Smart car

Detailed explanation of road test report of automatic driving vehicles in Beijing

The following is the Detailed explanation of road test report of automatic driving vehicles in Beijing recommended by And this article belongs to the classification: Smart car.

The most authoritative and referential annual report on automatic driving in China? It must be “Beijing road test report for autonomous vehicles”, which is the only official road test report for autonomous vehicles in China. Whether it is the inspection dimension, software and hardware indicators, testing standards, or the latest progress and direction, it is authoritative and comprehensive every year. Although the epidemic has an impact on the industry, the 2020 annual report, which has just been released, has many bright spots and made more progress than before.

For the first time, the unmanned test (similar to the unmanned test license that Baidu recently obtained in Silicon Valley, California), the landing trial operation of robotaxi, the special weather test, and the new news of sensors

Through this annual report, on the one hand, we can follow the current situation of China’s automatic driving technology, on the other hand, we can clearly see the industry pattern. In addition, we may be able to find out the potential reason why Baidu’s share price has risen sharply in recent years (4% after the report) and its market value is approaching 100 billion US dollars.

As usual, let’s pick the key points and explain them one by one.

New achievements: road test mileage increased significantly, safe driving mileage increased by 112.8%

Compared with last year, the most intuitive road test mileage has increased significantly.

As of December 2020, Beijing’s safe driving mileage has reached 2.2134 million km, an increase of 112.8% over the end of 2019.

According to the data, the epidemic has had a certain impact on automatic driving. The mileage growth slowed down in January and gradually recovered in March.

Among them, Baidu has no suspense in the mileage, with 1125300 km ranking first.

Baidu’s road test mileage is 20 times more than all other players combined. Or in other words, 96% of the mileage of Beijing road test this year is run by Baidu, which also exceeds the total mileage of automatic driving test last year.

Moreover, such test mileage is completed in different test areas and road grades.

Previously, Beijing’s road test difficulty was divided into four levels (R1-R4). At present, only one enterprise of Baidu has conducted R4 level test.

In addition, only Baidu has obtained the highest technical level, the highest standard, and the most difficult open road test qualification certification (Beijing T4 license).

Therefore, on the one hand, Beijing is putting forward higher dimensional regional test requirements, but on the other hand, baidu is the only one that has been allowed to win the high-level license.

In Beijing, Baidu’s snowball effect in automatic driving road test is quite obvious.

New news: robotaxi landing in Beijing for the first time, 90% of passengers think AI drivers are reliable

Besides road test mileage, manned test is another highlight of this year.

As the capital, Beijing has higher technical and safety requirements than other regions in terms of robotaxi manned.

From August 2020, Beijing robotaxi public trial operation was also initiated by Baidu. After December 2020, Xiaoma Zhixing and other companies began to join.

In fact, robotaxi has always been a key scene for the landing of automatic driving, and also a commercial focus of attention from the outside world. It is basically a “must fight” place in the field of manned automatic driving.

Robotaxi’s user experience is also an intuitive feedback on the level of technology.

According to the survey results, 63% of the interest groups in the test ride are “old drivers” (with more than 6 years of driving experience), and 60% are young people aged 18-35.

For robotaxi, the overall satisfaction of volunteers was high, and 97% of them even said they were willing to accept the paid service.

In terms of “automatic driving” and “human-computer interaction”, nearly 90% of the users gave “good” evaluation.

On the whole, robotaxi in Beijing (mainly Baidu) is still highly recognized, and its future commercial landing is expected.

New wind direction: start unmanned test for the first time

This year, Beijing also has the latest progress: (for the first time) to start the unmanned automatic driving test.

At present, 14 enterprises and 87 vehicles have obtained the general road test license of automatic driving in Beijing

However, after its first opening in 2020, only one company, Baidu, has obtained the unmanned test license.

In December 2020, baidu passed the unmanned special technical test for the first time and was awarded the first batch of 5 unmanned road test (phase I) “licenses”.

Compared with other cities, Beijing has set an example in terms of the length of open roads and the degree of standard setting.

In addition, a 64827 km unmanned test verification was set up to confirm the feasibility of the test technology and the reliability of the test methods and parameters.

As a wind vane, the first opening of the unmanned test means that the automatic driving technology is further away from the “safety officer”.

From the beginning of the public test of robotaxi, the ultimate commercial product form of automatic driving has made a faint sound.

New trend: accelerating the localization of core components

In the report, regarding the key parts of the automobile, it is also clearly disclosed that there is a trend of further localization.

