"DoubleCity" fully driverless vehicle commercialization, China's autonomous driving takes the lead on a "global level."

Author: Jaz

2022 may be a crucial turning point for China’s autonomous driving policy opening.

Recently, Chongqing and Wuhan respectively released the “Yongchuan District Intelligent Connected Vehicle Policy Pilot Area Road Testing and Application Management Measures” and the “Implementation Rules for Intelligent Connected Vehicle Road Testing and Demonstration Application Management (Pilot) in Wuhan”, becoming the first batch of fully autonomous driving commercial policies in China.

Among them, Baidu is the only company approved to operate in both “twin cities”. It is reported that Baidu’s automatic driving travel service platform, Apollo Go, will officially provide unmanned autonomous driving pay-per-ride service in the “twin cities”.

The “twin cities” of Chongqing and Wuhan have taken the lead in launching fully unmanned autonomous driving commercial demonstration and operation services in vehicles. This means that China’s autonomous driving has made a historic leap from “driverless” to “fully unmanned”, deeply activating the commercial landing of autonomous driving and bringing autonomous driving one step closer to the lives of ordinary people.

This “leap” has put China’s autonomous driving in a leading position in the world. Compared with the United States, which is home to many autonomous driving companies, China not only has prominent autonomous driving enterprises but also has more advanced policies.

It is reported that only California and Arizona in the United States can conduct commercial operations, while China has already had 12 cities that provide autonomous driving services to the public and 23 cities that conduct large-scale autonomous driving tests.

It can be predicted that as China’s autonomous driving policy goes deeper, autonomous driving will become a “hard technology trump card” for China.

It can be found that commercial autonomous driving is becoming a benchmark. One end is used to measure the progress of urban intelligence and modernization.

It can be seen that most first-tier cities in China have already embarked on the process of opening up autonomous driving policies, opening up imagination space for autonomous driving landing through legal support and pilot construction.

The other end measures the ability of technology companies to build autonomous driving closed loops. With the space to develop, domestic technology companies focusing on autonomous driving are increasingly “inward-looking”, from the number of autonomous driving fleets to technological breakthroughs. However, due to the high overall cost of autonomous driving and difficulties in commercialization, only Baidu has stood out among commercial operators.

The reason why Baidu became the only approved enterprise in the “twin cities” of Chongqing and Wuhan is precisely because of its confidence in commercial operation. This also means that Baidu has already stood at the forefront of the commercialization of autonomous driving.

Why is commercial operation the most critical part of autonomous driving?

Without commercial operations, autonomous driving cannot be established.

This may sound crude but it is true. Without commercial operations, companies cannot rely on autonomous driving to generate revenue, nor can they promote it in real life. Ultimately, this will push autonomous driving companies to the brink of death.In recent years, Cruise has laid off 150 people, Drive.ai has been acquired by Apple, Starsky Robotics has gone out of business, and Nuro, which was reported by foreign media The Information in 2020, is also seeking to sell itself. All of these directly confirm that slow business progress has become the biggest killer of autonomous driving companies.

In other words, before entering commercialization, autonomous driving was just a “high-end technology” in an ivory tower, and it is not until after commercialization that autonomous driving can be called a concrete and complete “industrial blue ocean”.

Fortunately, China’s autonomous driving policies and enterprises are gradually moving towards the correct path, creating a “China sample” for autonomous driving.

At present, autonomous driving has received policy support, and on the one hand, domestic policies are gradually becoming more open.

In the past two years, cities in China only allowed autonomous driving testing and test rides on some sections of roads in economic development zones. But in the past year, many cities have turned to commercial operation of autonomous driving on a large scale. Among them, they have shifted from manned operation to “no human behind the steering wheel” operation.

Beijing is a typical representative. In 2020, Beijing allowed autonomous driving testing and test rides in Yizhuang and Haidian. Just one year later, Yizhuang opened 100 kilometers of urban roads for unmanned testing and allowed robotaxis to operate without a human driver.

On the other hand, the regions covered by autonomous driving policies are spreading from super-large cities to first-tier cities across the country.

