Shifting from Tencent to Baidu Map, What’s the significance of Tesla in China?


Recently, Tesla‘s stock price has become a topic of concern in China and abroad.

After the company’s market value exceeded US $100 billion, the stock price rose 19.89% again on February 3, and since then, Tesla’s market value has gone up to US $140 billion. Tesla, which has become the second-largest automaker in the world by market value, has already pulled down luxury brands such as Mercedes-Benz and BMW. GM and Ford in the United States are far behind. The question is in 2020 can Tesla challenge the current position of Toyota in terms of market value, or become a landmark in the competition between automaker new force and traditional car makers.

In China, the news of Tesla changing its map service provider has also become the focus of public discussion. Tesla previously installed Tencent maps, it ended the partnership with Tencent since on January 17 this year, Tesla announced that its navigation map would be changed from the original Tencent map to the Baidu map.

Some analysts believe the reason why Tesla uses Baidu navigation is that Baidu will gradually be beneficial for automatic driving except Tencent maps has given Tesla users troubles in navigation. At the same time, various signs indicate that high-level technology for automatic driving is also rapidly commercializing with the rapid commercialization of AI and 5G, many Tesla owners, after using assisted driving and automatic driving, also said that the original automated driving is so close to us.

As a leader in the automated driving industry, Tesla naturally does not want to give up this once-in-a-lifetime opportunity.

Just on the evening of January 16, Tesla’s official Weibo announced that more than 600,000 cars under its brand have been equipped with fully automated driving chips.

Although Tesla’s current automated driving technology mainly relies on cameras and millimetre-wave radar for real-time calculations and does not rely on high-precision maps, some industry experts have analyzed and told us that for L3 and above levels of automatic technology would rely on high-precision maps.  In other words, the machine must completely replace human monitoring of the environment, and high-definition maps have become a necessary option for safety purposes.    

Therefore, Tesla’s replacement of Tencent maps at this time seems to serve two purposes: first, Tesla uses Baidu map navigation to improve the current dissatisfaction of Tesla users with car navigation; second, Baidu is in the field of high-precision maps, the layout and leading advantages may also support Tesla’s possible technical route changes and adjustments in the future.

01 – Autopilot

Self-driving cars, also known as automated cars, use artificial intelligence, visual computing, radar, monitoring devices, and global positioning systems to work together so that computers can automatically and safely operate motor vehicles without any human initiative.

According to the internationally defined standards of the American SAE Association, automatic driving can be divided into five levels according to the difference in the technical level. Tesla’s AutoPilot technology, which can be accessed daily, is a typical representative of Level 2 self-driving technology.

Automated driving technology can only land in a limited environment. It requires the support of positioning technology, high-precision map automatic generation technology, and reliable human-computer interaction. Even though the above-mentioned new technologies have all made improvements, their huge data scale and strict requirements on transmission rates have become a difficult problem that current 4G network technologies cannot effectively meet.

02 – Automated driving in the 5G era

The future is beautiful, but the reality is cloudy.

According to industry insiders, as of now, there is still no self-driving car in the world that reaches the L3 level and is produced on a mass level. Most of them are just in the testing stage and fail to meet the standards for safe roads.

5G refers to the fifth-generation mobile communication network. According to the vision of the International Telecommunication Union (ITU), the application scenarios of 5G should be divided into three types: Enhanced Mobile Broadband (eMBB), Massive Machine Type Communication (MMTC), and Ultra-Reliable and Low-Latency Communication (URLLC). According to the public data compiled by us, compared with 4G networks, the transmission rate of 5G is increased by 100 times, and the peak communication rate can reach 20Gbps. At the same time, the 5G network can simultaneously support 10 times the terminal equipment of 4G, and the air interface delay is as low as 1 millisecond.

According to China Mobile’s related estimates, under the existing 4G network delay conditions, when an automated vehicle is travelling at a steam speed of 100km/h, it will move at least 1.4 meters from the discovery of the obstacle to the activation of the braking system. However, in the case of a 5G communication network, it moved only about 3cm after receiving certain feedback. It is precise because 5G networks have low power consumption and large connections, low latency, high reliability, and large capacity and wide coverage that takes high-level automated driving from theory to reality.

Baidu Autonomous Driving

At present, the implementation and commercialization of 5G in the Chinese market is far ahead of other countries, which makes it possible for China’s automated driving industry to occupy the highest point in the industry. Among them, Baidu’s Apollo plans is at a leading position in China.

03 – How likely the possibility of Baidu & Tesla’s cooperation?

Public information shows that Tencent owns 5% of Tesla’s equity and ranks as its fifth-largest shareholder. Although it is only a financial investor, Tesla is so “disregarding the situation”, it seems to have a deep meaning that Tencent maps are replaced by Baidu maps. In addition to the current poor navigation experience of Tesla Chinese users, the gradual approach to self-driving has also become an important factor.

Baidu autonomous vehicle manufactured by FAW-Hongqi

If global automated driving is led by Waymo, Tesla and other companies, China’s automated driving industry is represented by Baidu Apollo.

As early as April 19, 2017, Baidu released its new plan called “Apollo”, which aims to provide an open, complete, and secure software platform to partners in the automotive industry. Give full play to Baidu’s technological advantages in the field of artificial intelligence to promote the development and popularizing of automated driving technology.

As of now, the “Apollo” plan has attracted the participation of Chinese top car companies such as BYD Auto. On February 15, 2018, an automated driving team consisting of Baidu, BYD, Golden Dragon Bus, and IDRIVERPLUS appeared on the Spring Festival Gala. The strength of “Apollo” was once again displayed in front of Chinese.

Baidu Apolong Unmanned Vehicle

Baidu develops self-driven vehicles for many reasons: First, automated driving, as one of the three major application scenarios in the field of AI, is fully compatible with Baidu’s “ALL IN AI” strategy; second, Baidu’s has done many years of research and development for technology accumulation in the field of automatic driving; and finally, Baidu’s technology and data accumulation in the field of maps, gives it certain advantages in the field of high-precision maps.

As Tesla’s self-driving technology is self-developed intelligence, its compatibility with Baidu Apollo’s platform-based system is weak. Therefore, some people think that Tesla values the high-precision maps for Baidu’s automatic driving. Up to now, Baidu has independently collected and produced high-precision maps to record complete three-dimensional road information, and can achieve vehicle positioning with centimetre-level accuracy.

Tesla’s current automated driving technology mainly relies on cameras and millimetre-wave radar for real-time calculations and does not rely on maps. However, industry experts have analyzed that for the automatic technology at the L3 level and above, because of the need to free people’s eyes, that is to say, the machine must completely replace the human monitoring of the environment, high-definition maps have become a necessary option for safety purposes.

Up to now, high-precision maps have been adopted for Audi A8, Cadillac Super Cruise and other models that have been in mass production or are close to mass production. Level 4 autopilot solutions from manufacturers such as Waymo and Baidu have also introduced high-precision maps.

Just on the evening of January 16, Tesla’s official Weibo announced that more than 600,000 cars under its brand have been equipped with fully automated driving chips. It is reported that the chip called a “full self-driving computer (FSD)” has 6 billion transistors, it is able to perform 144 trillion calculations per second, as well as simultaneously processing 2300 frames of images per second.

As the most valued self-driving technology in the field of EV brands, the landing of high-level self-driving technology is of great significance to Tesla. Although Musk once again announced that he will achieve fully automatic driving in 2019, the release of Tesla’s fully automatic driving chip is undoubtedly an important development.

In the future, we will continue to pay attention to whether Tesla’s self-driving technology will adopt high-definition maps and whether it will cooperate more widely with Baidu in the field of automated driving.