After Apollo buses go into mass production, what’s the next step for Baidu’s autonomous driving?

On July 4, at the Baidu AI Developer Conference, Robin Li announced that Apollo, the world's first L4-level mass-produced autonomous driving bus jointly developed by Baidu and King Long Bus, has officially rolled off the production line.

At this conference, Baidu also officially released the 3.0 version of Baidu Apollo autonomous driving platform, and introduced Xiaodu in-vehicle OS and four mass production solutions: autonomous parking, unmanned work vehicles, and unmanned shuttle buses.

These achievements show that Baidu's autonomous driving technology has reached a new level, and the goals of mass production and commercial operations are gradually being achieved.

After fulfilling its promise to mass-produce Apollo, what will be the next step for Baidu's driverless car development?

Open Apollo gathers more strength

Baidu's driverless car project started in 2013. Baidu has achieved certain results in the field of driverless cars and has gradually formed its own industrial chain.

In April 2017, Baidu officially announced that Apollo would provide an open, complete and secure software platform to partners in the automotive industry and autonomous driving field to help them combine vehicles and hardware systems to quickly build a complete autonomous driving system of their own.

In just over a year, Baidu Apollo has grown from 1.0, 1.5, 2.0 to 2.5, and then to the just-released 3.0. The speed of its development is astonishing.

Baidu's open plan shows that Apollo 2.5 will support visual high-speed autonomous driving in limited areas, the 3.0 stage will achieve mass production of low-speed campus autonomous driving, and 4.0 will support autonomous driving on highways and urban roads in limited areas.

The mass production of Apollo proves that the goal of Apollo 3.0 has been achieved, and Apollo will further advance to 4.0.

Although Baidu's autonomous driving project started earlier in China and its technology is the most mature, it is still far behind the international leading level. For example, Waymo, a subsidiary of Google, already has tens of thousands of cars in commercial operation of autonomous driving. In terms of data accumulation, Baidu also has a huge gap with leading foreign companies.

In recent years, Baidu has gradually fallen behind in the market value competition among the three giants BAT. Compared with the rapid development of Tencent and Alibaba, Baidu's market value has stagnated. This directly restricts Baidu's investment scale in the fields of AI and autonomous driving.

Compared with the tens of billions of dollars invested by international giants in autonomous driving, Baidu's investment in Apollo is only in the tens of billions of dollars.

Insufficient capital and a late start require Baidu to take a light-asset R&D path for autonomous driving. The open Apollo model is Baidu's secret to catching up.

The open Apollo allows Baidu to absorb domestic competitors into its own platform, forming a joint force and avoiding duplication of work to the greatest extent possible. The strategy of focusing only on autonomous driving software systems and platforms from the beginning allows Baidu to avoid taking on too heavy a burden and invest limited funds in the development of technology.

Open source platforms are not conducive to the formation of monopolies, but at the same time provide Baidu with a broader space for development. By only providing technology and platforms, rather than monopolizing the entire industry chain as much as possible, Baidu can gain more partners with an open attitude.

At this year's Developer Conference, the list of partners for the Apollo project was further expanded: 16 companies including Jaguar Land Rover, Valeo, Byton, Flextronics, ADI, and Suning Logistics joined, bringing the number of Apollo ecosystem partners to 116.

The numerous partners have undoubtedly enhanced Baidu's ability to integrate resources and shape the industry ecosystem. Division of labor and collaboration, allowing professionals to do professional things, can also further improve efficiency.

In fact, openness has gradually become a trend in the field of autonomous driving. Waymo, a subsidiary of Google, has terminated its strategy of building its own cars and instead strengthened cooperation with traditional car companies.

Since its founding, Tesla has always adopted the "single-handed" model of building its own cars, insisting on independently developing complete vehicles and autonomous driving technology. Now, not only is Tesla's electric vehicle production capacity limited, but its autonomous driving technology is also gradually lagging behind other players.

At home, latecomers may also win

In July last year, the State Council issued the "New Generation Artificial Intelligence Development Plan", and artificial intelligence has risen to a national strategy. Up to now, local policies on autonomous driving have been introduced one after another, further promoting the development of the autonomous driving industry.

Baidu Apollo was selected as a national new generation artificial intelligence open innovation platform in November last year. In March this year, it was the first batch of temporary license plates for Beijing's autonomous driving test, which created an effective road test environment for Baidu's driverless cars.

