of the functions controlled by suppliers are difficult to catch up with the ever-changing pace of Tesla in a short time. Even current Tier1 suppliers, such as Bosch and Fario, have set up separate software development teams to “tap into” the whole package of intelligent functions under the new electronic and electrical architecture. However, there are still many “obsolete” problems such as organizational structure, RESEARCH and development mode and talent type.
Although it is unclear exactly what the future of cars will look like, the general direction must be to become more “intelligent”. If we want to achieve such as automatic driving, FOTA upgrade and a series of intelligent functions, we need to fundamentally develop the underlying electronic and electrical architecture of the vehicle with “function-oriented”, and at the same time, we can build a good development foundation by combining hardware and software. Huawei, on the other hand, has just such strength.
Huawei not only has the accumulation of AI, 5G and other technologies, but also can integrate them with its huge software development resources, both in terms of its understanding of electronic and electrical architecture and its strength in software development. So that’s why Huawei started to enter into the “Internet of Vehicles Lab” in 6 years, and then went deeper into the automotive field. Last year, huawei launched the autonomous driving computing platform and just released the HI full stack solution.
In other words, Huawei takes a fancy to this huge cake of the future. In the early stage, it cooperated deeply with automobile OEM manufacturers for initial function implementation, growth and development. Then, I gradually established my own core barriers, had a voice in the automotive field, and gradually changed my identity to become a new Tier1 supplier in the automotive industry.
The advantage of doing so is that many auto companies do not have the ability and strength to catch up with the “new energy era” of automobiles, or the will to do so; However, Huawei provides a set of relatively mature solutions, as long as a few are not unreasonable, for OEM factories, far cheaper than their own research and development. Not only that, even though auto companies have their own research and development capabilities, Huawei’s ecosystem already has a considerable influence, and auto companies joining huawei’s “camp” may also bring a lot of competitiveness to their products.
In addition, although people still have various doubts about autonomous driving technology, once it is implemented, it will not only have huge market potential in long-distance transportation and public transportation, but also the usage scenario of taking cars will become a new traffic entry point, which may have much more commercial value than mobile phones. Huawei not only wants to take the cost of the solution away from the OEM factories, but also wants to get more value from the C-end users.
So it’s not surprising that Huawei doesn’t build a car, because it doesn’t need to build a car at all, and can make more money than “build a car”.