laitimes

Dialogue with Gu Junli, CEO of Dazhuo Intelligence: Do not do intelligent driving that burns money, and achieve double million mass production in 2025

author:Leifeng.com
Dialogue with Gu Junli, CEO of Dazhuo Intelligence: Do not do intelligent driving that burns money, and achieve double million mass production in 2025

A year ago, there were still various speculations in the industry about Chery Intelligent Driving and Gu Junli's joining. Soon after, in February 2023, Chery established an autonomous driving company named Dazhuo Intelligence, and Gu Junli officially returned to the industry's attention as the CEO of Dazhuo.

Recently, Dazhuo Intelligence, which has been established for nearly a year, recently held the first Ecological Day to show the results, product planning and development strategies of Chery and Dazhuo to the outside world.

Compared with their peers, Chery and Dazhuo have not made much progress in autonomous driving, so how to catch up with the overall progress of the industry?

The answer given by Chery and Dazhuo is "full-stack controllable", not "full-stack self-developed". Dazhuo is establishing an ecological innovation alliance to work with suppliers to serve Chery and install intelligent driving solutions into more models. Dazhuo's first batch of ecological partners include Bosch, Horizon, Heduo, NavInfo and other manufacturers.

The Return of Gu Junli: The Transition from Scientist to CEO

Gu Junli has worked in Tesla and Xiaopeng Motors. Since leaving Xiaopeng in March 2020, Gu Junli has rarely appeared in public in the industry for nearly two years. After joining Chery, Gu Junli has become one of the most critical figures in Chery's intelligent driving sector.

In addition to being the CEO of Dazhuo, Gu Junli also leads Chery's entire intelligent driving team and assists High-tech Group in planning the intelligent driving sector of the Group.

Different from Tesla and Xpeng in the early days, Chery Automobile has many brands and models, and it is oriented to more than 80 countries and regions around the world, so the challenge of mass production is greater. On the other hand, Gu Junli's role has also changed, from the previous role of a scientist to the CEO of the company, and the focus is naturally different.

"I went through the stage where the technology was not bad for money, and at that time I didn't have any business thinking, that is, I made a budget at the end of the year, and then asked the boss to cut it. ”

Gu Junli said that in Dazhuo, we need to change our thinking and cognition, from technology traction to industry traction, from technology to market demand, to have a business mindset, and turn technology into a priceless product.

Gu Junli said frankly that Chery has a very serious family motto for Dazhuo, Chery is not a money-burning company, and never does money-burning business, for this, Dazhuo must also inherit.

Therefore, Dazhuo adopts a product line mechanism, and the investment of each product line is linked to the business, market, and pricing, and is accounted for by the return in the next three years.

"Frankly speaking, our investment in product development is not as good as that of the new forces and Huawei, and we position ourselves to use one-third of the cost of the new forces to invest in production, research and development, and it is a dual-purpose for both business and use. ”

Gu Junli believes that only in this way can you make money and run to the end.

This also means that the technical route adopted by Dazhuo Intelligent pays attention to cost performance. Gu Junli believes that lidar and high-precision maps are both "crutches" for China's complex scenes, and Dazhuo should gradually remove the "crutches" and take a visual-led technology stack. To achieve this path, we need to flow big data, which is strongly related to the scale of Dazhuo's business.

In fact, this is also the change that the autonomous driving industry is currently experiencing, from technology to commercialization and cost reduction. At present, Dazhuo is raising funds from outside and seeking external capital for gradual development.

Gu Junli said that in the era of capital tightening, Dazhuo's market + technology team combination has an advantage at the financing level. "In terms of the current effect, it is still more than expected, and there are some surprises. ”

Dialogue with Gu Junli, CEO of Dazhuo Intelligence: Do not do intelligent driving that burns money, and achieve double million mass production in 2025

At the beginning of the establishment of Dazhuo, there were many opinions that the development path of Dazhuo would follow the model of the Great Wall and Momo Zhixing, that is, to become independent and gradually go public.

Gu Junli said that at the passenger car level, Dazhuo will not be half-hearted, but will build cost value around Chery's 13 million users around the world, and achieve the intelligent transformation of Chery models similar to Tesla, while at the commercial vehicle level, Dazhuo can go out, which also matches Chery's commercial vehicle strategy and broadens the intelligent network application market for commercial vehicles.

From Tesla, Xiaopeng, and now Chery, Gu Junli admitted frankly that the switch is not easy.

In her view, in the past, many autonomous driving teams would only make one product and serve one model, but now the scale of the service models is very large, and Dazhuo needs to learn product development methodology, project group management methodology, market analysis methodology, and user development operation methodology.

Specific to the company's business strategy, Dazhuo is learning Huawei's IPD process, studying the value of the market, users, and customers of each product line, as well as the value and pricing of Dazhuo.

Secondly, the ability to manage the programme is also important. Gu Junli believes that the industry needs to be coordinated, and it is impossible to form industrialization by burning money to build a product point, and how to face one or twenty mass production projects at the same time is a problem that all autonomous driving companies need to solve.

