SAIC high-end pure electric project "sprint": Zhiji automobile accelerates mass production and delivery

In 2021, the Shanxi Guoxin Energy Corporation Limited(600617) auto market broke out, and the emerging brands that have just started in the past two years have also accelerated the pace of mass production.

On December 26, the first batch of 200 beta experience versions of Zhiji L7 were delivered offline. According to the plan, in the next few months, the 200 vehicles will collect real scene data in 9 cities across the country. Zhiji automobile will optimize the functions of vehicles in intelligent cockpit, intelligent driving and so on.

Zhiji automobile describes this process as a "data cooker". After the layer by layer iteration of "data cooker", Zhiji L7 test drive vehicles will successively enter offline stores by the end of February next year. In early March, interested users can start a large-scale real vehicle test drive and make a "big decision".

Zhiji is a high-end intelligent pure electric vehicle brand incubated by Saic Motor Corporation Limited(600104) , just as jikrypton is to Geely, lantu is to Dongfeng and Jihu is to BAIC. Its main operator Zhiji automobile was jointly funded by Saic Motor Corporation Limited(600104) , Alibaba group and Shanghai Zhangjiang Hi-Tech Park Development Co.Ltd(600895) and was established on December 25, 2020. The predecessor of Zhiji Automobile - Saic Motor Corporation Limited(600104) "L" project has been started as early as March 2020. Since then, Zhiji has only spent 22 months from the project establishment to the delivery of the first vehicle.

The accelerated arrival of intelligent electric wave makes the "new generation" of Zhiji automobile go all out.

"When the project was established last year, we predicted that the whole intelligent electric vehicle market would have a considerable and rapid development as Saic Motor Corporation Limited(600104) As the carrier of transformation, Zhiji automobile must quickly speed up product R & D, quickly occupy the market and quickly establish a new high-end brand image. However, at that time, we did not predict that the current market evolution has reached such a degree. " Recently, on the eve of the delivery of Zhiji L7 Bata experience version, Liu Tao, CO CEO of Zhiji automobile, expressed emotion to the 21st Century Business Herald reporter.

China Shanxi Guoxin Energy Corporation Limited(600617) automobile is accelerating its penetration at a speed exceeding market expectations. According to the data, in the first eight months of 2021, the penetration rate of Shanxi Guoxin Energy Corporation Limited(600617) vehicles increased continuously month on month. After a slight decline in September and October, it returned to a high of 17.8% in November, while in 2020, the figure was only 5.8%.

Behind the accelerated penetration of new energy vehicles, the Shanxi Guoxin Energy Corporation Limited(600617) automobile brand has also completed the first stage of "attacking cities and land", and the competition in the high-end market is unprecedented fierce. Weilai, ideal and Xiaopeng, the three new car makers, have ranked among the top ten in the sales list of new energy vehicle manufacturers (the sales volume in the first 11 months of this year). At the same time, new energy vehicle brands incubated by traditional vehicle enterprises such as Jihu, lantu and jikrypton have also started batch delivery. On the list of entering the market in the future, avita Salon and Huawei's deeply enabled AITO are also "eyeing" -- it is very important for any player to seize the current window period.

The delivery of the first batch of models shall be completed within 22 months

The 200 units of Zhiji L7 beta experience version are already relatively mature models, "fully meeting the legal and regulatory requirements for driving on open roads". Liu Tao introduced that the performance of the new car chassis is close to a relatively complete state, but there is still a small part in the iteration in terms of intelligent cabin and intelligent driving.

In addition to the offline of Zhiji L7 beta experience version, Zhiji has also completed the planning of the whole product layout, including 2 product lines and 6 models: t (Tech) line and a (Art) line. Among them, there are 4 products of T line, including L7 and LS7 that have been released, and the other 2 have been frozen; Line a is a model developed based on the new thinking of the next development of intelligent electric vehicles. The first car is the mass production concept car AIRO released at the Guangzhou auto show, and the second car is also under pre research.

In the past 22 months, Zhiji automobile has completed the breakthrough of "from 0 to 1". This is not a simple task. Liu Tao said that "multi innovation" is a huge challenge for this project. It is a synthetic development of "platform + three electricity + software + power" - three electricity system, new domain integrated central computing digital architecture, SOA atomic software architecture, intelligent driving system, car modeling, interior lifting screen, etc. are all used for the first time, It is by no means a simple model renewal.

In conclusion, the core difficulty in the development of Zhiji L7 lies in four new product architectures: IO origin high-end pure electric architecture, providing four-wheel steering and four-wheel drive; The new generation domain integrates the central computing digital architecture, supports the vehicle in-depth OTA and refreshes the cockpit experience; SOA Software Architecture breaks the boundaries between functions from the bottom; Im ad intelligent driving system, focusing on the future intelligent driving experience.

