Autonomous driving in limited scenarios is moving towards commercialization and will enter a new stage of development within three years

Autonomous driving in limited scenarios is moving towards commercialization and will enter a new stage of development within three years

The picture comes from “”

Under the epidemic, the concept of “unmanned” unexpectedly ushered in the first ray of dawn in 2020.

In the days when people are forced to “stay at home”, short videos related to unmanned vehicles helping the epidemic quietly “become popular” on Douyin. On the screen, an unmanned delivery car with the logo of “JD Logistics” is in Qingshan District, Wuhan. Jilin Street “comes and goes freely”, and after passing through the intersection of Jianshe 2nd Road, the supplies are finally delivered to Wuhan No. 9 Hospital.

In addition to, many companies such as Huawei, Gosuncn, Uisee Technology, Zhixingzhi, Neolithic, Purdue Technology also participated in this power-up competition. In Wuhan and even in major cities and hospitals across the country, disinfection, food delivery, logistics, temperature measurement and other services are provided for local residents. This also marks the first time that models such as unmanned disinfection vehicles, unmanned sweeping vehicles, and unmanned delivery vehicles have truly entered the lives of ordinary people.

Many netizens commented: Technology changes life. However, the reason why the unmanned vehicles that have been “fry” for a long time can provide services at key nodes is largely because the above vehicles are all self-driving vehicles in limited scenarios, with lower speeds and simpler surrounding environments.

The so-called restricted scene is a specific area with geographic constraints. At present, the working scenarios of unmanned vehicles providing services to epidemic areas are closed/semi-closed areas such as hospitals and residential areas, and abnormal situations occur less frequently. As for some applications on some public roads, it is also due to the scarcity of vehicles and pedestrians during the epidemic, which is theoretically close to the semi-closed scene of the park. Another fact is that the delivery distance of unmanned vehicles on public roads, such as JD’s logistics trolley, is not too long (hundreds of meters). The superposition of all these factors gives unmanned vehicles a good opportunity to “show their talents”.

On the other hand, the timely application of autonomous vehicles under the epidemic also highlights the rapid development and possibility of this technology being implemented. Today, the autonomous driving technology in limited scenarios such as ports, parks, airports, mining areas, and parking lots is gradually becoming mature and is in the middle stage of pilot operation and commercialization. Its initial pilot operations date back to two years ago.

Looking back at the node in 2020, Yiou Automobile believes that autonomous driving in limited scenarios may enter a new stage of development within three years.

new development opportunities

According to past impressions, most people think that autonomous driving is a proper term for passenger cars: tap on the mobile APP, the unmanned car will come in a moment, and carry passengers to the destination automatically. This is not only the understanding of autonomous vehicles in most people’s minds, but also the future vision of the capital market’s overweight layout of RoboTaxi.

However, when the capital market environment becomes more severe, the bottleneck of RoboTaxi’s own development is gradually emerging – the technology is difficult to achieve, and it is difficult to achieve large-scale implementation. It will take a long time to make a real profit, and its popularity will naturally decrease. According to the observation of Yiou Automobile, in the past year, the financing events and financing amount in the field of RoboTaxi have generally decreased. On the contrary, many companies in the field of autonomous driving with limited scenarios have completed a new round of financing last year, including Huituo, Tage Zhixing, and Xidizhi. Enterprises that focus on mining scene scenarios such as Jiajia and Yikong Zhijia, as well as enterprises that apply mainline technology and Neolithic tools in ports and parks.

The attitude of capital reflects the market prospects of the industry and the commercialization hopes of enterprises. “Autonomous driving needs to be realized in combination with scene applications.” Chen Long, CEO of Huituo, once told Yiou Automobile. Chen Chaozhuo, vice president of Zongmu Technology, also stated in the Beijing Internet of Vehicles and Intelligent Driving Forum in 2017 that the maturity of autonomous driving should be analyzed from the two dimensions of automatic driving classification and geofence. In his view, autonomous driving technology can mature earlier in scenarios with no/few people and low speed. This is the distinguishing feature of autonomous driving in limited scenarios.


