新興車企戰略分析（2022）：理想汽車

理想汽車2022年上半年銷量60801輛，同比增長99.1%。 從車型來看，麗ONE上半年依然是主力。 隨著2022年9月L8的上市，相信Li ONE將逐漸退出生產線，L8、L7、L9成為未來產銷重心。

本報告重點關注中國新興汽車製造商理想汽車，提供有關公司概況、近期業績趨勢、營銷和服務結構、應用、電氣技術、電氣和電子架構等增長的信息。

內容

新興汽車製造商銷量前 10 名（2021 年）

第 1 章 Li 汽車簡介

• 基本信息
• 開發成果
• 收入、銷量、計劃
• 汽車製造平台和產品規劃
• 財務結果
• 投資結果
• 研發投資和方向
• 生產佈局
• 營銷服務

第2章理想汽車App用戶操作

• 理想汽車應用版本迭代（一）
• 理想汽車應用版本迭代（2）
• 理想汽車應用版本迭代（3）
• 首頁用戶畫像分析
• 功能板塊分析——社區板塊/新聞板塊
• 功能板分析 - 車輛控制板
• 功能板塊分析-理想汽車商城產品類別/特點
• 功能板分析 - 我的板
• 管理策略/用戶增長系統
• 用戶增加系統-積分功能及獲取方式
• 用戶增值系統-積分消費模式
• 用戶增長系統 - 徽章/會員等級

第三章理想汽車電氣技術

• 比較銷售模型之間的動態性能參數
• 電氣技術佈局
• 電力系統特點
• 第一代車輛增程式電氣解決方案
• 第二代汽車增程式電氣解決方案
• Li ONE熱管理系統架構及供應商
• Li ONE 的熱管理系統策略
• Li L9 熱管理系統和驅動系統供應商
• 車輛供能方式
• 超級充電網絡規劃
• 電氣化動態

第4章鋰汽車電子電氣結構

• EEA 演變：LEEA 1.0 到 LEEA 3.0
• LEEA 2.0 硬件架構：AD Max
• LEEA 2.0 硬件架構：自動駕駛算法
• LEEA 2.0 硬件架構：智能座艙
• LEEA 2.0 硬件架構：中央域控制器 (XCU)
• LEEA 2.0 硬件架構：中央計算平台 + 4 個區域控制器
• LEEA 3.0 硬件架構：中央計算機單元 (CCU) (1)
• LEEA 3.0 硬件架構：中央計算機單元 (CCU) (2)
• LEEA 3.0 硬件架構：區域控制器
• LEEA 3.0 通信架構：PCIe 交換機、TSN 交換機
• LEEA 3.0 軟件架構：定義和部署多級服務
• LEEA 3.0 軟件架構：LiOS (Li Auto OS)

第五章理想汽車智能駕駛技術

• 自動駕駛發展路線圖
• 自動駕駛芯片佈局
• 智能駕駛研發模式
• 智能駕駛算法（一）
• 智能駕駛算法（二）
• 智能駕駛數據的積累和處理
• AD系統及典型模型
• 廣告最大系統
• AD Pro 系統
• ADAS 硬件迭代和供應商
• ADAS 硬件 - 在 LiDAR 解決方案中比較蔚來汽車、理想汽車和小鵬汽車
• ADAS 的軟件迭代
• ADAS - NOA 功能
• ADAS - 比較 NOA 解決方案中的蔚來汽車、理想汽車、小鵬汽車和特斯拉
• 自動泊車系統發展路線圖
• 自動泊車的功能演變
• 全場景可視化泊車
• 自動駕駛中的合作動力學

第6章理想汽車的智能座艙與車聯網技術

• OTA更新模式
• OTA 更新分析 - 按頻率
• OTA 更新分析 - 按年份
• OTA 更新分析 - 按類別
• OTA 更新的主要變化 (1)
• OTA 更新的重大變化 (2)
• OTA更新計劃
• 智能座艙配置
• L9、L8 和 L7 的座椅配置
• 李艾
• 李艾互動（一）
• 李艾互動（二）
• 李艾互動（3）
• 李愛的應用實例：L9
• 其他智能座艙供應商
• 智能座艙研發計劃
• 智能語音系統
• 汽車生態學
• 車聯網安全實踐
• IoV 中的合作動力學

Research on Emerging Automaker Strategy: the strategic layout of Li Auto in electric vehicles, cockpits and autonomous driving

Li Auto will shift from the single extended-range route to the "extended-range + high-voltage battery-electric" route of in 2023.

In the first half of 2022, Li Auto sold 60,801 vehicles, up 99.1% year-on-year. In terms of models, Li ONE still played a main role in the first half of the year. With the launch of L8 in September 2022, Li ONE will be gradually withdrawn from the production line, while L8, L7 and L9 will be the focus of production and marketing in the future.

