[Autohome Industry] After intelligent interconnection became the "new darling" of the car industry, more and more new cars on the market began to have a unique change, that is, the physical buttons in the car There are fewer, but the screens in cars are getting bigger and more numerous. Cars are no longer simply means of transportation, but rather "little partners" in travel. Through the car screen, you can order takeaways, locate restaurants, and make contactless payments, etc. If the early competition between car companies was mainly concentrated in the 1.0 era in areas such as engines and chassis, now it is the 2.0 era in which the battle for intelligent systems behind the "screen" is underway. Wang Jian, Chairman of the Technical Committee of Alibaba Group, once said: "It is a shame for car people to use a mobile phone in a car." So, what kind of car-machine experience do we need? Can the current large car screens satisfy people's imagination of smart cars? Why will in-vehicle OS usher in competition from technology companies represented by BAT? Who can ride the wave in the treasure land of car OS?
■Real demand or false outlet? Judging from the car consoles (mainly referring to infotainment systems) seen on the market, they can be divided into two categories: one is that the car connects to the mobile phone, and the mobile phone screen is projected onto the car console, using Apple CarPlay, Google's Android Auto and Baidu Carlife is represented; the other type is car-level systems, which inject application ecosystems suitable for car scenarios by improving the computing power, performance and functional experience of the car, such as Ali AliOS, Huawei Hongmeng OS, etc. Which category is better? At present, users prefer mobile phone mapping solutions represented by CarPlay. However, in the long run, this method seems simple and convenient, but it does not access the underlying architecture of the vehicle, whether in terms of computing power or derived functions. All have great shortcomings. Even Huawei's latest HUAWEI HiCar is considered by the industry to be a transitional solution for mobile phone mapping. One of the real purposes of Huawei's launch of HiCar is due to Huawei's high mobile phone market share. Using Huawei mobile phones to "get on the car" is to take advantage of the situation. The development of cars and machines will be able to feed back Huawei's mobile phone business ecosystem. It is reported that the new Baojun RC-6 will use Huawei’s mobile phone mapping solution. So there is a second type of car machine - based on its own car OS (Operating? System, car operating system), and based on this set of car-level operating systems, more efficient applications such as Zebra AR navigation are created. At present, although there are endless in-vehicle systems "mixed in the market", many car companies only attract attention with their vehicle infotainment system solutions. For example, in order to make their systems "featured", many car companies have added various ecosystems to their cars, but they have not provided an upgrade mechanism for these third-party applications, resulting in a gap between the application experience of cars and mobile phones. Getting bigger, especially navigation. In fact, many in-car needs are mostly created out of hand. Compared with trivial needs such as "booking tickets", "shopping on the car", and "telling jokes", users need more software services such as improved driving and vehicle-related software. . What can be confirmed is that a good car computer needs to have functional safety and data security. Simply "stacking" entertainment functions is not the future of the car system. ?
■ Why do technology giants and car companies gather together for in-vehicle OS? In the future, a real car-machine will need to combine voice, gestures, perception, and even driver expression interaction forms to cater for the multi-modal interaction of smart cockpits. Therefore, a mobile phone holder or a projection solution is far from satisfying people's imagination of smart cars. The most important thing is that considering that the future automotive ecosystem (especially connected cars) will “nugget” the long-tail value of users and the subsequent revenue generated from data, it is the car companies that continue to actively develop in-vehicle operating systems. Explore the reasons. It seems that whoever occupies this screen controls the future of the Internet of Vehicles. Limited by policies, regulations, technology, cost and other reasons, domestic car companies are deploying in-vehicle OS much faster than autonomous driving. However, in the mainstream view, the development path of the in-vehicle OS field is not clearer than that of autonomous driving. Each manufacturer's own car and engine are different, and even the car and engine of different models of the same brand will be different. Without standards, it is difficult to interconnect. This requires companies to establish their own standards. Whoever masters the standards will take the initiative. ? In fact, the car OS launched by technology giants are trying to establish their own standards and compete for the very important screen of the car. Recently, the competition has been heating up dramatically.
『Tencent AutoLink TAI3.0 first rides on Haval F7, with in-car WeChat』 In June 2020, Tencent Mobility’s newly upgraded TAI3.0 ecological Internet of Vehicles tested real driving scenarios on the new Haval F7 Function display; In May, the strategic reorganization of Zebra Network and AliOS was officially completed. After opening, it will cooperate with SAIC Roewe, Ford, Citroen, Baojun, Qoros and Skoda. Roewe RX5? PLUS will be the first model equipped with the new Zebra Smart Travel VENUS system. ; On the other side of the battle, Ford's new sixth-generation Explorer equipped with Baidu's in-vehicle OS system was launched on June 16; Chery New Energy's official Weibo mentioned in February that the new car will be equipped with Huawei's Hongmeng car system... This This competition has become a team battle in which leading companies lead their respective “side cards”. Behind this team battle, in addition to the car and machine being the next entrance to mobile Internet, what companies value more is the service ecosystem development of in-vehicle OS, which can meet the respective strategic layouts of car companies. For example, AI In Car is Tencent’s pawn in the field of automotive OS. "Tencent Auto Connect does not manufacture hardware or develop operating systems. In cooperation with car companies, it always plays a supporting role. It connects Tencent's ecosystem with the traditional automobile industry through AI in Car." Tencent Auto Connect Zhong Xuedan, general manager of the business unit, said this. The Xiaodu OS created by Baidu with massive data is also aimed at the application level, and there is still a lot of room for improvement in the control of the underlying system of the car.