Finally, ecological management, Dazhuo needs to calculate costs and promote technology complementarity to build an ecosystem for passenger cars and commercial vehicles, which is a great test of the industry's judgment. "For each model, let's say there are 100 people, which will not be able to meet the expansion and stickiness of the business volume. ”

Returning to the intelligent driving industry as the CEO of Dazhuo, Gu Junli has a deeper understanding of business and management, but at the same time, the burden on her shoulders is also heavier. Whether Chery's intelligent driving can catch up, Dazhuo Intelligence is the most critical link.

Full-stack controllable, rather than full-stack self-developed

Dazhuo's first appearance after its establishment was at Chery's 2023 Yaoguang Science and Technology Festival. Chery has straightened out the intelligent division of labor and built Dazhuo Intelligent Driving as Chery's autonomous driving technology brand, which is one of the five technical cornerstones of Chery Automobile's "Yaoguang 2025" strategy alongside Mars Architecture, Kunpeng Power, Lion Intelligent Cabin, and Galaxy Ecology.

Although it entered the game late, in the past year, Dazhuo has come up with some achievements - it took half a year to launch L2 assisted driving products, and has been mass-produced and applied on two models, Xingtu Yaoguang and Tiggo 9. Today, Dazhuo has established R&D bases in Wuhu, Shanghai, Shenzhen, and the United States, with a team of more than 500 people, and more than 60% of them are masters and doctors.

Dazhuo has two product lines, namely Pilot and Drive.

The Pilot product line mainly focuses on passenger cars, from L2 assisted driving, to L2+ based on high-speed NOA, and then to L2++ based on urban NOA. At present, L2 assisted driving, including AEB and LCC, will be mass-produced in 2023, and high-speed NOA products will also be mass-produced in 2024.

It is reported that Dazhuo will be based on LiDAR + BEV + Transformer + OCC multi-sensor fusion technology route, to achieve full-scene intelligent driving on streets, rural roads, highways, as well as cross-layer memory parking, valet parking high-end intelligent driving functions, this product will be completed in 2025 mass production and loading.

Gu Junli believes that in the next three years, the penetration rate of intelligence in passenger cars will expand rapidly. By 2025, Dazhuo aims to achieve the goal of more than 80% of the overall ADAS configuration rate of its main sales models and more than 30% of the L2+ configuration rate.

This goal will be achieved by Dazhuo and its suppliers.

According to Gu Junli, Chery positions Dazhuo as an autonomous driving solution company and an integrated platform, and Dazhuo strives to achieve 300,000 to 500,000 units in the L2 product line this year, and more than 900,000 units based on the supplier's solution.

By 2025, Dazhuo hopes to achieve the goal of "double million", with the domestic and international loading volume of independent mass production solutions reaching one million vehicles, and the loading volume of intelligent driving solutions from cooperative suppliers will exceed 2 million.

The Drive product line includes Robotaxi, commercial vehicles and unmanned vehicles.

Gu Junli said that dual-use is a core development strategy that Dazhuo can reduce costs in technology, "Most of the technologies have become mature, and we probably know which technologies can be used for both passenger and commercial purposes." ”

At present, Dazhuo has launched an unmanned tractor project in Chery's Super Plant 1, covering the scene from the dock to the factory. By 2026, Dazhuo hopes to launch 10,000 smart commercial vehicles. In the field of Robotaxi, Dazhuo has obtained the first autonomous driving test license in Wuhu City and is conducting road tests.

Relying on Chery's years of overseas accumulation, the international market is also a key market for Dazhuo. Dazhuo intends to build itself into an industry platform for autonomous driving to go overseas, and lead the suppliers who join the ecological alliance to go overseas together.

Chery and Dazhuo chose to cooperate with other suppliers to build an ecological alliance mainly due to two considerations.

On the one hand, Dazhuo is able to use the ecological alliance to rapidly develop its own research capabilities.

Gao Jiabing, deputy general manager of Dazhuo Intelligence, gave an example of this move, saying that Dazhuo can achieve mass production within 6 months, which is inseparable from the cooperation with ecological partner Horizon. Horizon provides technical empowerment for Dazhuo, and Dazhuo is responsible for the implementation of algorithms. In this process, Dazhuo not only completed the project, but also built the product development process and system.

Dialogue with Gu Junli, CEO of Dazhuo Intelligence: Do not do intelligent driving that burns money, and achieve double million mass production in 2025

On the other hand, it is also related to Chery's huge size. In 2023, Chery sold a total of 1.881 million vehicles, a year-on-year increase of 52.6%. At present, Chery Group has more than 13 million global automobile users, including 3.35 million overseas users.

Gao Xinhua, President of Chery R&D Institute and Chairman of Dazhuo Intelligence, said that Chery's goal is to spread out the platform, build a full-stack controllable technology platform, and operate on the platform with benign partners to make the cake bigger. On this basis, Chery wants to build Dazhuo Intelligence into a benchmark in the entire intelligent driving field, and achieve the highest end of the industry in terms of technology.

At the same time, Gao Xinhua mentioned that Dazhuo will not easily touch the hardware market, mainly doing software and asset-light operations.

Gu Junli said that she will take the lead in setting up an intelligent product planning committee to translate the requirements of Chery's various business divisions and regions into specific division of labor - which can be self-developed and which need industry partners.

"In the next three years, it will be a process of mutual growth. ”

Leifeng.com Leifeng.com