According to the relevant information released by Tesla's official website a few days ago, the price of its advanced driver assistance software Full Self-Driving (FSD) beta has been reduced from $15,000 to $12,000 in the US market, and the price in Canada has been reduced from $19,500 to $16,000, but the subscription price of FSD remains unchanged, still at $199 per month.
It is reported that Tesla's self-driving products are divided into 3 categories, namely AP, EAP and FSD. Among them, AP is the most basic version; In order to enhance assisted driving, EAP adds functions such as intelligent summoning, automatic parking, automatic assisted navigation driving, and automatic assisted lane change compared with AP. FSD, on the other hand, adds functions such as recognizing and reacting to traffic lights and stop signs, and automatically assisting driving on city streets. At a time when major car companies have increased their investment in automatic driving technology and made automatic driving a selling point for new cars, Tesla took the lead in reducing self-driving fees, do you want to set off an automatic driving price war after the whole vehicle?
Tesla FSD is not yet a "price anchor"
Similar to Tesla, domestic car companies have also begun to charge or stealth charges in the field of automatic driving.
NIO's assisted driving function adopts a subscription system, with a high-speed pilot assistance fee of 380 yuan/month, and a city pilot fee of 680 yuan/month. Xpeng uses the G6 as a reference, and the MAX version is 20,000 yuan higher than the PRO version (others include leather seats). The ideal AD MAX function is also standard on MAX models, which is 40,000 yuan higher than the previous stage (including one screen + two speakers in the rear). Zhiji LS7 can only be equipped with lidar in the Lux version, with an optional installation fee of 22,000 yuan (lidar) plus a functional package of 36,800 yuan, with a total price of 58,800 yuan. Huawei's smart driving fee is the same as the pricing of the M5 Smart Driving Edition, Avita 11, and Extreme Fox Alpha HI Edition, with a subscription fee of 720 yuan per month, 7,200 yuan per year, and a buyout of 18,000 yuan.
Zhuang Jingqian, associate partner of Roland Berger, analyzes that from the perspective of price: although Tesla FSD is priced at 64,000 yuan, because the current FSD function is not yet open in the Chinese market, few consumers directly buy Tesla FSD products at present, most consumers choose Tesla's EAP solution, the price is 32,000 yuan, which can achieve L2+ automatic driving. Compared with the L2 solution of domestic mainstream players, including Huawei ADS2.0, Xpeng XGNP, Ideal NOA, and NIO NOP+, this price is basically the same, and there are no significant price advantages and disadvantages.
From the perspective of performance: simply comparing the L2+ automatic driving solution, because Tesla adopts a pure visual solution, does not use LiDAR/advanced map as auxiliary input, and superimposes the complexity of China's road conditions, the performance of Tesla's EAP system is not as "silky" as the domestic brand solution in some specific scenarios. For example, due to the lack of high-definition maps, Tesla's EAP system needs manual guidance from the driver to enter the ramp in special scenarios such as three-way lanes. In the scene of unclear old and new lanes, Tesla's body will have a certain shaking phenomenon. Overall, Model 3's perception is powerful, its style is aggressive, its security is low, and it needs to be taken over manually. At the same time, Tesla relies on visual perception results, and its integration with navigation and maps is relatively low.
But for higher levels of autonomous driving, Tesla is completely based on vision + deep learning, strong software algorithms; China adopts the multi-sensor fusion scheme of "vision + radar", and the hardware is stacked. Tesla has a lower cost advantage and more powerful software capabilities supported by huge amounts of data collected around the world. "Of course, there are certain risks, Tesla every vehicle is collecting information in real time, in the complex international situation, regulatory policies may have certain restrictions on the real landing of Tesla FSD." Zhuang Jingqian said.
It is difficult to see a "price war" in the short term
If Tesla's self-driving charges at this stage are not much different from domestic car companies, and Tesla takes the lead in reducing prices in foreign markets, will this idea continue to China and will it set off a round of autopilot price wars?
