36Kr Exclusive | NIO's Intelligent Driving Adjustment: Ren Shaoqing Directly Leads the Large Model Team to Sprint for End-to-End Delivery

Author | Anqi Li   Qin Li

Editor | Qin Li   Xuan Yang

In the fierce "end-to-end" competition in the intelligent driving industry, NIO plans to accelerate the delivery pace of the "end-to-end" high-level intelligent driving solution.

Exclusive to 36Kr, today, the intelligent driving R & D department of NIO announced a series of organizational structure adjustment plans. The intelligent driving department will establish a new technical committee to be responsible for the comprehensive capacity building. It is understood that this adjustment is mainly to strengthen the capacity of NIO's intelligent driving delivery team to ensure the delivery of the solution.

Ren Shaoqing, the person in charge of NIO's intelligent driving, will further directly manage the team and personally lead the large model department. Previously, this department was led by Peng Chao, the person in charge of perception algorithms, and reported to Ren Shaoqing. After this adjustment, the persons in charge of key areas such as the driving and parking solution and the data closed-loop solution will directly report to Ren Shaoqing.

The intelligent driving platform department of NIO includes the large model department, the deployment architecture and solution department, and the system department. It is reported that NIO expects that after this adjustment, the organizational structure can better match the large model-based intelligent driving technical architecture, thereby further improving the R & D efficiency and delivery speed, and providing the middle platform capabilities for the three brands of NIO's main brand, LeDao, and Firefly.

In response to the above information, NIO responded that: The company makes active adjustments to ensure that advanced functions are better delivered to users, while meeting the business needs of multiple platforms and multiple functions, and responding agilely to the rapid changes in technology and products.

An employee who is an insider told 36Kr that the core of this adjustment is actually to break down departmental barriers and accelerate the end-to-end vehicle delivery speed. "Therefore, it can be seen that the departments with relatively large impacts from the architectural adjustment are some L2 department leaders." 36Kr Auto has learned that Yuan Hongyuan, the person in charge of the spatiotemporal information department, announced his resignation from NIO not long ago.

L2 corresponds to the director level internally. NIO is one of the car manufacturing companies that started to focus on end-to-end intelligent driving earlier in the industry. In June this year, NIO also took the lead in announcing the organizational structure adjustment around the "end-to-end" intelligent driving technology, and established the large model department, the deployment architecture and solution department, and the spatiotemporal information department. The persons in charge are Peng Chao, Qin Hailong, and Yuan Hongyuan, respectively, and all report to Ren Shaoqing of NIO.

NIO's end-to-end solution also started as early as April and May. Peng Chao is mainly responsible for the end-to-end model R & D work, and this department integrates the model departments of the intelligent driving perception and regulation control departments; Qin Hailong is mainly responsible for the mass production and delivery of the vehicle.

However, in the opinion of many NIO intelligent driving employees, this organizational adjustment that started earlier did not bring the expected results. "Obviously, the earliest plan for the end-to-end intelligent driving was to be launched on 930 (September 30), and the pace is obviously slower."

In the opinion of many people, the key reason for the progress not meeting expectations is that "the organization needs to be further adjusted".

The poor coordination between departments is a major factor that troubles employees. For example, the map department led by Yuan Hongyuan is also responsible for a considerable part of the end-to-end data work, and this part of the work needs to be deeply coordinated with the large model department in order to be advanced quickly.

The regulation control department led by Qin Hailong also needs to cooperate and support the data with the large model department. However, in the opinion of some internal employees, these cooperations have not been smooth in the past few months. "It's very simple. After the end-to-end large model technology is mature, it is necessary to replace some of the original work, which is the conflict of interests faced by the entire industry."

This year, during the end-to-end adjustment process, Li Auto also carried out a significant staff reduction, and during Tesla's transformation to end-to-end, the core team was also quickly compressed to about 200 people.

Therefore, this time, NIO's establishment of a large technical committee and a more thorough architectural adjustment is also expected by many employees. Some people told 36Kr directly, "This adjustment is definitely a good thing. It may have a relatively large impact on department leaders, but for the people working below, the coordination will be much smoother."

In their opinion, NIO has a relatively luxurious thousand-person intelligent driving team, four Orin intelligent driving computing hardware, and a considerable investment in resources. If the team efficiency can be further improved, the end-to-end intelligent driving progress will not fall behind the mainstream rhythm of the industry.

A top end-to-end intelligent driving engineer told 36Kr directly that in fact, the transformation of intelligent driving to end-to-end "is to use the end-to-end team of about a hundred people to revolutionize the original intelligent driving team. Only by realizing this can the adjustment and transformation be resolute."

According to 36Kr Auto, after this organizational structure adjustment, NIO's "world model" intelligent driving solution plans to be launched in the first quarter of next year.

In November this year, Li Bin, the CEO of NIO, issued a warning in the all-staff letter on the 10th anniversary of the establishment. The next two or three years will be the most intense and cruel stage in the entire industry, and only a few excellent enterprises can survive. "We have no retreat, and we should not leave a retreat for ourselves." Li Bin said.

NIO has made a long-term investment in technologies such as intelligent driving, with a huge cost. It has also reached a critical moment to use the delivery results to enhance the comprehensive competitiveness of the enterprise.