For example, the brand of laser radar adopted by domestic autopilot cars is 71% made in China, and the number of lines is below 40 lines.

Compared with 60% in 2019 and 80% in 2018, velodyne, the originator of lidar, is no longer dominant, with its share falling to 29%. Instead, it is domestic lidar.

And the car itself, especially the new energy vehicles, also has the trend of further localization, and even the pure electric models have all been localized this year.

What’s more gratifying is that China has also made a set of testing scheme with lower cost and higher practicability.

This year, Beijing intelligent car alliance and Baidu jointly developed a hybrid reality test system, which can inject some traffic targets into the real field to realize scene test.

If the car is tested in this closed test field, the mileage effect can reach more than 100 times of the open road test.

In addition, in terms of automatic driving test standards, Beijing has also improved this year: in the test scenario, special weather has been added for the first time.

From the results, the perception performance of automatic driving vehicles in foggy days can only reach 45% of that in the daytime, which means that there is still a certain time for automatic driving vehicles to be on the road all day.

However, it also shows that the requirements of Beijing road test are both strict and close to the real extreme environment.

Therefore, even if there are many cities that are carrying out road test now, automobile enterprises may still have to “conquer” Beijing road test standards in the future to truly prove the level of automatic driving.

New challenge: centimeter level positioning, key reasons for departure

After the closed field test this year, an average of 18100 km, there will be a collision accident (not avoiding pedestrians and running over the road in time). Compared with last year, the average distance has increased by nearly 46%.

On average, 67% of these accidents are caused by abnormal positioning (unless centimeter level positioning is achieved), and the remaining 33% are caused by perception errors.

Further improving the positioning and perception is the fundamental method to ensure that the automatic driving vehicles have fewer accidents.

In the test of open road, separation is also one of the important reference indexes.

In the critical disengagement, the disengagement between traffic participants and targets accounts for a large proportion (about 93%) of all critical disengagements. The most common situation is “target occupied lane”, accounting for 25%.

Road construction (22%) and intersection game (17%) are also the key reasons for automatic driving separation. In addition, retrograde, cross, cut, too close and other reasons, accounted for a total of 36%.

Fundamentally, these detachment is mainly caused by the automatic driving system “no violation of traffic rules”. As a result, the test vehicle can not pass through flexibly like human drivers.

In reality, it is inevitable that there will be human violations of traffic rules and other emergency situations, which is a part of automatic driving landing must be considered, and vehicle road cooperation is one of the current solutions.

Of course, understanding a series of “new” knowledge points disclosed in this year’s road test report, as an annual “answer sheet”, can also be used as the latest evaluation of strength level.

Which is the best auto driving company in China?

Still Baidu Apollo.

Continuing the road test report in Beijing last year, baidu Apollo is still firmly in the leading position this year.

In terms of test mileage, among the 14 car companies in Beijing, the ladder gap is still very obvious.

First of all, in terms of echelon division, Baidu and Xiaoma Zhixing have opened a big gap for other players.

Secondly, the actual difference between the first Baidu and the second Xiaoma Zhixing is still very big.

At present, Baidu’s accumulated mileage has reached 2.0192 million kilometers, while Xiaoma Zhixing, ranked second, is 1.632 million kilometers. Other players, however, have almost reached a fraction of them.

From the perspective of duration, Baidu’s automatic driving road test is also in the absolute lead.

In addition, judging from the current progress level of manned test and unmanned test, baidu Apollo manned test has entered the third stage, and unmanned test has officially entered the first stage, and there are no other players at the same stage.

In short, as the only official automatic driving road test report in China, Beijing automatic driving road test annual report reveals many new directions, but it is still the old pattern

Baidu Apollo, the more you run, the more nobody can.

And this objective result, in fact, also provides a new “material” for another hot topic recently.

How much is Baidu Apollo autopilot worth?

From the second half of 2020, the automatic driving industry will start a new wave of high incidence of valuation and financing.

Rough statistics show that since 2020, there have been nearly 30 financing cases for autonomous driving, with a total amount of RMB 100 billion.

Among them, waymo’s financing exceeded $3 billion, cruise’s new financing led by Microsoft was $2 billion. In addition, on the start-up track, the leading company pony Zhihang also gained $267 million in investment, accumulating more than $1 billion in financing.

Among these investments and financing, waymo, the global pioneer of automatic driving, is estimated to be worth US $105 billion, cruise is estimated to be worth US $30 billion after the latest investment, and pony wisdom bank is estimated to be worth more than US $5.3 billion.

