HomeArticle

Driverless vehicles are working vigorously in rural areas.

36氪的朋友们2025-11-24 18:39
Unmanned vehicles reduce costs and increase efficiency, enabling large-scale application in the express delivery industry.

Inexpensive and useful - these are the reasons why Guo Haiying keeps repurchasing unmanned vehicles. Compared with large trucks that need to load 1,500 parcels before departure, an unmanned vehicle can set off once it is filled with 600 parcels, and it can make three or four trips a day. Without the need for a driver, it is not restricted by time and never gets tired. Even in the most remote villages, it can accurately identify the route and make timely deliveries.

At 4:30 p.m. on November 18th, the temperature in Jinghe County, Bortala Mongol Autonomous Prefecture, Xinjiang, had dropped to -7°C. Lan Shuxin, the owner of a Zhongtong delivery point, took out his phone from his thick down jacket, opened the Neolix app, and clicked the departure button. A white unmanned vehicle slowly drove out of the warehouse and embarked on its delivery route for the day.

This vehicle has no cockpit and is only as tall as an adult's shoulder. It looks like a large iron box, relying entirely on two large laser radars in the front, resembling big eyes, to identify objects and routes. It runs to eight stations regularly every day and delivers over 500 parcels. Lan Shuxin roughly calculated that just on fuel and labor costs, he can save 200 yuan a day. If the daily delivery volume can reach 1,000 parcels, the vehicle can pay for itself in a year.

In a sparsely - populated county like Jinghe, unmanned vehicles can be particularly useful. The local area is 24 times the size of Chaoyang District in Beijing, but the population is only 1/24 of that of Chaoyang. Some villages are 60 kilometers away from the delivery point, and often only two or three express parcels are delivered in one trip, which is not cost - effective. Unmanned vehicles fill the "unprofitable zone" where drivers are reluctant to go but the delivery point cannot afford to abandon.

When the vehicle first hit the road, Lan Shuxin specifically followed it on his electric bike. He watched it slowly move forward on the deserted Gobi Desert. When it encountered pedestrians, it would automatically broadcast "Please keep a two - meter distance", and when it met an obstacle, it would automatically stop until the station owner manually unlocked it. "It drives more regularly than I do," Lan Shuxin said. Later, he stopped following it and only checked its operating status on his phone. He plans to buy two more vehicles.

Lan Shuxin's attempt is not an isolated case. In the past two years, the application of unmanned vehicles has been accelerating. Data from the State Post Bureau shows that as of the end of 2024, the number of unmanned delivery vehicles in use has exceeded 6,000. This number far exceeds that of unmanned taxis, another area where autonomous driving technology is applied. The total number of vehicles in operation by leading companies in the unmanned taxi field is 3,000.

The express delivery and logistics industry has the most extensive application of unmanned delivery vehicles. This year, Zhongtong has put nearly 3,000 unmanned vehicles into use; J&T Express has more than 1,000; SF Express has about 1,800; and China Post announced in October that it will purchase 7,000 unmanned vehicles in a concentrated manner in the next three to four years.

Two leading unmanned vehicle companies, Neolix and Jiushi Intelligence, have received multiple rounds of large - scale financing worth hundreds of millions of yuan this year and have entered the unicorn club. Among them, Neolix completed a 1 - billion - yuan Series C+ financing and a Series D financing of over 600 million US dollars. As of the time of writing, the latter is still the largest private placement financing in China's autonomous driving field. Jiushi Intelligence completed two Series B financings this year, totaling 400 million US dollars.

Reduce Losses and Supplement Manpower

Provinces using unmanned vehicles to deliver express parcels are not limited to Xinjiang. In Gansu, you can also see them shuttling through villages and towns.

As early as 2023, Guo Haiying, the person in charge of the J&T Express delivery point in Jiayuguan City, began to pay attention to unmanned vehicles. In early 2024, she bought her first one. The bare vehicle and five - year service fee totaled 150,000 yuan, which was a significant investment. After using it for a month, she resolutely bought a second and a third one. In 2025, she directly placed an order for a fourth one in a live - streaming room. By then, the price of the bare vehicle plus service fee had dropped to less than 100,000 yuan, one - third cheaper than a year ago.

J&T Express unmanned vehicle delivering parcels in Jiuquan, Gansu (Photo provided by the interviewee)

Inexpensive and useful - these are the reasons why Guo Haiying keeps repurchasing. Compared with large trucks that need to load 1,500 parcels before departure, an unmanned vehicle can set off once it is filled with 600 parcels, and it can make three or four trips a day. Without the need for a driver, it is not restricted by time and never gets tired. Even in the most remote villages, it can accurately identify the route and make timely deliveries.