As for product planning, all the models currently being sold by Li Auto are extended-range electric vehicles. However, Li Auto plans to launch at least two high-voltage battery-electric vehicles every year from 2023 onward. For the purpose of high-voltage super-fast charging, Li Auto deploys the following four aspects: First, 4C batteries. Second, application of SiC technology. Third, thermal management system. Fourth, 400KW charging network.

According to its plan, Li Auto will produce the third-generation semiconductor SiC power chip in 2024. At the same current of the high-voltage platform, this chip is 70% smaller than an IGBT chip, with the comprehensive efficiency being improved by 6%. The layout of Li Auto's 800V high-voltage battery-electric technology reveals that one of the selling points of new cars in the future will be reflected in the charging speed.

Li Auto has self-developed AEB and NOA and laid out autonomous driving chips to progress on intelligent driving

As for the progress of intelligent driving, Li Auto has developed AEB system by itself as a "latecomer". In the future, Li Auto will provide all open source codes of its AEB system to improve traffic safety.

In addition, Li Auto added NOA to 2021 Li ONE in December 2021, improved the performance of AEB, and optimize the detection and fusion of cameras and radar. Since 2022, all new cars have been equipped with Li Auto's self-developed hardware compatible with L4 autonomous driving as standard. Li Auto plans to make urban NGP functions available in Li L9 through OTA in 2023, and install L4 autonomous driving capability on production vehicles via OTA around 2024.

Regarding the core underlying technology layout of intelligent driving, Li Auto established Sichuan Lixiang Intelligent Technology Co., Ltd. in May 2022 to design chips. In August 2022, Xie Yan, the former vice president of Huawei Software, joined Li Auto as the head of system R&D division. The system R&D division is mainly responsible for R&D of some underlying intelligent technologies, including Li Auto's self-developed operating system and computing platform. Li Auto's computing platform business also includes its self-developed intelligent driving chip project.

For the intelligent driving algorithm, Li Auto uses BEV framework similar to that of Tesla, that is, it utilizes pure vision for motion perception prediction. On the basis of BEV visual information, Li Auto exploits additional LiDAR and HD map information input to implement the BEV fusion algorithm, and adds a visual security module and a LiDAR security module which are redundant with BEV framework model for the sake of an extra layer of protection.

The cockpit of Li Auto upgrades from 2D interaction to 3D interaction.

The cockpit multi-modal interaction represents development trend of human-machine co-driving era. As per three new cars launched in 2022, Li Auto upgrades the past four-screen 2D interaction in Li ONE to current five-screen 3D interaction, and realizes "voice+gesture" multi-modal interaction.

For example, the five screens of Li L9 include a safe driving interactive screen, a W-HUD with a projected area of 13.35 inches, a 15.7-inch integrated center console screen and co-driver screen, and a 15.7-inch rear entertainment screen. The in-vehicle 3D ToF sensor perceives the cockpit environment in real time. Plus, 6 microphones, 7.3.4 panoramic sound layout, 5G dual-operator automotive communication network, and multi-modal spatial interaction technology developed by Li Auto based on deep learning enable the three-dimensional interaction in the cockpit.

In terms of perception, Li AI, the intelligent cockpit space, imitates the coordination of human ears and eyes to attain the three-dimensional information perception inside the vehicle under the influence of multi-modal attention technology by a distributed hexasilicon microphone, an IR 3D ToF sensor, MIMO-Net six-vocal-range enhancement network and MVS-Net multinocular & multi-view visual fusion network.

As for understanding and expression, Li AI restores the multi-source heterogeneous data sensed by fusion perception to concrete events in the network, and fulfills further abstract understanding. Ultimately, knowledge linking, knowledge completion and logical reasoning form an event graph, allowing machines to have their own understanding and decision-making capabilities.

Regarding voice technology, Li Auto defines its voice assistant "Lixiang Tongxue" as the user's housekeeper (current stage) and family (future goal), and plans a three-stage product upgrade. At present, the goals of the first two stages have been achieved through OTA: The first stage: Li Auto's self-developed "Lixiang Tongxue" engine replaces the underlying capabilities with Horizon + AIspeech + Microsoft, etc.

The second stage: "what you see is what you can say", four-vocal-range locking and other functions.

In the future, the voice system will offer functions such as "from application-on-demand to network-on-demand", cross-screen multi-person dialogue, and "the front passenger can pick up the conversation after the driver finishes speaking".