So how much should Baidu Apollo’s autopilot business be valued? In the hot market, many well-known analysts have different opinions on the valuation of China’s autonomous driving geese.

For example, Goldman Sachs, for the automatic driving business, has given Baidu Apollo a valuation of $20 billion in the previous research report, but combined with the comparison of reference frame, the industry generally feels that it is undervalued.

On the one hand, waymo in the United States and Apollo in China have become a phenomenon in the field of automatic driving. Moreover, according to the landing speed of automatic driving, baidu Apollo, which is supported by v2x in China, may be able to actually land automatic driving faster than waymo.

This is also the key reason why Cathie wood, the “Queen of the bull market”, is optimistic about and bought Baidu on a large scale after Tesla’s share price and market value soared due to its automatic driving.

Moreover, in the latest report of Navigant research, an authoritative research organization in the industry, baidu Apollo, waymo and cruise are also ranked in the first echelon.

Cruise, on the other hand, is now offering a mass production car and robotaxi solution valued at $30 billion. And this, is not in the same echelon of Baidu Apollo’s automatic driving business?

As far as the automatic driving business is concerned, there are many similarities between them.

Cruise is backed by GM, with a fleet size of about 300-500 vehicles; Baidu is mainly based on Hongqi EV, and FAW has provided Baidu with a L4 front loading production line in Changchun. At present, the overall scale of Baidu Apollo vehicles is about 500.

In terms of business model, cruise takes the route of robotaxi + car making, and Baidu also defines the robotaxi + Maas (travel as a service) platform in terms of automatic driving business.

So cruise is really suitable for Baidu Apollo’s automatic driving business.

But this is not road test, because thanks to Baidu Apollo’s recognition at home and abroad, road test is one position ahead of cruise.

Both of them have obtained all unmanned licenses in California, and both of them have entered urban road sections, except that Baidu Apollo, in addition to cruise’s California road test, also has nearly 30 cities / regions in China, such as Beijing and Changsha.

In terms of mileage, cruise is currently 3.2 million km, while Baidu has broken through 7 million km.

Therefore, if the valuation of Baidu Apollo autopilot is “only a lot more” than Cruise’s $30 billion, it is not groundless.

Of course, there are also some analysis from investment banks. Before that, UBS valued Apollo’s autopilot at $10 billion, which is one third of waymo’s, which is obviously underestimated. And Morgan Stanley has given waymo a valuation of $105 billion. According to a third of UBS’s logic, Apollo’s autopilot part is valued at more than $30 billion.

This is also why Apollo’s current autopilot business has an “external evaluation” with a valuation of more than $30 billion.

However, it should be noted that the above valuations occurred before the annual report of Beijing road test. After the report is disclosed, it may be further revalued.

what you think?

Asahi Shimbun: China overtakes McKinsey in the field of automatic driving technology: China will become the largest automatic driving market in the world Too horrible to look at, the 2030 auto driving and travel service revenue will exceed 500 billion dollars Rice University & Texas Tech University: research finds that self driving cars will make drivers feel complacent. ReportLinker:70% American drivers love to drive their own cars. The ranking of global auto driving strength is announced: Google first, Baidu seventh, apple miserable Ipsos: the survey shows that compared with using automatic driving Most people love to drive in California. 2019, the automatic driving system is divorced from the TOYOTA super Baidu Apollo 270 times Frost: automatic driving will contribute tens of billions of pounds to the British economy every year. The Japan economic news: Google’s Waymo is “taking the lead” in the field of automatic driving technology Navigant Research:2019 auto driving power ranked the bottom of the survey report, which shows that people do not like full automatic driving but the car companies do not think so. Reuters & amp; Ipsos: the survey found that 2/3 Americans do not believe in the autopilot car Bloomberg Charity Fund: 2017 global smart driving city distribution J.D.Power: China’s consumer love driving tendency survey

If you want to get the full report, you can contact us by leaving us the comment. If you think the information here might be helpful to others, please actively share it. If you want others to see your attitude towards this report, please actively comment and discuss it. Please stay tuned to us, we will keep updating as much as possible to record future development trends. is a website that focuses on future technologies, markets and user trends. We are responsible for collecting the latest research data, authority data, industry research and analysis reports. We are committed to becoming a data and report sharing platform for professionals and decision makers. We look forward to working with you to record the development trends of today’s economy, technology, industrial chain and business model.Welcome to follow, comment and bookmark us, and hope to share the future with you, and look forward to your success with our help.

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button