Now, in the delivery point managed by Guo Haiying, all rural parcels are delivered by unmanned vehicles. Every day, four unmanned vehicles depart from the express delivery and logistics park in Jiayuguan City, covering three towns and six stations in the city. In the past, due to the long distance, it took more than five hours to make a round - trip to the farthest village, and the number of parcels was small, which was not cost - effective. Now, when using unmanned vehicles for delivery, even if the vehicle is not fully loaded, it only costs a few yuan in electricity for one trip, far less than the fuel consumption of traditional vehicles.

Guo Haiying calculated that using unmanned vehicles can save 25,000 yuan in fuel costs, 20,000 yuan in insurance and maintenance costs, and 20,000 yuan in labor costs a year. It won't be long before the investment pays off.

There have also been changes on the station side. In the past, villagers' pick - up times were uncertain, and couriers often had to wait all day. It was difficult to recruit suitable rural couriers. Now, there are self - service express cabinets in the villages. When the unmanned vehicle arrives, the station staff comes to work and puts the parcels into the cabinets. It only takes two hours a day, and the owner of the village grocery store can take on this part - time job.

In Dunhuang, the express delivery point faces another problem - it can't recruit enough people. Liu Jun, the person in charge of the J&T Express delivery point in Dunhuang City, once offered a monthly salary of 6,000 yuan, but still no one was willing to deliver express parcels. Most locals are engaged in the tourism industry, which offers high incomes, so they are naturally reluctant to do express delivery. Now, three unmanned vehicles undertake more than 40% of the delivery tasks at this delivery point. Employees only need to load and unload the parcels, which lowers the employment threshold.

The use of unmanned vehicles is not limited to the northwest region. In Shenzhen, Qingdao and other places, unmanned vehicles are gradually becoming an important part of the urban logistics network.

In Qingdao, Shandong Province, the number of vehicles deployed by Neolix has exceeded 1,200, making Qingdao one of the cities with the highest density of unmanned vehicles in the world. In July this year, Didi Delivery and Neolix launched a pilot cooperation in Qingdao. They opened an "unmanned delivery" option on the platform, allowing users to place orders through the Didi app. The system then deploys unmanned vehicles for fixed - point delivery as needed.

This pilot focuses on the niche scenarios where "drivers are reluctant to accept orders": tasks during high - temperature periods, peak hours, at night, and in some remote locations. Unmanned vehicles serve as a supplement to the transportation capacity. For example, a food ingredient supply chain warehouse in Qingdao needs to deliver fresh vegetables to four schools between 11 p.m. and 3 a.m. every night. It is difficult to find night - shift drivers, and the delivery time is hard to guarantee. Now, unmanned vehicles have solved this problem and can depart on a fixed route on time and arrive as scheduled.

According to the "Report on the Operation and Development of Functional Unmanned Vehicles" released by Shenzhen in October, Shenzhen has opened 1,257 routes for unmanned vehicles. Among the 798 functional unmanned vehicles deployed in the city, more than half are unmanned logistics vehicles. Currently, SF Express operates 135 unmanned vehicles, mainly serving the last - mile delivery in closed parks, the last 100 - meter delivery in buildings, the last 3 - kilometer delivery from the delivery point to lower - level stations, and the line transportation from the site to the delivery point. SF Express official data shows that in October just passed, the number of unmanned vehicle delivery orders reached 680,000, a month - on - month increase of 55%.

SF Express unmanned vehicle delivering parcels in Shenzhen (Photo by Zhou Yue)

Behind the diverse application scenarios is the continuous evolution of unmanned vehicle technology.

Zhang Xuchen, the co - founder and COO of Jiushi Intelligence, told the Economic Observer that Jiushi Intelligence almost conducts a complete iteration of the vehicle model every year, including sensor layout, structural details, and the vehicle's overall adaptability, enabling the unmanned vehicle to more stably handle different climates and road conditions in various regions. Since the end of 2023, Jiushi Intelligence has successively deployed vehicles in Fuxin, Liaoning, and other places with a long ice - and - snow - covered period and temperatures as low as - 30°C for regular operations. As of the first half of 2024, these unmanned vehicles had been operating continuously in the extremely cold environment for more than half a year and can work in a temperature range of - 30°C to 60°C right out of the factory.