China is the world's largest automobile consumer market, and its environmental pollution and traffic congestion problems are more serious than those in Western countries. Considering the need for industrial upgrading and overtaking on the curve, China's demand for driverless cars is more urgent.

In China, people are more accepting of autonomous driving.

Last year, JD Power's "Chinese Consumers' Autonomous Driving Preference Survey" showed that 10% and 68% of Chinese consumers said they "completely trust" and "may trust" fully autonomous driving technology, and only 4% of consumers "completely distrust" fully autonomous driving technology.

A similar study by JD Power in the United States showed that 35% of Americans "do not trust" fully autonomous driving technology at all, and more than a quarter of Americans believe that autonomous driving technology has no benefits.

China's urban structure can provide richer traffic scenarios for road testing of driverless cars. As China further opens up road testing of autonomous vehicles, companies such as Baidu will obtain a large amount of autonomous driving data, which will create strong conditions for the continuous improvement of autonomous driving.

More road test scenarios will mean faster iteration and more mature technology, which will accelerate the transition of autonomous driving technology from the laboratory to practical use. `

The standards for autonomous driving vary from state to state in the United States, and overall the control over autonomous driving is stricter.

Compared with the United States, China has stronger infrastructure construction capabilities. Once the technology matures, the Chinese government will be able to deploy autonomous vehicles more efficiently.

Collaborative mass production accelerates the implementation of autonomous driving technology

With the continuous upgrading of Apollo versions, the openness of Baidu's autonomous driving technology has gradually deepened.

In this Apollo 3.0 version, the original vehicle reference platform was upgraded to a vehicle certification platform, and the hardware reference platform was upgraded to a hardware open platform. The upgraded hardware development platform also added a bottom-level software abstraction layer, which is used to seamlessly connect the upper-level software and device drivers.

After the Apollo upgrade, Baidu will be able to participate more deeply in the autonomous driving industry chain and provide more convenient services. This means that Baidu will gain more partners and resources. The threshold for various start-ups to join will also be further lowered, and the ecological chain built by Baidu will be more open and dynamic.

Different from last year’s slogan of “All in AI”, Baidu’s slogan this year is “Everyone Can AI”, which further emphasizes the importance of platform openness.

AI is Baidu's layout for the future, and autonomous driving is an important part of Baidu's AI layout. At the same time, Baidu's autonomous driving can get support from its other AI departments, which is also Baidu's advantage in developing autonomous driving.

At this year's Developer Conference, Baidu launched four mass-production solutions: Xiaodu in-vehicle OS and autonomous parking, unmanned work vehicles, and unmanned shuttle buses.

Baidu claims that OEMs and Tier 1 can upgrade traditional cars to smart cars in as little as 30 days through Xiaodu Car OS. Currently, Xiaodu Car OS already has mass production partners such as Ford, Beijing Hyundai, Dongfeng Yueda Kia, Chery, Byton, and Dongfeng Xiaokang.

The Neolithic unmanned vehicle launched at this conference is also a representative work of Baidu's cooperation in the autonomous driving industry chain: Baidu provides autonomous driving software and algorithms, CHJ provides vehicle hardware platform, and Neolithic is responsible for integration, product implementation and operation. The promotion of multi-party cooperation has undoubtedly accelerated the speed of technology implementation.

With the implementation of these mass production plans, the development of Baidu's autonomous driving technology will be further accelerated. As the actual implementation scenarios increase, Baidu will have more opportunities to train algorithm models, and then continuously adjust and optimize algorithms and improve technologies.

Apollo's partners bring Baidu not only various resources, but also provide Baidu with many new ideas. Through numerous partners, Baidu can better understand the market implementation of segmented fields, which will help Baidu find more mass production implementation scenarios.

As a winner of Toutiao's Qingyun Plan and Baijiahao's Bai+ Plan, the 2019 Baidu Digital Author of the Year, the Baijiahao's Most Popular Author in the Technology Field, the 2019 Sogou Technology and Culture Author, and the 2021 Baijiahao Quarterly Influential Creator, he has won many awards, including the 2013 Sohu Best Industry Media Person, the 2015 China New Media Entrepreneurship Competition Beijing Third Place, the 2015 Guangmang Experience Award, the 2015 China New Media Entrepreneurship Competition Finals Third Place, and the 2018 Baidu Dynamic Annual Powerful Celebrity.