[Big read] Guangzhou remains at the forefront of the nascent autonomous car industry

On a phone app, I typed in my pick-up point on Daguan Road in Guangzhou's Huangpu district, along with my destination. My booking was confirmed in seconds, and three minutes later, a Hongqi self-driving robotaxi slowly pulled up. The robotaxi does not look very different from other taxis; one easily overlooks its roof sensor module in the sea of cars on the road.

According to Guangzhou's regulations on the commercialisation of autonomous vehicle pilot services, while the car is on self-drive, a human driver, or "safety driver" (安全员), needs to sit in the driver's seat to monitor the driving situation throughout the journey. The safety driver also verifies the passengers' details and reminds them to fasten their seatbelts before the vehicle moves off. Since passengers are only permitted in the backseat, the robotaxi can only accommodate a maximum of three people.

On my journey, the safety driver's hands were not on the wheel but always near it. He told Lianhe Zaobao that all safety drivers undergo rigorous training and assessments. They do not need to drive the car and would only intervene in the event of an emergency.

The robotaxi drove steadily and was able to give way to other vehicles and change lanes. But it drove slower than regular vehicles due to speed limit requirements. The six-kilometre journey took 15 minutes to complete. The ride would have normally cost 36 RMB (US$4.90), but I only paid a promotional price of 3.58 RMB as a first-time user of the robotaxi.

I sat in the Apollo Go robotaxi, which operates in designated areas of Guangzhou and is one of the autonomous vehicles operating in the city's autonomous driving demonstration zone. In 2018, Guangzhou introduced plans to push for smart car development. In 2019, it issued the first batch of licences to allow road tests for self-driving cars and later expanded the scope and subject of the road tests.

Currently, Guangzhou has issued on-road testing licences to 348 intelligent connected vehicles (ICVs) from 14 test entities, including Pony.ai, WeRide, Baidu Apollo, GAC Group and XPeng.

In designated self-driving areas, robotaxis have become a travel option for locals. While a safety driver is still present in most cars, driverless robotaxis are on the rise.

Guangzhou permitted the operation of driverless robotaxis on the roads since April this year, signalling a new phase in autonomous vehicle testing. Guangzhou-based Pony.ai was the first to receive a self-driving passenger taxi permit, offering fully driverless robotaxi services to select users in the Nansha district. During the car ride, passengers can communicate with remote specialists in real time to receive remote assistance.

Besides self-driving robotaxis and robobuses that are providing public transportation, self-driving cargo trucks and sanitation vehicles have also been put into operation in Guangzhou.

A government official of Guangzhou's Haizhu district told Lianhe Zaobao that Guangzhou planned to launch six self-driving operation demonstration lines in 2021 to develop an ICV demonstration operation programme. Among them, the line in Haizhu district was launched for public testing in 2022 and offers feeder services from subway stations, tram stations and car parks to workplaces, providing convenient last-mile rides to commuters.

The official pointed out that Haizhu's Pazhou Artificial Intelligence and Digital Economy Pilot Zone is currently planning and exploring more intelligent transportation application scenarios, such as smart junctions and automated valet parking, that would solve parking and retrieval problems.

Autonomous driving is considered "the crown jewel of artificial intelligence", and is a huge industry. Besides automobile and tech companies, China's major cities have accelerated the development of the autonomous driving sector in an effort to establish a local industrial chain and obtain market dominance.

According to incomplete statistics cited in a report by The Beijing News, more than 40 provinces and cities in China, including Beijing, Shanghai, Guangzhou, Shenzhen, Chongqing and Changsha, have been actively introducing relevant regulations and plans to manage autonomous driving. These cities are exploring the possibilities in domains such as operational testing, driverless testing, manned testing and road testing.

Guangzhou is an important base for China's automobile industry and is at the forefront of developing the ICV industry. Besides being one of the first cities in China to begin testing autonomous driving, it is also the first to authorise the testing of 5G remote-operated vehicles, recognise the on-road testing licence of ICVs from other regions and introduce policies for the commercial operation of autonomous driving.

In order to establish secure industrial and supply chains for the automotive industry, Guangzhou is also developing the Cooperative Vehicle-Infrastructure Systems (CVIS) to coordinate communication between vehicles and roadside infrastructure, and an Internet of Vehicles (IoV) application that connects and exchanges vehicle data over the Internet.

According to Guangzhou's development goals, the city looks to achieve an automobile industry above designated size (规模以上汽车制造业) worth 1 trillion RMB by 2025, becoming a globally recognised "smart transportation city".

Acting mayor of Guangzhou Sun Zhiyang, who recently took office in October, has close ties to the automobile industry - he was a long-time employee of China FAW Group, and was in charge of developing Guangdong's automobile industrial chain during his time as the vice-governor of Guangdong. Sun's extensive experience in the automobile industry has led observers to be more convinced of Guangzhou's goal to be a "smart transportation city".

WeRide, an autonomous driving tech company headquartered in Guangzhou, told Lianhe Zaobao that Guangzhou's open policies and well-established and complementary automobile industrial chain are key factors in attracting businesses to set up shop there.

