Autonomous driving technology is an important part of driving innovation and promoting the intelligent and digital upgrading of transportation facilities. Recently, this reporter invited Ayushun, executive vice president of Yinche Technology (Shanghai) Co., Ltd. (hereinafter referred to as Yinche Technology), to discuss the application of automatic driving in commercial scenarios, ask him to tell the industry's first autonomous driving heavy truck mass production and commercial operation model, and the development path of using automatic driving technology to empower trunk logistics.
Full-stack self-developed autonomous driving technology
Trunk logistics as an important part of the market supply chain system, the road scene is relatively standardized than urban traffic, many experts believe that trunk logistics is expected to become the first trial field of automatic driving landing, during this period also emerged a lot of excellent logistics technology enterprises, Yingche Technology is one of them.
Founded in 2018, Wincher Technology is an autonomous driving technology and operating company. Ayushun introduced that Yinche Technology focuses on the trunk logistics scenario, and through adhering to the core strategy of "full-stack self-development + mass production drive + in-depth operation", independently develops full-stack L3 and L4 level autonomous driving technologies, and works closely with the automotive industry to achieve front-loading mass-produced autonomous driving heavy trucks, providing logistics customers with a safer, more efficient and better cost of a new generation of transportation as a service freight network (TaaS).
Ayushun introduced that automatic driving involves a number of software and hardware technologies, in order to make complex software and hardware better integrated, Yinche chose to create an automatic driving system and provide services to customers by means of full-stack self-development from algorithms, software, computing platforms to the chassis interface of wire control.
The Yinche Xuanyuan system has a unique algorithm for trunk logistics heavy truck scenarios: the ultra-long distance perception algorithm can achieve 1000 meters of three-dimensional perception, the error control is below 5%; the adaptive and robust control algorithm can effectively cope with the vehicle fluctuations and the flexible link characteristics of the front and the car hanging, and the vehicle lateral control error is maintained within 7 centimeters under different load and road conditions; the fuel saving algorithm can analyze the driving behavior data and select the optimal route, so that the comprehensive fuel consumption is comparable to that of experienced old drivers. Freight logistics is easier, safer and more environmentally friendly through such autonomous driving technology.
By the end of 2021, the L3-level self-driving truck equipped with the Yinche Xuanyuan system has reached mass production and rolled off the production line. The mass-produced models of Yinche Technology have achieved breakthroughs in heavy-duty truck automatic driving in five aspects, including a fully redundant wire control chassis, L3 and L4 truck specification-level hardware sets, L3-oriented human-computer interaction systems, network security design schemes, and rigorous and perfect test and verification systems. The Eincher Regulus system is the industry's first truck autonomous driving system directly for mass production and the first to be put into commercial operation.
Deeply operate the "escort" logistics industry
According to statistics, the current market size of trunk logistics in the mainland is nearly 4 trillion yuan, and the potential is huge. However, there have long been prominent pain points in the market such as fatigue driving, driver shortage and high fuel consumption, which have brought about the problem of low efficiency of fleet operation and high cost that needs to be solved. Autonomous driving technology can effectively alleviate these problems, improve management efficiency for enterprises, and reduce operating costs.
The market for autonomous driving-related enterprises is hot, but the key to winning the market is a clear commercial landing scenario and long-term customer value. Ayushun told reporters that Yinche Technology is not a simple technology company, but a technical company in the logistics soil, since its inception has put forward the "technology + operation" two-wheel drive business model, and set up an operation team to create an operation network, the first to provide logistics customers with kilometer-per-kilometer payment of standardized intercity heavy-duty truck transportation services, has formed a nationwide, based on commercial contracts, pay-per-kilometer capacity network. Since the end of 2021, the front-loading mass production of L3-level autonomous trucks has been carried out with many industry head cargo owners on multiple lines. The cost advantage has been verified, and the scale of operation has grown rapidly.
At the end of February this year, Wincher Technology received a US$188 million B+ equity financing led by Sequoia China and Legend Capital.
