If in the past ten years, smartphones, as a killer application, have driven the entire industry forward, in the next ten years, smart cars are being considered as the ultimate driver in many fields.

As the so-called industry rules are often changed by alien species, it is often not your peers who beat you. How to follow the general trend and stand on the wind? Facing the coming new era of smart cars, outside the traditional car factories, technology giants have begun to plan ahead and enter the game first.


The development of automobiles has gone through hundreds of years. Since the era of electrification, more and more semiconductor Electronic components have been used in automobiles, and innovations in the automobile field have also been triggered by this. Automobiles are no longer mechanized in the traditional sense. The device is like a mobile computing center on four wheels.

Facing the vast territory of traditional car factories, technology upstarts such as Internet giants can’t wait.

On January 11, Baidu officially announced that it will form a smart car company to enter the automotive industry as a vehicle manufacturer, and Geely Holding Group will become a strategic partner of the new company. Baidu said in a statement that the newly formed Baidu Auto Company will target the passenger car market and allow users to purchase more extreme smart electric vehicles.

Tech giants are rushing to enter the game, and the smart car industry is on the forefront?

As an Internet company started as a search engine, it is not surprising that Baidu makes a car. In the past few years, Baidu has been exploring the field of automotive intelligence for many years. The Apollo Apollo autonomous driving platform launched has a large number of ecological partners from hardware, software to OEMs. This time Baidu established an automobile company, and its core technologies such as artificial intelligence, Apollo autonomous driving, Xiaodu vehicle, and Baidu map will fully empower the new automobile company. Next, Baidu Auto will jointly build the next generation of smart cars based on Geely’s newly developed pure electric architecture – the vast SEA intelligent evolution experience architecture.

On January 4, Byton Motors signed an agreement with Foxconn Technology Group to accelerate the mass production of Byton’s new energy vehicles. Foxconn said that this is an important measure for Foxconn to deploy in the electric vehicle field. Foxconn will work together with Byton to mass-produce it as soon as possible to achieve mutual benefit and win-win results.

Prior to this, Alibaba and SAIC Group had announced the launch of the high-end smart electric vehicle brand “Zhiji Auto”, and two Zhiji electric vehicles were first released on January 13. Zhiji Auto said that under the conditions of national regulations and the opening of high-precision maps, it will have the ability to take over automatic driving from point to point by the end of 2021; it will also apply automatic valet parking and car call functions in selected commercial centers in first-tier cities during the same period. Alibaba’s rich big data and DAMO Academy’s technical support will help Zhiji Auto achieve a deeply intelligent and personalized experience.

Korean media recently reported that Apple and Hyundai Motor are planning to sign a partnership agreement in March on electrified self-driving cars, with production expected to begin in the U.S. as early as 2024. Like Baidu and Ali, Apple’s car project focuses on developing autonomous driving technology. Correspondingly, Taiwanese media reported that Apple is working with TSMC to develop autonomous vehicle chips. As a trend-setter in the era of smartphones, in the era of smart cars, Apple is naturally unwilling to be left behind. To continue to dominate the industry, it seems that it is the only choice to find cooperation with traditional car manufacturers. As of now, Apple has not given an official response to this.

Unlike the above manufacturers who end up building cars in person, Huawei has clearly stated that it will not build cars, but it has positioned itself to focus on ICT technology and is committed to becoming an incremental component supplier for Internet-connected cars. At the software level, Huawei has launched a smart cockpit operating system (based on Hongmeng microkernel), smart driving operating system, and smart car control operating system. At the same time, in terms of hardware, Huawei’s processor chips and 5G communication chips have begun to deploy in-vehicle applications , With its leading position in 5G technology, Huawei’s “5G automotive ecosystem” has included many traditional car manufacturers such as BYD and FAW. The newly launched BYD Han uses Huawei’s 5G communication module.

Like Huawei, Waymo, an autonomous driving pioneer independent from Google, also stated that it does not build cars and will focus on autonomous driving technology research. At present, Waymo has reached cooperation agreements with traditional car companies such as Chrysler and Daimler.

On the road to smart electric vehicles, traditional car factories are obviously unable to do what they can do, and technology upstarts from the Internet and mobile phone fields are entering the field one after another with their strong technologies in the field of software and hardware, and a new round of staking has begun. He Xiaopeng, founder of Xiaopeng Motors, said in a Weibo post: “As far as I know, there are probably several technology companies that may become new car-making forces in different models this year. Positive acceleration impact.”

Can self-driving smart cars accelerate their arrival?

The emergence of new car-making forces will definitely strengthen the strength of smart car manufacturers, but when highly intelligent autonomous vehicles can be used, the industry has different expectations.

At the World Artificial Intelligence Conference held last year, Tesla CEO Elon Musk said: “At Tesla, we are very close to L5 level automatic driving, using Tesla’s current hardware, only need to improve the software, I have Confidence to complete the basic functions of the L5 level this year. I think there are no fundamental challenges at the bottom to achieve the L5 level of autonomous driving, but there are many details, and the challenge we face is to solve all these small problems.”

But in fact, there are still very few brands that achieve the L4 level in commercial vehicles, and the realization of some scenarios based on the L4 level has become the primary goal of the current autonomous driving manufacturers.

In the 0-5 level of driving automation, it is generally believed that the L4 level is the beginning of the true sense of autonomous driving. A car that reaches the L4 level can drive on the road in a specific area without human intervention, including automatic overtaking, braking, driving into the destination, etc. In order to achieve full sense of autonomous driving without a driver, it needs to reach the L5 level.

At present, Tesla, Weilai, Xiaopeng, Weimar and other new car manufacturers have tested the L4 automatic driving function in some scenarios. For example, automatic parking, smart calling, etc. However, as radical as Tesla, it only calls its Autopilot an automatic assisted driving function in the publicity of its official website. However, Tesla said that it will continuously improve the function through OTA software updates and will achieve fully autonomous driving in the future.

Waymo and Baidu, which focus on autonomous driving platform technology, have conducted a large number of road test experiments for highly autonomous vehicles. In October last year, Baidu launched a self-driving taxi service in Beijing. Citizens can book a ride at the self-driving taxi site. Baidu self-driving taxis are equipped with safety officers, but there is almost no need for safety officers to intervene during the whole process, and the vehicles can accelerate automatically. , avoid, turn, wait for the red light, and drive into the destination. Waymo has also launched self-driving road test experiments in many cities in the United States, especially in Phoenix, the United States has opened a quasi-fully self-driving road test without a safety officer, but the test area is limited to a limited range, and the actual road conditions are far away Not as busy and crowded as big city roads.

The intelligence required for autonomous driving not only requires strong computing power support, but also requires massive measured data to provide simulation training. The data from the RAND Corporation in the United States shows that if it is to prove that the fatality rate of traffic accidents caused by autonomous vehicles is lower than that of human driving , requiring at least 5 billion miles of actual road testing. This is a huge challenge for the commercial use of autonomous vehicles.

Unlike Musk’s extreme optimism, Waymo CEO John Krafic said in an interview with the media, “(Making a self-driving car) is more challenging than launching a rocket and sending it into orbit around the earth, because it The test has to be done safely over and over again.” He also believes that L5 fully autonomous driving is almost impossible to achieve, one of the main reasons is that the problem of self-driving car sensors not working properly in rain or snow has not yet been solved.

In addition to autonomous driving, smart cockpits, battery life, safety, etc. are all challenges for the future of smart cars. Who can win this competition? Stay tuned!

The Links:   CM300DU-24F FS100R12KE3-B3