200 experience models are offline, and Zhiji automobile has preliminarily completed the development and integration of these technologies. According to Liu Tao, Zhiji was able to do this because he learned the logic and concept of "agile development" from the software industry and carried out innovative practice in the field of vehicle intelligence.

The essence of "agile development" is that "design, development, integration and iteration" can significantly reduce the development cycle. In addition, Zhiji has also cooperated with companies in zebra, zero beam, Yanfeng and other industrial chains to overcome difficulties one by one in the form of multiple rounds of "battle to overcome difficulties" and improve development quality and efficiency.

It is said that in the past month, Zhiji L7 software has iterated a total of 88 versions, and the iteration of Zhiji function software is completed every three days. The user contacts the "Im Zhiji" app to realize the normalized update frequency of 2 weeks and 1 iteration.

Zhiji relies on SAIC and Ali, which is also an important reason why the first model can be quickly developed and offline. Liu Tao frankly said that SAIC has invested heavily in many technologies in the past few years. For example, the IO origin architecture used by Zhiji L7 is the embryo of the chassis created by SAIC's "project X" project around 2015. In terms of vehicle architecture, SAIC began to develop in 2016, and the intelligent driving scheme, including the technical scheme of the chip used, was also developed very early.

Ali has also helped Zhiji in intelligence. Alibaba has obvious advantages in data-driven computing power improvement and intelligent scene iteration. "Alibaba's cloud computing and cloud supercomputing rank first in the world. It has more than 1.5 million servers and 2800 nodes. In addition, Alibaba cloud ranks very high in the world in terms of large-scale data concurrency, cloud computing, communication, security, network and low latency." Liu Tao said, "these are the core starting points for the bottom technical cooperation between Zhiji and Ali."

Data should be further accumulated before formal mass production

"Fully comply with the legal and regulatory requirements for driving on open roads", which means that Zhiji L7 beta experience version can be delivered in mass production. However, Zhiji did not start mass production at this time node, but let the first 200 vehicles off the line run on the ground for the last round of data collection and function iteration before mass production.

According to Zhiji, after these models are delivered to system test users, a "cruise" covering 9 cities across the country will be started to start the process of "Data-Driven whole journey optimization" in the real environment. The tour will take Shanghai as its first stop, covering popular routes in urban business circles such as the Bund and Shanghai Lujiazui Finance & Trade Zone Development Co.Ltd(600663) .

What functions will Zhiji L7 optimize? It is reported that through the data verification of real use scenarios in 9 metropolises, Zhiji will fine tune the intelligent performance of user contacts and product software systems; At the same time, the perception system, domain controller software and human-computer interaction of the product are optimally calibrated for the high load use scenario of mass production vehicles.

In fact, in the development process over the past year, Zhiji has completed the measurement of intelligent driving function in five cities: Shanghai, Guangzhou, Chengdu, Hangzhou and Suzhou, but these tests are not enough for Zhiji's car to pursue the "most human like" intelligent driving experience.

"Zhiji's intelligent driving team in Shanghai has carried out algorithm iteration and data training day and night. We fought in downtown Shanghai, expressway in Chengdu, night peak in Suzhou and Hangzhou and night intelligent driving in Guangzhou." Liu Tao explained that next, Zhiji will collect the actual road conditions in more cities. On the one hand, it will measure whether the algorithm in the laboratory is accurate. On the other hand, it will also provide the most real data to the data factory and data engine, so that Zhiji's intelligent driving algorithm will become better and better and more close to the feeling of human driving.

In fact, Zhiji emphasized the concept of "user co creation" at the beginning of its establishment. Specifically, the data generated by users in the process of using cars will directly promote the evolution and iteration of products, services and brands through automatic in-depth learning of artificial intelligence - at that time, it was also questioned that for a new brand just established, How does the first product accumulate data without users? Now, Zhiji has given its problem-solving ideas.

In the view of Zhiji's team, this method also innovates the traditional automobile development process - on the basis of strict quality verification processes such as DV (design verification) and PV (product verification), Zhiji pioneered a new link of DFV (data feed validation of real scene) to continuously verify and polish the user experience, "This new paradigm redefines the new standard of smart car development process. It will use the real scene data as the 'golden key' to rigorously verify the intelligent performance of deep smart cars and bring higher quality user delivery experience. This is the core difference between smart cars and traditional cars in the development process." Zhiji automobile said.

(21st Century Business Herald)

 

- Advertisment -