Easier for drivers And businesses are more efficient
Whether it is express delivery, express transportation or LTL logistics, it is now in a state of rapid integration, and the concentration of the industry is getting higher and higher, but the phenomenon of "small scattering" in the capacity market is still prominent.
Ayushun believes that the management of cars is essentially the management of drivers, and a large number of inefficient driver management work is the bottleneck of the expansion of the industry, resulting in limited overall industry efficiency.
The application of automatic driving technology has made the driving that was highly dependent on drivers in the past standardized, starting from the L3 manned automatic driving, the large cargo driver has become a simple and easy safety officer, which greatly improves the management efficiency of the fleet and fundamentally unlocks the bottleneck of efficiency and management in the freight industry.
For long-distance freight at a certain distance, it used to be necessary to work shifts with 2 to 3 truck drivers, but now with L3 automatic driving technology, only 1 truck driver is needed as a safety officer, which greatly improves the overall management ability of the fleet, and can gradually reduce the uncertainty caused by the driver's personalized behavior, realize standardized transportation, and improve transportation safety.
L3-level autonomous driving technology has reduced long-distance trunk dual-driving drivers to single-driver drivers.
Intitut Technology insists on starting from the full life cycle cost (TCO) to provide full-stack asset management and capacity service solutions. By connecting the entire life cycle cost of the vehicle, the fuel and driver costs, which account for 50% of the full life cycle cost, will be standardized. Wincher Technology can calculate the cost of each kilometer of self-driving trucks, in this way, it can not only reduce costs for cargo owners in a more targeted manner, but also provide a stable and economical combination of goods to the fleet, making the transportation service more efficient.
At present, Yinche Technology not only cooperates with China's leading platform logistics and express delivery companies, such as Debon, Shentong, Aneng, etc., but also provides autonomous driving capacity services for nestlé and other cargo owners.
During the "Double Eleven" period in 2021, Yinche Technology and Debon Express opened the first intelligent driving freight line in China to carry out freight business on the trunk line between Shanghai and Jinan. Since October 31 last year, this self-driving "Double Eleven" special line has been running twice a day, 880 kilometers a way, and each trip is completed by a driver independently. According to system data, more than 90% of the transport mileage of each trip is assisted by the Yingche Xuanyuan system, and the average speed of the whole journey can reach 74 kilometers per hour, and the time efficiency is 100%.
During the "Double Eleven" period in 2021, the Intelligent Truck of Yinche Technology is ready to depart at the Debon Express Jinan Distribution Center.
Cloud computing data intelligence "bottom" technology iteration
After the commercial operation is launched, the massive data recharge generated in the real business scenario will become a valuable data asset. Efficient data processing and analysis often relies on cloud computing as a support. Cloud computing can not only reduce hardware costs and improve data analysis capabilities for enterprises, but also effectively reduce the risk of data loss and ensure the flexibility of data use.
Ayushun introduced that Yinche Technology chose to carry out in-depth cooperation with Alibaba Cloud to build an industry-leading autonomous driving data intelligence system. Using Alibaba Cloud Data Lake as the foundation of The Data System, Intour has built a highly elastic computing resource integration and scheduling, which can effectively improve the efficiency of computing resources, greatly shorten the time of labeling, model training, and simulation, and lay a good foundation for large-scale commercial applications.
The large screen of human-computer interaction of the intelligent truck of Yingche Technology displays the perception status of the vehicle.
It is estimated that by the end of 2022, the cumulative mileage of Yinche Technology's autonomous driving operation is expected to exceed 100 million kilometers. The high-quality data mined from the massive amount of real operation data can help Wincher iterate faster on autonomous driving technology, continuously improve customer experience and customer value, and achieve L4 level unmanned driving earlier.
Author 丨 Wu Bowen
Picture 丨Reporter Zhang Xingyu
Editor 丨 Wu Bowen Liao Qian
Audit 丨 Jiao Yang
Producer 丨Sun Yingli