The recent market entrance of electric vehicles (EVs) from leading US and China-based firms, such as Tesla and BYD, have challenged the dominance of traditional gas-powered cars and ushered in the future of on-car intelligence features that control throttle, brakes, and steering. However, these vehicles still require human intervention to operate safely and cannot anticipate or avoid potential accidents due to the limits of their computing platforms and smart road technologies. Leading Chinese technology firms Baidu and Didi Chuxing are looking to overcome these technological hurdles and create true autonomous vehicles which rely on high-powered sensors, AI accelerators, and smart road infrastructure to alert motorists of traffic accidents and provide real-time road information within milliseconds.
Supported by the extensive financial resources of Beijing and substantial private investment, these firms have undertaken commercial joint ventures with local telecom providers and foreign AI companies to gain access to state-of-the-art 5G technologies and semiconductor chips. While the global mobility market remains focused on selling personally owned vehicles to consumers, China’s leading AV firms are committing to these technologies with the certainty that autonomous driving will replace traditional vehicles over the next few decades. As Beijing comes upon the half-way mark of its ten-year state initiative “Made in China 2025,” the ruling party is in a unique position to reduce its dependence on foreign technology and compete directly with Washington for the future of the world economy. AVs are one such battleground.
Strategic Reform and Investment
The success of modern economic reform within the Middle Kingdom can be traced to its state-led industrial policies. In particular, Beijing’s latest economic policy “Made in China 2025” has rapidly updated the country’s manufacturing base through direct government subsidies and state-owned enterprise (SOE) investment. As outlined within the European Union Chamber of Commerce’s report “China Manufacturing 2025,” the central government established over 300 active state-investment funds in 2015 to support the priorities outlined in CM2025. These investment funds, which have been reported to hold over US$200 billion of assets, have slowly injected capital into the autonomous vehicle industry with an average SOE investment of US$4 billion from 2016 to 2019. This capital inflow has brought vital resources to the industry which has modernized the production lines, tools, and machinery used to produce vehicle hardware and software and expected to generate $1.1 trillion from mobility services and $0.9 trillion from autonomous vehicles sales by 2040.
Autonomous Driving Industry Investment Records in China
Investment and strategic reform within the AV industry has also taken place through foreign and private sector sources. In Q2 2020, Japanese conglomerate Softbank announced a US$500 million investment in leading autonomous share-riding service, Didi Autonomous Driving. This foreign direct investment represents the single largest fundraising round in the history of China’s autonomous driving industry and could encourage further foreign investment, which Beijing seeks to bring to its economic shores. From a domestic perspective, China’s tech giants continue to fight for the future of the MaaS market through similar audacious commitments of financial capital designed to speed up their research and development programs. In 2017, artificial intelligence company Baidu launched its US$1.5 billion autonomous driving fund, Apollo, to invest in 100 different autonomous driving projects over the next three years. Tencent has followed suit in Q2 2020 to pledge over US$70 billion in areas from autonomous vehicles and cloud computing to artificial intelligence.
MaaS Technology Stack
These SOE and private investment sources of capital have provided Chinese firms with the financing required to build the domestic AV market. However, the direction and growth of this market will depend on the research and development of emerging technologies and whether full-scale adoption of self-driving vehicles will occur within China. Recent activity by leading firms Baidu and Didi Chuxing suggest AV firms are focusing their short-term efforts on ride-sharing services to drive profitability and will gradually phase in autonomous driving technology as an alternative to traditional automotive travel.
An analysis of the future Mobility-as-a-Service (MaaS) market from leading US-based management consulting firm Mckinsey & Company suggests the opportunity to capitalize on the market for autonomous vehicles will largely depend on Chinese firms’ ability to identify and secure the development of the strategic elements of the AV ecosystem and technology stack. For autonomous vehicles, this technology stack includes the following layers: (A) Sensors, (B) Computing platforms, (C) Mathematical algorithms, (D) AV system integration, (E) Whole-car integration, and (F) Mapping and location-based services.
Given China currently lacks the technology to build the most advanced semiconductor chips and software used within CPU, GPU, and AI accelerator computing platforms, most domestic firms have outsourced these technological capabilities as well as invited foreign collaboration over hardware components and AV system integration. In particular, Baidu’s partnership with Nvidia to use the US firm’s Drive AGX platform for their AV computing power highlights the use of foreign technology to fully develop domestic AV services. While foreign partnerships allow Beijing to import needed technological expertise and in some cases, result in forced technology transfer as the price for entering the Chinese markets, certain technology stack layers, including location based services (LBS) and data-cloud solutions are being produced within the country due to security concerns.
The Race to Autonomy
Understanding the sources of public and private investment as well as China’s industry focus within the AV technology stack provides a clearer picture into the progress of the country’s race for autonomous transportation. While many firms have entered the AV space to position themselves for the coming decades, two companies in particular are leading stars within the industry: Baidu and Didi Autonomous Driving. We examine each firm below in further detail and ultimately consider the likelihood of full-scale adoption within China.
