The “irrational exuberance” of smart cockpits

What do you expect from the smart cockpit?

As the originator of automotive intelligence, Tesla seems to have been working on building its cockpit into a “game hall”, from the introduction of the earliest “ATV Racing” to “Happy Landlords” and “Happy Landlords”, which are popular among domestic consumers. Happy Mahjong”, to today’s “Cyberpunk 2077” and other 3A game masterpieces.

Xiaopeng Motors, which is the benchmark with Tesla, has no shortage of gimmicks in its cockpit. The P5 can open the sleeping space with the seat completely flat with one button. It can also watch movies, watch balls, play games, and sing K. Rounding up is equivalent to having a mobile phone. entertainment venues.

The development of smart cockpits in smart cars has been going on for several years. Looking back, is smart cockpits really “smart” enough? Or is this intelligence really what users want?

The smart cockpit is entrusted with “important tasks”

When it comes to autonomous driving, AI algorithms at the decision-making level are of course important. However, under the algorithm, as the most easily perceived intelligent experience for drivers and passengers, the interactive design of the intelligent cockpit is also an important reason for consumers to buy a car or not.

According to the newly released passenger car models in China in 2021, the penetration rate of smart cockpits has reached 50.6% in 2021, and the price range of the models with the highest penetration rate is 200,000-250,000 yuan, up to 58.4%. All vehicles with smart cockpits Among them, the penetration rate of the central control color screen is as high as 97.9%. At the same time, in the process of intelligent development of cockpit, the function of OTA (remote online upgrade technology) is becoming more and more popular, and the current penetration rate of OTA has reached 50.9%.

From the perspective of the market size of smart cockpits, according to previous forecasts by IHS, the global smart cockpit market space will exceed US$40 billion in 2021, and the market size will reach US$68.1 billion in 2030. According to ICVTank’s forecast, China’s smart cockpit market will reach 103 billion in 2025. According to the growth rate, China may become the world’s largest smart cockpit market in the future.

At present, mainstream car companies are all focusing on smart cockpits. In fact, smart cockpits have become a “strife” in the electric vehicle market. Looking at the recently unveiled new cars, the Lili L9 focuses on five-screen linkage, and the Xiaopeng G9 focuses on the Internet of Vehicles. In addition to the new car-making forces, Internet companies have also deployed smart cockpits, such as the Jidu built by Baidu and Geely, the Extreme Fox Alpha from Huawei and BAIC, and the Sky ME5 from JD.com and Tianji Automobile.

Why are OEMs and Internet companies betting on smart cockpits?

This question seems like a good answer, but “Intelligent Relativity” believes that this question needs to be divided into two parts. On the one hand, the smart cockpit is an indispensable part of the autonomous driving technology, and the “intelligence” in the autonomous driving technology also needs to be reflected in the interior space of the car.

On the other hand, and more importantly, the current development of autonomous driving technology has slowed down, or there has been no major technological breakthroughs, resulting in higher vehicle prices and stronger competition with peers. Competitiveness, OEMs use smart cockpits as a “selling point”.

According to the forecast of Roland Berger International Management Consulting, by 2030, the penetration rate of L4 and L5 high-level autonomous driving will reach 5%, that of L2 and L3 autonomous driving will reach 60%, and the primary level of autonomous driving assistance will reach 35%. %about. From the actual results, this forecast is obviously too optimistic.

Relevant statistics show that the penetration rate of passenger car assisted driving systems in 2021 will be about 37%, of which L1 accounts for 21% and L2 accounts for 18%. At the stage when the penetration rate of the L2 level has not reached 50%, it is still too early to talk about the large-scale landing of the L3 level and even the L5 level.

In fact, the current autonomous driving technology is stuck at the L2 level. In this regard, Tan Xuguang, chairman and CEO of Weichai Power, said directly to the public, “Assistant driving is to hope that a specific environment can be realized. For example, in a port, it is very expensive to replace the driverless driver. Don’t listen to them. L1 is okay, L2 is so-so, and L3 is impossible.”

The implementation of high-level assisted driving (autonomous driving) is very complicated. Many OEMs and technology companies are technically easy to grasp and implement. configuration.

To sum up, due to various reasons, the actual technical application of autonomous driving cannot be promoted to the high-end for the time being. In order to introduce new products from the old and create “selling points”, consumers are most likely to perceive and care about the interior of the car. Interaction design becomes a good entry point.

The research of Yiou think tank shows that among the 7 major reference factors affecting car purchase, intelligent experience has been ranked fourth, second only to quality, performance, and design factors. As consumer demands have been upgraded from physiological needs such as safety and comfort to psychological needs such as emotion and belonging, the importance of smart cockpits as an extension of the social relationship between drivers and passengers will also increase in the future.

However, can the current smart cockpit really shoulder this “heavy task”?