In addition to hardware, Jiushi Intelligence's software also maintains a high - frequency update rhythm. Zhang Xuchen said that the autonomous driving algorithm is basically updated monthly. For common weather conditions in the north such as sandstorms, fog, and snow, the software can continuously optimize the recognition model through a large amount of real - road data, allowing the vehicle to maintain stable obstacle avoidance in low - visibility and high - reflection - interference situations. Currently, more than 15,000 unmanned vehicles deployed by Jiushi Intelligence across the country are continuously operating on real roads in different climate zones. The data transmitted back will be used for the next round of algorithm training and vehicle model improvement.

Resolve the Right - of - Way Issue for Smooth Operation

Unmanned vehicles are not something new. As early as 2018, Neolix launched an unmanned delivery vehicle, but at that time, the price was high and the policy was unclear, making it difficult for them to operate in practice. Real breakthroughs often come from on - the - ground attempts.

Mao Jiandong, the person in charge of the J&T Express delivery point in Jiuquan City, was one of the first in the northwest region to "take the plunge". In 2023, he spent 220,000 yuan to buy his first unmanned vehicle.

Like human - driven cars, unmanned vehicles also need a license. Mao Jiandong said that he had multiple discussions with the Jiuquan government. He arranged a specific work team to be responsible for the daily scheduling, fault handling, and safety inspections of the unmanned delivery vehicles, and monitored and managed the operation process of the unmanned delivery vehicles in real - time. He also submitted a commitment letter to the government department, guaranteeing that someone would be held accountable in case of an accident. Only then were the unmanned vehicles officially put into operation.

Lan Shuxin in Jinghe County, Xinjiang, also went through a similar process. After purchasing his first unmanned vehicle, he collected road condition data with the manufacturer's team and customized the driving route. At the same time, he held consecutive seminars with local postal, transportation, and public security departments. After nearly a month of observation and testing, the vehicle was finally allowed to hit the road.

As unmanned vehicles have appeared in multiple counties and cities in the northwest, express delivery points in Ningxia, Qinghai, and other places have started to wait and see. In the second half of this year, Huang Jinyin, the person in charge of public affairs in the Gansu - Qinghai - Ningxia region of J&T Express, organized two unmanned vehicle departure ceremonies. Express delivery companies from Qinghai and Ningxia came to observe, and they are all considering whether to use unmanned vehicles. However, whether they can be actually put into use ultimately depends on whether the local government opens up the right - of - way.

As of the first half of 2025, more than a hundred cities across the country have opened up the right - of - way for unmanned delivery vehicles, but the distribution is extremely uneven.

Take Gansu as an example. Among its 14 prefecture - level cities, only a few areas such as Jiuquan, Zhangye, Jiayuguan, and Dunhuang have opened up the right - of - way. These cities that allow unmanned vehicles on the road have some common features: they are sparsely populated, have fewer mountains, and have good traffic conditions. In addition, the local governments have a positive attitude towards unmanned vehicles on the road. For example, the Jiayuguan government issued the "Several Measures (Trial) to Promote the High - Quality Development of the Postal and Express Delivery Industry in Jiayuguan City" last year. For postal and express delivery enterprises that purchase unmanned vehicles, unmanned intelligent equipment, etc., with an investment of more than 1 million yuan, a one - time subsidy of 5% of the investment amount will be given, with a maximum of 200,000 yuan per enterprise.

Policy approval is just the first step. It is the exploration of front - line couriers and station operators that really gets this technology "on the road". To ensure the safety of unmanned vehicles as much as possible, operators have pasted their phone numbers on the vehicles so that pedestrians can contact them if there is a problem. When designing the express delivery routes, they avoid busy streets, commercial areas, and schools... They also try to avoid places where it is difficult to park and avoid peak hours.

"We can't let one vehicle's non - compliance affect the operation of all unmanned vehicles in the city," Guo Haiying said. Prudent and standardized use of unmanned vehicles is an unspoken rule among express delivery bosses.

In a situation where policies vary greatly from place to place, unmanned vehicle companies also hope to promote the formulation of regulations and on - the - ground pilot projects.

On the one hand, Neolix participates in the formulation of national and local technical standards and road - test rules for unmanned delivery vehicles. On the other hand, it has established a local - docking team to cooperate with transportation, market supervision, and other departments to adjust the filing materials and vehicle parameters. For express delivery systems with franchise models such as Zhongtong and YTO, Neolix uses the method of headquarters certification and regional pilot projects, allowing different stations to decide whether to introduce unmanned vehicles on their own.

Zhongtong unmanned vehicle delivering parcels in Xinjiang (Photo provided by the interviewee)

Zhang Xuchen introduced that after unmanned delivery vehicles enter the delivery points, the first people to come into contact with them are couriers, co - distribution center staff, and station employees. They can directly feel the operation difficulty and delivery efficiency.