Cities like Beijing and Shenzhen are no longer exceptions. The “Double City” of Wuhan and Chongqing released the first batch of national policy for fully unmanned commercial pilot of autonomous driving, which represents that autonomous driving is breaking through in first-tier cities across the country.

According to autoX, companies that obtain operating licenses after the policy is implemented can conduct remote testing, demonstrations and commercial operations of unmanned vehicles in Chongqing and Wuhan.

The opening up of the “Double City” of Chongqing and Wuhan means that autonomous driving policies have entered a new open stage.

From the perspective of enterprises, it is not difficult to understand why Baidu became the only company to obtain an operating license in the “Double City”. This is not only a demonstration of Baidu’s autonomous driving strength but also the accumulation of advantages in autonomous driving commercial operations.

For example, the mountainous terrain in Chongqing is not exaggerated, and there are many large overpasses. For example, the Huangjuewan Bridge in Chongqing has a very complex road condition. This bridge has five layers, each leading to different directions, and has 15 ramps, connecting important nodes in the city.

These complex conditions put extremely high safety requirements on autonomous driving. The prerequisite for obtaining an operating license is safety and technological leadership to ensure full redundancy. Therefore, Baidu’s acquisition of an operating license in the “Double City” obviously indicates that Baidu has passed the most fundamental technical challenge in autonomous driving.In addition, the common point of the policies released by the “Twin Cities” is that they both allow commercial operation demonstrations, which requires companies’ autonomous driving commercial operation capabilities. Baidu Robotaxi has the most mature commercial operation experience and has entered the commercial operation stage since the launch of “Luobo Kuaipao” last year.

Therefore, Baidu’s acquisition of the “Twin Cities” is not accidental, but the inevitable result of many years of deep cultivation in the field of autonomous driving.

Baidu “Breakthrough” and Enter the Autonomous Driving “No Man’s Land”

In the field of autonomous driving, there are many “specialists” and few “all-rounders.”

If compared in terms of individual capabilities, some companies win with their Robotaxi fleet vehicles, some have outstanding performance in actual test mileage, and some have relatively strong financing capabilities.

However, the longboard does not bring incremental benefits, and the short-board effect that followed also made it difficult for these “specialists” to move forward. For example, Zoox, which once set a single-round financing record for independent startups, was sold to Amazon for $1 billion in 2020, focusing on unmanned distribution.

Baidu’s autonomous driving is different. It is more like an “all-rounder.”

Currently, Baidu has achieved a breakthrough in autonomous driving technology, safety, products, and operations.

In terms of technology, Apollo autonomous driving platform is the foundation. In terms of safety, there are 5G cloud valet and a vehicle’s multiple redundant system. In terms of products, there is Apollo RT6, a mass-produced unmanned vehicle for autonomous driving. In terms of operations, there are Luobo Kuaipao and the leading commercial operation experience.

From the core technology point of view, Baidu is the “Old Money” in the field of autonomous driving. It began to lay out autonomous driving in 2013, and the Apollo autonomous driving open platform was launched in 2017. According to the latest data, Baidu autonomous driving has accumulated more than 4,000 China patents, ranking first in China.

One of the key indicators for measuring autonomous driving technology is the scale and performance of autonomous driving tests. Baidu Apollo’s test fleet has not only more than 500 vehicles and a total test mileage of more than 32 million kilometers, but also the testing environment on domestic roads is more complex than that in California and Arizona, where Waymo and Cruise are located. Nevertheless, Baidu still achieves safe operation with zero accidents.

In 2020, Baidu mentioned the three essential elements for fully autonomous driving at the World Conference: “mass-produced vehicles,” “AI drivers,” and “5G cloud valet.” The supporting service, “5G cloud valet,” was born to ensure the safety of the full-car autonomous driving scenario. This system can ensure that the “online” valet remote controls the vehicle in case of sudden situations such as temporary road changes and traffic restrictions, thereby ensuring the safety of the vehicle.Only one year later, Baidu has upgraded its consideration of autonomous driving safety once again and has extended its safety considerations from technical penetration to autonomous driving products.