Compared with the open road scene, the number of types of vehicles and pedestrians in the limited scene is less, and the dependence on sensing equipment such as lidar is reduced, and the overall cost of the autonomous driving solution is reduced accordingly. At the same time, since the autonomous vehicle travels slowly in a limited scenario, its decision-making pressure is also reduced. If there is an unfortunate problem during driving, the vehicle can also pull over and stop in time to enhance driving safety.

In addition, due to the geographically constrained nature of the limited scene, the range of abnormal situations encountered by autonomous vehicles is relatively controllable, and there are fewer corner cases, which reduces the difficulty of technology research and development. Therefore, compared with open road scenarios, autonomous driving technology will mature faster in limited scenarios, and will be the first to usher in commercialization.

At present, companies such as Mainline Technology, Tage Zhixing, Huituo, Zhixingzhi, Uisee Technology, and Neolithic have all implemented pilot operations in mining areas, ports, airports, parks and other scenarios. Through exchanges with industry insiders, Yiou Automobile believes that the scope of the pilot project of this scenario will be expanded on a large scale in the next three years, and commercial operation will be realized in about five years.

Standing in the future, three changes

As autonomous driving in limited scenarios gradually moves from pilot to commercial operation, relevant companies will also undergo corresponding changes. Yiou Auto learned through exchanges with many industry insiders that autonomous driving in limited scenarios is currently on the eve of commercialization, and the focus of competition among related companies in the future will shift from technology to operation.

First, the cost of parts and components in the autonomous driving industry will be further reduced, and the supply chain will be further mature. At CES in early 2020, autonomous driving lidars can be called “a hundred flowers blooming” and become a major attraction of the exhibition. Companies such as Sagitar Juchuang, Leishen Intelligence, DJI, Hesai Technology have displayed their latest products. At present, there are already a number of companies whose products have dropped the price to around $1,000 while stabilizing their performance. In order to maintain accuracy and safety, autonomous driving in limited scenarios still requires the assistance of lidar, and the low cost and high performance of this essential kit for unmanned vehicles will complement the commercialization of autonomous driving.

In addition to lidar, a suitable wire-controlled chassis is also a basic kit for autonomous driving. With the development of technology, the role of autonomous drive-by-wire chassis as an independent supplier began to become clear around 2018. Under the autonomous driving in limited scenarios, low-speed electric vehicle companies and passenger car companies are trying to enter the market. At the same time, some robot companies are also making corresponding attempts.

Secondly, in the process of evolution to large-scale pilot applications, vehicles with limited-scenario autonomous driving will be closer to the product state – stable enough performance, relatively low cost, strong scene replication capabilities, and good integration with the operating platform. These four points are also considered by the industry to be the key factors for the implementation of autonomous driving. Providers of autonomous driving solutions will focus on technology development and focus more on the operation and service capabilities of enterprises.

“Don’t expect autonomous driving technology to solve all problems. Some problems that cannot be solved by technology can be solved through operation.” Xia Tian, ​​CTO of White Rhino, once told Yiou Automobile. Compared with open road scenarios, there are relatively few corner cases encountered by autonomous vehicles in limited scenarios. During the actual road test and trial operation, the machine can collect problems in a timely manner and learn them in the shortest possible time. The solution, so iteratively, grows in its ability to deal with abnormal situations.

To increase safety, most autonomous driving companies focusing on limited scenarios have equipped their vehicles with remote takeover systems. At the same time that occupations such as takeaways and logistics drivers disappear, new positions such as automatic driving remote receivers, automatic driving operation and maintenance personnel, and automatic driving platform administrators will appear.

At the same time, the competitive situation of autonomous driving companies with limited scenarios will also shift from technical competition to operational competition, in order to meet the needs of users to the greatest extent.

The Links:   NHD-24064WG-ATFH-VZ LM150X08-TLC1