"Ford's new sixth-generation Explorer equipped with Baidu in-vehicle OS system"
■Who can ride the wind and waves? The next era will be revealed. It can be said that the current state of competition is a chaotic war, with old and new forces competing for territory. In fact, strictly speaking from the underlying architecture, QNX and Linux are the originators of automotive OS. Google's Android is deeply developed based on the Linux kernel. These three are also the three major camps in the current market. Among the new teams, Alibaba AliOS and Huawei Hongmeng OS are developed based on the Linux kernel and are deeply customized (including system kernel modifications) on top of the underlying architecture system. At the car company level, in addition to Volkswagen and Tesla's in-depth self-developed operating systems, car companies such as Audi, NIO, Xpeng, BYD and other car companies generally choose to conduct customized development based on the open Android system, which does not involve changes to the system kernel and is semi-autonomous. R&D. Mainly because it is not easy to develop a complete underlying OS on your own, and you need to overcome countless pitfalls. The Volkswagen ID.4 has experienced release delays again and again due to software overturns.
『Tesla Version developed based on Linux』 "Several automobile companies with Internet background particularly want to master core technologies. This is closely related to the Internet gene. The most important products and core of all Internet companies Technology must be mastered independently," said Zhuang Li, former vice president of NIO Software Development Department. In her view, it is natural to hope that vehicle digital experience technology will be firmly in the hands of car companies to ensure their core competitiveness and differentiated characteristics. Customized development can allow OEMs to greatly shorten the development process and reduce development costs, while also gaining control over the operating system. But even if there are countless pitfalls, Volkswagen still wants to "harden" Tesla. "Because the software stack is fully transparent, it is easier to optimize and iterate the system, and it is easier to gain control over product technology. However, the operating systems of traditional car manufacturers are more based on black box software from other suppliers and are more restricted. It is difficult to modify and optimize independently," said Li Xingyu, Vice President of Horizon Strategic Planning. It can be seen that due to its open source nature and strong customization capabilities, Android has basically become the first choice of major car companies and OS suppliers in recent years. However, the low operating stability of the Andriod system itself will be brought into deeply customized car OSs. , some models that have been mass-produced and delivered have black screens and crashes. Secondly, when the Andriod kernel is optimized and upgraded to a new version, developers will inevitably need to consider re-adapting it to their own systems. Although vehicles can be upgraded through OTA, the labor costs caused by software adaptation development and testing are still prohibitive. Don't underestimate it. Once serious vulnerabilities are not discovered during the adaptation process, the losses will be huge.
Domestic OEMs are also wary of this, which brings expansion opportunities to other operating systems. For example, AliOS has been applied to at least 9 car brands, and more than 700,000 vehicles are equipped with the Zebra Smart Mobility system; and the independent Huawei Hongmeng The importance of OS to the entire industry is even more self-evident. According to Huawei Consumer Business CEO Yu Chengdong’s description, the Hongmeng operating system underlying Hongmeng cars has cross-platform capabilities, including supporting all scenarios and across multiple devices and platforms. Opening up these ecosystems means that more and more intelligent terminals can be controlled by Hongmeng vehicles, and car owners can enjoy more ecological services. Does that mean Hongmeng OS can replace Andriod OS? At present, the possibility is very small. As a latecomer, Huawei, even if it develops its own operating system, it is too difficult to break the network effect of Andriod and form its own ecosystem. This is also what Ali?OS faces* **Same question. Ren Zhengfei, founder of Huawei, also made it clear in the interview that "the Hongmeng OS launched by Huawei will not replace Andriod OS. If Hongmeng can have part of the ecosystem, it will be a complement, not a substitute." In fact, Huawei's real strategy is not to surpass Android and other camps, but to rush all the way to the next era. Huawei's 5G adopts this strategy. It is precisely because Huawei began to lay out 5G before 3G was popularized. Only ten years later this year, Huawei's 5G is far ahead. The same is true for Hongmeng OS. Its power is that it can integrate the large smart terminal of the car into the entire Internet of Everything network layout, and realize the perfect migration of data and experience to multiple terminals. At this level, the car side is very important and is an indispensable and important puzzle piece. However, Huawei's competitors at the automotive OS system level have not shown similarly huge plans. Who will be the winner of the operating system in the era of Internet of Everything? Maybe it’s Hongmeng, maybe it’s another strong rival, Google Fuschia.
■Written at the end At this stage, most vehicles can only meet users’ entertainment needs and cannot provide users with deeper services. In addition to the 4G network's inability to meet in-vehicle needs in terms of transmission speed and latency, the performance of the in-vehicle OS has not reached the same level as current mobile phone and computer operating systems, whether it is the system itself or external computing algorithms. All still need to be developed. Whether it is Tesla in the car industry or Huawei in the Internet, they have made a big move into the automotive field, combining their IT genes with the genes inherent in cars to create new species that are difficult for ordinary companies to copy. If Chinese brand car companies want to break the situation, they still have to start from their own fundamentals and find a suitable development path. In the future, as in-vehicle OS experiences become increasingly richer, vehicle safety will become the primary issue facing software suppliers and vehicle manufacturers. How to create excellent stability and reliability of in-vehicle OS, whether it is more advanced technology or more rigorous testing systems, requires more in-depth cooperation and efforts between enterprises. (Text/Autohome?Peng Fei)