At the recently-concluded 2022 World Congress, Baidu’s mass-production vehicle was upgraded to the new Apollo RT6, a next-generation autonomous vehicle. In addition to the 5G cloud driving, the seven-layer safety redundancy of the Apollo RT6 has added a new layer of security to autonomous driving.

It is reported that Apollo RT6 has seven-fold redundant vehicle systems including architecture redundancy, sensor redundancy, computing unit redundancy, brake system redundancy, steering system redundancy, power supply redundancy, and the industry’s first dual 5G communication redundancy.

At the same time, the Apollo RT6 software is equipped with a fault diagnosis and risk reduction system that integrates the whole vehicle and automatic driving system as one.

In fact, Apollo RT6 is the last hurdle for Baidu to move towards commercial operation. As the world’s first truly mass-producible autonomous vehicle, Apollo RT6 has passed the safety, technology, and price hurdles with seven redundancy, eight lidar, and Apollo autonomous driving technology, at a cost of only CNY 250,000.

Finally, with the gradual opening of policies, the problems that previously plagued the commercialization of autonomous driving, such as the inability to achieve true unmanned and low prices, have been solved.

It is reported that the monthly cost of Apollo RT6 is only CNY 4,100, which will be lower than that of ordinary online car-hailing and taxi prices after putting into operation.

If technology, safety, and products have paved the way for Baidu’s commercialization, then Baidu has relied on Luobu Racing to gain a first-mover advantage in autonomous driving commercialization.

Last year, Luobu Racing welcomed the first commercialization of autonomous driving in Beijing. As of now, Baidu Luobu Racing has landed in more than 10 cities such as Beijing, Shanghai, Shenzhen, Guangzhou, Chongqing, and Wuhan, with an order volume of 1 million.

With the operation of Apollo RT6 in Luobu Racing, the number of large-scale autonomous driving commercialization orders and landing cities will continue to rise. It is reported that by 2025, Luobu Racing will be put into operation in 65 cities.

Thus, Baidu’s autonomous driving layout has taken shape.

The meeting between Baidu and Tesla, but missed chances to cooperate

When Baidu took the step towards commercialization, more and more voices emerged comparing Baidu to Tesla.

Baidu and Tesla’s meeting in the middle? Perhaps Baidu, which once made autonomous driving, and Tesla, which makes cars, were not related, but Baidu is now making “autonomous vehicles” with Apollo RT6, while Tesla is entering Robotaxi, making the world aware that Baidu and Tesla have stood on the same stage.But even on the same stage, their styles are vastly different.

First of all, Baidu’s autonomous driving technology takes a completely different path from Tesla’s. Baidu’s approach involves vehicle-to-road coordination.

Baidu founder, chairman and CEO Robin Li has stated that using vehicle-to-road coordination for autonomous driving is the technological path that Baidu insists on and is optimistic about. Vehicle-to-road coordination means real-time dynamic interaction among vehicles, roads, people and the environment.

Compared to Tesla’s primarily vision-based autonomous driving, which has developed rapidly in the early stages with fewer limitations, Tesla has overlooked future traffic patterns in society from a long-term perspective.

Baidu’s vehicle-to-road coordination not only has a safer technology path than Tesla’s single vehicle intelligence line, but it also places autonomous driving in a broader context – where autonomous driving is merely the starting point for intelligent transportation, intelligent cities and even intelligent societies.

This means that Baidu’s autonomous driving, other than Robotaxi, can have a wider participation in China’s intelligent transportation construction.

Secondly, China’s autonomous driving policy has entered into a gradual period of openness, which will provide more room for Baidu’s autonomous driving to perform.

Especially with the addition of the “twin cities” representated by Chongqing and Wuhan, it further illustrates the “openness” trend. Meanwhile in the United States where Tesla is located, attitudes toward autonomous driving differ from state to state, and the commercial operation of Robotaxi has yet to be further expanded. For example, the average quarterly order volume for Waymo and Cruise is only 33,000 units, which is far from the number of orders on Chinese ride-hail platform Didi Chuxing.

Lastly, Baidu has a first-mover advantage in Robotaxi, as Baidu’s unmanned vehicles have been updated to the sixth generation, with Apollo RT6 dedicated to autonomous driving research, while Tesla’s unmanned vehicles are still in the planning phase.