Is it difficult to “stack” the future of smart cockpits?

After experiencing the motion recognition of a car factory’s smart cockpit, a car reviewer complained that it was a function that he felt made himself look stupid: “I want to nod to activate a certain switch, but the setting it recognizes is’ I’m going to put my head in a fixed place’, and I can’t lift my head up when I nod it down… Who can drive a car and put his face on the big screen and nod his head and still not ‘rebound’?”

According to the previous statistics of “Automobile”, there are at least 40 technologies related to the smart cockpit, and “action recognition” belongs to the human-computer interaction part of the smart cockpit. Among the many technologies of the smart cockpit, “safety assistance” and “human-computer interaction” are the perception technology segment closest to the “intelligent experience”, and are also an important battlefield for many Tier 1, artificial intelligence and software companies to fight.

The current smart cockpit market is a “three-way wrestling” battlefield. In addition to the self-research by OEMs, most of the smart cockpits on the market are provided by Tier 1 suppliers such as Qualcomm, Bosch, Continental, Visteon, Texas Instruments, etc., and technology companies such as Huawei, Tencent, and Baidu have also come to an end.

The AITO M5, a smart car jointly developed by Huawei and Xiaokang Celis, is equipped with a HarmonyOS-based smart cockpit. It is understood that the HarmonyOS smart cockpit currently covers entertainment, games, parent-child and other vehicle scenarios, and has more than 100 applications.

In November last year, Tencent released the newly upgraded TAI4.0 smart cockpit solution. In order to improve the driving experience, Tencent has also designed a new generation of maps for people and vehicles – “Tencent Smart Driving Map”, which realizes “one map” of the same source of data for high-precision maps, ADAS maps, and standard maps. In addition, its TAI4.0 smart cockpit solution has upgraded the in-vehicle version of WeChat, adding basic applications such as car-machine navigation and location sharing, and other functions such as national karaoke, Tencent conference, and 3D city exploration are also ready.

In addition, in the upcoming official reveal of the ideal L9, it is said that its smart cockpit is a 5-screen interactive mode, with an oversized HUD as standard, and a safe driving interactive screen above the steering wheel. In addition, the vehicle center control screen, passenger entertainment screen, rear cabin The display uses three 15.7-inch car-grade OLED screens.

If you take a closer look, whether it is the display screen in the car, or the installation of WeChat and National K-Song software, can you really find out what are the substantial differences between the smart cockpit solutions of different car companies or suppliers? It can be said that although different OEMs are claiming the “innovation” of their smart cockpits, strictly speaking, this “innovation” is just another sense of “innovation”.

In fact, large screens + multi-screens in smart cockpits are not uncommon, and functional configurations such as smart voice, HUD, OTA, and DMS are also very common. Behind the convergence of smart cockpit functions, 90% of the solutions provided by most suppliers to different brands are the same, which is obviously difficult to meet the needs of future smart cockpit development trends.

Why is the current homogenization problem of smart cockpits more serious? Or why is it difficult for smart cockpits to be truly innovative?

On the one hand, it originates from the design ideas of current car companies. The logic that the current OEMs adhere to for smart cockpits is “no one has me”, resulting in a large number of functions and designs.

An auto industry analyst once said that after “Wei Xiaoli” came out, to a certain extent, this is a very cruel mentality that the traditional auto industry bears – to find a new car that is far better than the other party in the car designed by yourself. The bright spot, then I must have someone else’s. For example, a car manufacturer priced its car at less than 60,000 yuan, but also wanted to have all the sensing functions as standard.

The mentality of “wanting everything” leads to the result that everything has to be installed. From the perspective of the entire industry, this “involution” idea has also caused the convergence of different car companies and brands in the smart cockpit. This is the design idea level of the smart cockpit. From a technical point of view, the relatively low implementation threshold of the smart cockpit is also one of the reasons.

Compared with automatic driving, the implementation of intelligent cockpit functions does not involve chassis control, has low safety pressure, low technical implementation difficulty, and easy perception of results, which will help to rapidly improve product differentiation competitiveness. Therefore, OEMs are waiting for the key technologies of autonomous driving to mature. Stalls, began to gradually shift energy to the landing of the smart cockpit.

For car companies, to solve the current homogenization problem of smart cockpits, it is necessary to design smart cockpits that meet their own tonality according to the characteristics of the brand, rather than just working on stacking materials.

The design of smart cockpits by car companies is also similar to “training dogs” to some extent. It is well known that Bian Mu is the smartest dog, but some dog training experts said that Bian Mu is the smartest only because the project to evaluate the IQ of dogs is most in line with Bian Mu’s characteristics. In fact, each kind of dog has its own unique attributes. They are the smartest in their respective fields. This may be the development idea of ​​the future smart cockpit. Only the most suitable, not the smartest, and the clearest brand tone and user needs can create the smartest smart cockpit.