how is policy shifting?
- self-driving cars get legal status in Shenzhen
- owner/managers liable for accidents
- manufacture/seller liable if defects cause accidents
- simulation platforms and vehicle-to-everything infrastructure encouraged
Fully autonomous vehicles (AVs) can run on certain roads in Shenzhen from 1 Aug 2022. Thanks to new local regulations commercial scale rollout of AVs will likely speed up in the Greater Bay Area’s buzzing innovation hub. Shenzhen is home to BYD—that recently overtook Tesla as the world’s top EV producer by sales.
The new regulations fill in critical policy and legal gaps in the smart vehicle industry at a time when Beijing banks on a strong auto industry to drive growth (see: H1 exports keep economy afloat, for now.) If duplicated to cover wider regions, this Shenzhen legislation, with its detailed guidelines, could set a clearer path towards achieving Beijing’s 2025 goal of scaled production of conditionally self-driving cars.
self-driving cars officially street-legal
On par with the Society of Automotive Engineer (SAE)’sL3, L4, and L5 typology, three categories of self-driving vehicles are specified
‘Conditionally and ‘highly’ AVs are required to come with human control equipment and drivers, requirements that are waived for fully autonomous vehicles. AVs can only operate unmanned in sections or roads designated by public security or transport authorities. They must be able to turn on hazard lights automatically, pull over and be subject to remote human override in unsafe scenarios.
Applying both to commercial and personal vehicles, these stipulations imply that robotaxis and other commercial applications of AVs now have the legal underpinning to scale up in Shenzhen. A minor caveat here is that the vehicles are required to come equipped with an exterior signal light for autonomous driving, which must turn on when autonomous driving mode is engaged.
clarity on accident liability
The regulations clarify rules for liability in accidents. The driver will be liable for accidents and relevant compensation for AVs with a driver. Owner/managers will be responsible for fully AVs without a driver. The regulations add that where accident damage is due to defects of the AV, its driver/owner/user can sue the AV manufacturer or seller after fronting up their fines, compensation, etc.
A road accident involving a ‘conditional’ AV prompted debate on whether manufacturers touting AV functions should be liable if something occurs. The Shenzhen legislation helps by mandating full liability for drivers—unless the accident is caused by defects of the vehicle’s AV features. Insurance coverage will be impacted: the regulations encourage insurers to cover the full chain of autonomous vehicle design, manufacturing, personal and commercial use, plus data and algorithm services.
The regulations encourage simulation platforms to be built to test and verify AV tech. Unlike closed-course or public road testing, simulation is typically performed in cyberspace, using parameters mimicking real-world driving conditions. Simulation can generate data more efficiently than real-world testing, at lower cost or risk. With TAD Sim (Tencent Autonomous Driving Simulation) at the forefront, the simulation industry in the PRC will likely expand with this new policy support.
Also highlighted in the regulations is vehicle road coordination, also known as V2X (vehicle-to-everything). Planning and building of communication, sensory and computing infrastructure for V2X will be supported, along with sharing relevant data and network resources. With continuous improvement of its algorithms, Shenzhen will thus gain an edge in improving AV safety.
The Shenzhen regulations will bring new momentum to commercialising AV tech, and to other PRC jurisdictions and national policies. It also brings the PRC closer to advanced AV policy in countries like the US and Germany. Despite its groundbreaking significance, the Shenzhen regulations leave blanks needing to be filled later, not least around the state’s role in promoting AVs, and criteria for determining fault and enforcing compensation in the case of accident.
who is shifting policy
Gu Dasong 顾大松 | Southeast University Transportation and Development Research Centre executive director
Shenzhen’s regulations fail to specify what ‘human driving modes and equipment’ entail—traditionally this implies steering wheels, accelerators and brake pedals used by human drivers. Removing steering wheels is one of the ultimate goals of autonomous driving, diverging from traditional design. How Shenzhen specifies local standards in relation to this remains to be seen. Its regulations reflects the urgent need for local laws for AVs and for localities with AV industry clusters, so are a welcome step forward. No blanket conclusion for accident liability will hold: different types of vehicle and levels of automation need to be distinguished. Input from industry, management and legal experts is needed to build consensus.
An associate professor with a constitutional and administrative law background, Gu served on urban governance expert committees in Nanjing and Beijing. He holds a doctoral degree in law from Wuhan University, with post-doc work at Southeast University.
Cao Jianfeng 曹建峰 | Tencent Research Institute researcher
The Shenzhen regulations clarify in legal terms the rules for product market access and local standards for AVs, laying legal foundations for commercialising them on a large scale. Companies are most likely to go for more intelligent and connected vehicle (ICV) sales, and commercial operation of driverless taxis and buses in Shenzhen to explore early stage prospects. Highlights include
- support for simulation in relevant technology development and testing
- opening all areas to road testing, application demonstration and commercialisation pilots
- giving the districts authority to manage these activities
- emphasis given to commercial operations
These new policies will guide the future direction of the industry, and bring new opportunities to innovative companies.
A ‘young expert’ of the Internet Society of China (ISC), Cao is a special researcher on AI ethical decision-making at Hunan Normal University, focusing on the intersection of AI, law and ethics. He holds a master’s degree in intellectual property law from Southwest University of Political Science and Law.
in case you missed it…
5-year plans—live updates
5-year plans: 150-odd and counting…