As early as April 2019, Elon Musk proposed the Robotaxi plan, with a deployment of one million autonomous taxis planned by the end of 2020.

However, although more than half of 2021 has passed, there is still no news of phased progress for Tesla’s Robotaxi. The only known information is that Musk revealed in April that Tesla’s Robotaxi model is in the development stage, without pedals or a steering wheel, and equipped with the FSD fully autonomous driving system.

The meeting between Baidu and Tesla is not a matter of who is superior or inferior. Based on the current progress, according to Baidu’s plans for Apollo RT6, the vehicles can be put into trial operation on Didi Chuxing as early as the second half of next year, while Tesla’s Robotaxi is still nowhere to be seen.There are still a lot of technical barriers between making cars and making autonomous vehicles, advanced autonomous driving and fully autonomous driving, which makes it difficult for Tesla to catch up with Baidu’s progress in autonomous driving commercialization even if it launches Robotaxi in 2024 as planned.

Recently, Tesla was accused of false advertising regarding its Autopilot and FSD autonomous driving systems by the California Department of Motor Vehicles (DMV).

In its filing, the DMV stated that Tesla has not truly achieved autonomous driving and advertised that “drivers can take their hands off the wheel” in its commercials.

Compared with Baidu’s “all-in-one” approach to autonomous driving, Tesla faces numerous embarrassing issues such as slow progress in making autonomous vehicles and difficulties with policies in entering the Robotaxi market. Therefore, although Baidu and Tesla have met in the middle ground of Robotaxi, there are already signs of Baidu surpassing Tesla.

To summarize, Baidu’s goal in the short term is to win the city breakout battle and capturing two cities marks a milestone. This type of urban breakout battle can fundamentally renovate the business model of the autonomous driving industry.

IHS Markit once had the view that relying solely on the private car market cannot achieve the rapid development of overall L4 autonomous driving as users and carmakers are price-sensitive. The future market potential mainly relies on the commercial model promotion of Robotaxi.

Therefore, the breakout battle of Baidu’s autonomous driving is not only driving the intelligence of urban transportation but also upgrading the entire industry. In combination with more open policies, the commercial operation of Baidu’s autonomous driving can implement national-level incentive measures and achieve positive results in the market.

It is predicted that the penetration rate of new vehicles with L2 or higher will reach at least 45% by 2025 and over 80% by 2030.

According to the current situation of autonomous driving city operations in China, there are 12 cities in China that provide autonomous driving travel services to the public, and 24 cities that carry out large-scale autonomous driving tests. Among them, Beijing Yizhuang, as the earliest city in China to open up autonomous driving, has been crowned “the first city of autonomous driving” by many media outlets.

However, although Beijing was the first to initiate the policy, its progress in open policies compared to Chongqing and Wuhan’s “fully autonomous commercial operation” is relatively slow. At present, it is still a commercialized driverless city with a safety officer in the passenger seat and the main driver seat that can be commercialized.

Earlier, Auto Home had interviewed industry experts on this issue. Professor Deng Zhidong, the director of the Visual Intelligence Research Center of the Artificial Intelligence Research Institute of Tsinghua University, believes that, under the premise of ensuring absolute safety, relevant government departments can encourage innovation through appropriate advanced policy planning and deployment.

From this perspective, if domestic policies on autonomous driving can be opened up as soon as possible, it will massively drive the prosperity of the autonomous driving industry and further build smart urban transportation.# From Myth to Reality: Baidu Leads the Way in Commercialization of Autonomous Driving

In the past, there was a misconception about the autonomous driving industry, with exaggerated financing and worship of autonomous driving geniuses. However, Baidu’s step-by-step breakthrough has truly brought autonomous driving to the ground, from the laboratory to the road, and from high-end technology to the specific industrial chain.

Baidu continues to cultivate the field of autonomous driving, but it is certain that Baidu has become the leader in the commercialization of autonomous driving on a large scale.

This article is a translation by ChatGPT of a Chinese report from 42HOW. If you have any questions about it, please email bd@42how.com.