A person in charge at WeRide said that Guangzhou has always adopted open policies to create a relaxed business environment, and boost the development of the ICV industry. The city also looks to implement a CVIS-based "one system, four zones and full interconnectedness" (一带四区、区域互联) model, with the construction of IoV demonstration zones in Huangpu, Huadu, Pazhou and Panyu, and interconnectedness between regions, car companies, platforms, terminals and vehicle models.

This is a great confidence booster for businesses in the autonomous driving industry who wish to set up shop in Guangzhou.

Hao Xinhua, director of the Industry Centre of the Guangzhou Urban Planning & Design Survey Research Institute, told Lianhe Zaobao that Guangzhou's automobile industry has a solid foundation. A total of 3,136,800 cars were produced in 2022 in Guangzhou, topping all the cities in China for several consecutive years.

Also, the output value of Guangzhou's automobile industry accounts for nearly 70% of the total output value for the entire province. The large-scale vehicle production has resulted in the formation of a sizeable industry cluster in Guangzhou and the surrounding regions.

Hao pointed out that Guangzhou's automobile industry has strong innovative capabilities. Of the 196 national- or provincial-level engineering technology centres or laboratories relevant to the automobile industry in the Guangdong province, 54 can be found in Guangzhou itself. Moreover, Guangzhou has attracted a slew of cutting-edge autonomous driving businesses. Coupled with the innovative policies that have always been at the forefront of the country, these factors contribute to the advantageous environment for Guangzhou to develop autonomous driving.

Yet at the same time, the development of the autonomous driving sector in Guangzhou has met with some clear challenges. Hao stated that the automobile industry in Guangzhou is quite traditional, with the development of the ICV and new-energy sectors moving slowly. Just 313,700 new-energy vehicles were produced in 2022 in Guangzhou, ranking eighth nationally. Also, Guangzhou's development of parts such as chips, transducers, smart cockpits and other hardware relevant to smart driving is lagging behind.

Furthermore, the industrialisation of autonomous driving companies in Guangzhou is relatively weak. Hao said that companies in Guangzhou, such as Baidu's Apollo, WeRide and Pony.ai, mainly focus on autonomous driving for robotaxis, with little collaboration with local vehicle manufacturers such as GAC and XPeng. In August this year, Pony.ai, Toyota and GAC jointly established a new company to promote the mass production of robotaxis, primarily for commercial services.

Meanwhile, progress of the Advanced Driver Assistance Systems (ADAS) for private cars in Guangzhou is moving slower than that of autonomous driving for commercial vehicles. Hao notes that automotive companies in Guangzhou now primarily focus on autonomous driving for commercial vehicles, such as robotaxis, and have a weaker presence in the ADAS field. In contrast, Huawei's personal passenger vehicle autonomous driving systems are mainly deployed in the private car sector, making it easier to expand market reach.

With the rapid development of ICVs and the autonomous driving industry, autonomous driving in China has started to shift from the testing phase to the commercially operational phase, gradually changing how people get around and their lifestyles. However, several challenges must be overcome for autonomous driving to truly achieve commercial viability.

According to WeRide, in terms of technology, the stability and reliability of integrated software and hardware solutions for autonomous driving call for collaboration from various participants in the industrial chain. Such collaboration, especially in the early stages of vehicle design, is crucial for achieving mass production, in particular for vehicles with Level 4 autonomy (see table).

The person in charge at WeRide told Lianhe Zaobao that in terms of cost, whether the return on investment for autonomous driving is in line with market demand and expectations depends on various factors, including reducing hardware costs such as LiDAR, labour costs for safety personnel, and increasing algorithm reusability to reduce research and development (R&D) expenses.

Policy is also one of the barriers to the rapid development of autonomous driving, with practical issues including vehicles being unable to enter and operate in various cities, as well as difficulties in determining accident liability. The source at WeRide said that various regions currently have different policies on admission permits and the operational scope of autonomous driving, and overcoming these obstacles will require joint efforts from government agencies and businesses.

Chen Xijin, head of the Industrial Research Center at the Guangzhou Urban Planning & Design Survey Research Institute, told Lianhe Zaobao that due to factors such as policy openness, legal systems and technological maturity, the current phase of autonomous driving for robobuses and robotaxis is still largely centred on R&D, with small efforts towards commercial trials. It will be difficult for autonomous driving companies to achieve economic balance in the short term. However, the market is expected to quickly reach the monetisation stage for the high-level assisted driving field of ADAS.

Chen said that based on current trends in technology and policy, and with the gradual improvement of the legal framework, robobuses and robotaxis are expected to gradually play a bigger part in people's lives after 2025. In contrast, high-level smart driving in the personal passenger vehicle sector is expected to spread faster among the public.

Chen is more optimistic about the development of ADAS companies, noting that while autonomous driving companies continue to innovate in advanced R&D, timely introduction of ADAS-related products in the passenger vehicle sector will help enhance economic benefits and increase their influence.

This article was first published in Lianhe Zaobao as "打造万亿级"智车之城" 广州自动驾驶"加速度"".

Related: Has China surpassed the US in the 5G race? | China and the US, who innovates better? | Can the domestic market save jobs, livelihoods and companies in China?