While the artificial intelligence giant is most well known for its presence in the Chinese search engine market, Baidu has invested heavily in autonomous technology over the last half decade. The establishment of Baidu’s Apollo Fund in 2017 for US$1.5 billion signaled the firm’s intentions to directly rival foreign competitors Tesla and Google. Over the last few years, this investment has borne fruit in the development of Baidu PaddlePaddle, an open source deep-learning platform which has attracted research talent and shaped the growth of the autonomous technology industry within China. Building off its success, Baidu has released its autonomous vehicle platform, Apollo, to offer a variety of consumer solutions, including robotaxi services, food and supplies delivery, and municipal street cleaning. These services, although limited during the COVID pandemic, have been a vital resource to consumers and the government which has sought to limit the spread of the virus through unmanned technology.
Beijing’s desire to utilize emerging technologies to rebuild domestic infrastructure has also provided Baidu a leading voice in the development of smart cities. In 2020, Baidu won bids in Chongqing’s Yongchuan district, Shanxi province’s Yangquan city, and Anhui province’s Hefei city to build cellular-vehicle-to-everything (CV2X) pilot zones. These zones serve as urban testing grounds where Baidu will work with national telecom carriers to build 5G stations and smart traffic systems that relay information directly to their autonomous vehicle fleet. This new technology will lower driving data latency, or the time it takes for important traffic related data to reach an unmanned vehicle, and provide a safer driving experience for users. Given the 388 county-level cities, 293 prefectural-level cities, 4 municipalities, and 2 special administrative regions (SARs) within China, Baidu Apollo recently released the ACE Transportation Engine, a full-stack technology solution, to help cities build smart infrastructure. The ACE Transportation Engine software subsystem provides cities with a platform to incorporate AI into the equipment, services, and infrastructure being built in any given location. The release of this technology allows Baidu to further Beijing’s vision of a technologically advanced China and provides the company access to important consumer data for its Apollo program. Baidu has placed itself firmly in the driving seat in the race for transforming the traditional driving experience and potential profits which trail this new technology within the competitive AV industry.
Didi Autonomous Driving
Although the company’s CEO has said that it will take another decade of continued investment for AVs to meet regulatory milestones, Didi, the Beijing-based rideshare giant, is also well on its way to becoming a domestic leader in China’s autonomous vehicle industry. In 2016, Didi announced its autonomous vehicle division, and in 2019, the division broke off to form its own independent company. The firm is dedicated to producing Level 4 autonomous driving technologies, which implies high levels of driving automation but still allows for human overriding of vehicle functions. The firm has leveraged its high-profile investments, vast pools of traffic and user data, and key partnerships to differentiate itself.
Didi’s US$500 million fundraise in 2020, led by Softbank, was the largest single investment into autonomous vehicle technology in China to date. For data, Didi has the unique advantage of being able to leverage its ridesharing platform’s big data system, Didi Brain. Didi Brain processes over 100+ TB of new route data and nearly 4900 TB of traffic, user behavior, and driving behavior data every day. Access to pools of data is highly critical for training AV artificial intelligence algorithms and improving vehicle decision. For partnerships, Didi recently linked up with BAIC, the Chinese state-owned automotive manufacturer, to jointly-develop autonomous vehicle technology. While Didi excels at AV research and development, it will greatly benefit from BAIC’s know-how in auto manufacturing.
To say that Didi is well-equipped with resources in the race for autonomous vehicle market share would be an understatement, and the firm has already leveraged its resources to push its vehicles out for testing. The firm became one of the first in China to reach this level of automation, and now Didi has acquired public road testing licenses in Beijing, Shanghai, Suzhou, and California. Earlier this year, Didi began its first public testing trials in Shanghai via its rideshare service, offering robotaxis in open-traffic areas in Jiading District. For now, safety drivers remain in the front seat to oversee vehicle decisions, but, come 2030, these cars could very well be driving on their own.
Beijing’s directive to transform China’s infrastructure using emerging technologies like artificial intelligence has been promptly answered by domestic tech giants Baidu and Didi. Given the competition with foreign firms in America and Europe, these companies have invested heavily in the local autonomous vehicle market with the knowledge that full-scale adoption of these technologies within China could swiftly transform the global automotive industry. These firms and others will continue to build their capital base and expand their technology offerings in the hope that they can provide affordable and safe autonomous driving experiences.
While China may not be as technologically adept at producing certain high precision components of AV technology, including the semiconductors used within its computing platforms, the country can rely on its advantage in artificial intelligence software and foreign partnerships to effectively compete with its most adept rival in the industry, the United States. The current progress made within the Middle Kingdom’s private sector suggests that, while Beijing will not only reach but surpass its goals set within “Made in China 2025,” moving forward, drivers within China and around the world can expect significant changes as technology transforms cities and vehicles into autonomous machinery.