Zhang Kangkang, a senior analyst at Tsingyan Huake New Energy Research Institute, believes that at present, the automatic driving services of domestic car companies do not take Tesla as a pricing anchor, because Tesla FSD has not yet landed in the Chinese market. But from the perspective of the future, Tesla has accumulated a lot of data, talents and capabilities in the field of autonomous driving, and if one day it reaches the expected "technical singularity", it may be difficult for other car companies to catch up. At that time, its FSD will not only be used for itself, but may also be sold to other car companies, becoming an important money-making tool, which should not be underestimated.
Zhuang Jingqian also believes that the probability of autonomous driving price war in the short term is not large, the core reason is that its industrial development stage is quite different from that of vehicle products. The core logic of the vehicle price war in the first half of 2023 lies in the convergence of technical solutions, the reduction of production costs superimposed by scale effects, coupled with the inventory clearance demand of China VIA, which has triggered a number of traditional car companies to actively or passively adopt price reduction promotion methods. In the field of electric vehicles, it is the "catfish effect" driven by Tesla, and other OEMs are forced to follow suit. The reason is that the process simplification brought by Tesla's integrated die-casting technology and the pure vision algorithm scheme have brought production cost advantages. At the same time, the cost reduction of the supply chain and the pressure of delivery volume after the epidemic directly triggered Tesla to adopt price reduction promotion.
However, autonomous driving technologies and solutions, especially L2++/L3 and above high-level autonomous driving solutions, are still in the stage of continuous evolution, far from reaching the maturity and convergence of technology and commercialization solutions, and it is expected that as autonomous driving technology continues to evolve from L2+/L3 or even L4, its price will continue to rise, with three core driving factors.
First of all, the number of core components is increased: for example, L2/L2+ autonomous vehicles are generally equipped with at least 5 cameras, while L2++/L3 and L4 autonomous vehicles need to be equipped with at least 9 cameras. In addition, L2++/L3 and above vehicles are generally equipped with relatively expensive lidar products. Second, the price of core components has increased. Taking the most important core component, the autonomous driving domain controller, as an example, as the requirements for computing power increase, the price will also increase. At present, the price of L2 controller is 800~1000 yuan, while L2+ is 4000~6000 yuan, and the cost of L3 and L4 is more than 10,000 yuan and 20,000 yuan respectively. Finally, the value of software algorithms increases. With the increase of the complexity of application scenarios, the increase in the amount of data required to be processed, and the value of autonomous driving software is also increasing, still taking the value of the upper software algorithm of intelligent driving domain control as an example, the value of L2 software is about 80~100 yuan, but this value will grow rapidly to L2+, L3, L4 to 1000, 3000, 6000 yuan. In addition, in order to further enhance the long-term competitive advantage, Xpeng and other car companies began to build their own computing power centers, and the initial investment of hundreds of millions of yuan made OEMs need to maintain a high premium level for their autonomous driving solutions to share the huge upfront investment.
Consumers' willingness to pay needs to be improved
Facing the future, "no car, no intelligence" is becoming the consensus of car companies. As Li Xiang, the founder of Ideal Auto, said: "It's like buying a building with twenty or thirty floors, it is the difference between whether there is an elevator or not." In the future, in the (automobile) high-end market, if urban NOA cannot be provided, consumers are the difference between buying and not buying. ”
Zhang Kangkang believes that the reason why there will be no automatic driving price war at this stage is that the ultimate purpose of car companies to make automatic driving is to sell cars. Software appears as added value, not as a completely separate commodity. The main body of the price war is the car, not the autopilot/assisted driving software. Another trend can also be observed, Xpeng Motors, which used to charge for autonomous driving services separately, is now abolished, and there are signs of reducing or even not charging for autonomous driving services. "Of course, car companies will not invest in vain, the wool is out of the sheep, and these expenses will eventually manifest themselves in other forms." This won't be a simple software price war. Zhang Kangkang said.
Zhuang Jingqian pointed out that autonomous driving technology is still in the process of continuous evolution, and the competition pattern of high-level autonomous driving and advanced assisted driving systems (ADAS) will show certain differentiation in the future. As technology continues to evolve, existing ADAS solutions such as L1/L2 will gradually form an industry best practice, that is, the best performance, the lowest cost, its cost will be reduced on the existing basis, but it will also become a standardized solution, so there will be no significant difference in price, so the competition will gradually turn to cost oriented.
Text: Hao Wenli Editor: Guo Chen Layout: Liu Xiaoye