News on July 18, according to a report by "36Kr Auto", Li Auto has recently formed a new team focusing on "end-to-end autonomous driving". The team size is slightly more than 200 people, and has received the flexibility of other teams in the company. support. This newly established autonomous driving team is part of the Ideal Intelligent Driving Department, which has a total number of approximately 800 people and is mainly divided into two major groups: algorithm research and development and mass production research and development. Among them, the core research and development strength of the "end-to-end" autonomous driving project is concentrated in the algorithm research and development group, led by Jia Peng, the leader in the research and development of ideal intelligent driving technology.
1.
Perception Algorithm Group:
-
Behavioral Intelligence Group:
- Team Leader: Yang Yi
- Cover: end-to-end architecture, model, control Model
-
Cognitive intelligence team:
- Team leader: Zhan Kun
- Covering: cognitive model, cloud model
-
Mass production R&D team:
- Team leader: Wang Jiajia
- Covering: smart driving, smart parking, smart safety
-
Reporting relationship:
- Jia Peng and Wang Jiajia report to Lang Xianpeng (head of Lili Intelligent Driving) fenye1. Previously, Lixiang CEO Li Xiang in 2024 China It was revealed at the Automobile Chongqing Forum that the company plans to launch a supervised autonomous driving system based on end-to-end + VLM (visual language model) trained on 3 million Clips (pre-training model) to test users in the third quarter. .
- And between the end of this year and the beginning of next year, the company will also launch a more advanced supervised autonomous driving system that has been trained with more than 10 million Clips.
- At the Li Auto Intelligent Driving Summer Conference held on July 5, Li Auto also announced the end-to-end + visual language model early bird plan, aiming to make vehicles more intelligent and more humane.
- Li Auto emphasized that the advantages of the end-to-end model are reflected in the two aspects of "efficient transmission" and "efficient calculation". The model adopts an integrated design, and information is efficiently transmitted within the model, which improves the overall performance limit of the system and makes the system's actions and decisions more anthropomorphic.
- At the same time, this integrated model can complete inference in one go on the GPU, reducing end-to-end latency, ensuring timely response to vehicle actions, and bringing users a smoother and more natural driving experience. -->
The above is the detailed content of An elite team of 200 people! Li Auto was revealed to have established an “end-to-end autonomous driving” entity organization. For more information, please follow other related articles on the PHP Chinese website!
Statement:The content of this article is voluntarily contributed by netizens, and the copyright belongs to the original author. This site does not assume corresponding legal responsibility. If you find any content suspected of plagiarism or infringement, please contact admin@php.cn