Table of Contents

Top10 Emerging Automakers by Sales Volume in 2021

1 Profile of Li Auto

• 1.1 Basic Information
• 1.2 Development History
• 1.3 Revenue, Sales Volume and Planning
• 1.4 Carmaking Platform and Product Planning
• 1.5 Financing History
• 1.6 Investment History
• 1.7 R&D Investment and Direction
• 1.8 Production Layout
• 1.9 Marketing and Services
  • 1.9.1 Sales Models
  • 1.9.2 Distribution of Offline Service Outlets
  • 1.9.3 Sales Channels
  • 1.9.4 After-sales Services
  • 1.9.5 Financial Solutions

2 User Operation of Li Auto APP

• 2.1 Version Iteration of Li Auto APP (1)
• 2.1 Version Iteration of Li Auto APP (2)
• 2.1 Version Iteration of Li Auto APP (3)
• 2.2 User Portrait and Homepage Analysis
• 2.3 Functional Plate Analysis - Community Plate and News Plate
• 2.4 Functional Plate Analysis - Vehicle Control Plate
• 2.5 Functional Plate Analysis - Commodity Categories and Characteristics of Li Auto Mall
• 2.6 Functional Plate Analysis - My Plate
• 2.7 Operation Strategy and User Growth System
• 2.8 User Growth System - Points Features and How to Obtain Them
• 2.9 User Growth System - Points Consumption Modes
• 2.10 User Growth System - Badges and Membership Levels

3 Electric Technology of Li Auto

• 3.1 Comparison between Models on Sale in Dynamic Performance Parameters
• 3.2 Electric Technology Layout
• 3.3 Features of Power System
• 3.4 The First-generation Extended-range Electric Solution for Vehicles
• 3.5 The Second-generation Extended-range Electric Solution for Vehicles
• 3.6 Thermal Management System Structure and Suppliers of Li ONE
• 3.7 Thermal Management System Strategy of Li ONE
• 3.8 Thermal Management System and Drive System Suppliers of Li L9
• 3.9 Energy Supplement Modes of Vehicles
• 3.10 Super Charging Network Planning
• 3.11 Electrification Dynamics

4 Electronic and Electrical Architecture of Li Auto

• 4.1 Evolution of EEA: LEEA 1.0-LEEA 3.0
• 4.2 LEEA 2.0 Hardware Architecture: AD Max
• 4.3 LEEA 2.0 Hardware Architecture: Autonomous Driving Algorithm
• 4.4 LEEA 2.0 Hardware Architecture: Intelligent Cockpit
• 4.5 LEEA 2.0 Hardware Architecture: Central Domain Controller (XCU)
• 4.6 LEEA 2.0 Hardware Architecture: Central Computing Platform +4 Zonal Controllers
• 4.7 LEEA 3.0 Hardware Architecture: Central Computer Unit (CCU) (1)
• 4.8 LEEA 3.0 Hardware Architecture: Central Computer Unit (CCU) (2)
• 4.9 LEEA 3.0 Hardware Architecture: Zonal Controller
• 4.10 LEEA 3.0 Communication Architecture: PCIe Switch and TSN Switch
• 4.11 LEEA 3.0 Software Architecture: Definition and Deployment of Multi-level Services
• 4.12 LEEA 3.0 Software Architecture: LiOS (Li Auto OS)

5 Intelligent Driving Technology of Li Auto

• 5.1 Autonomous Driving Development Roadmap
• 5.2 Autonomous Driving Chip Layout
• 5.3 Intelligent Driving R&D Model
• 5.4 Intelligent Driving Algorithm (1)
• 5.4 Intelligent Driving Algorithm (2)
• 5.5 Intelligent Driving Data Accumulation and Processing
• 5.6 AD System and Typical Models
• 5.7 AD MAX System
• 5.8 AD Pro System
• 5.9 Hardware Iteration and Suppliers of ADAS
• 5.10 ADAS Hardware - Comparison between NIO, Li Auto and Xpeng in LiDAR Solutions
• 5.11 Software Iteration of ADAS
• 5.12 ADAS - NOA Functions
• 5.13 ADAS - Comparison between NIO, Li Auto, Xpeng and Tesla in NOA Solutions
• 5.14 Development Roadmap of Autonomous Parking System
• 5.15 Functional Evolution of Automated Parking
• 5.16 Visual Parking at All Scenarios
• 5.17 Dynamics of Cooperation in Autonomous Driving

6 Intelligent Cockpit and IoV Technology of Li Auto

• 6.1 OTA Update Modes
• 6.2 OTA Update Analysis - by Frequency
• 6.3 OTA Update Analysis - by Year
• 6.4 OTA Update Analysis - by Category
• 6.5 Major Changes in OTA Updates (1)
• 6.5 Major Changes in OTA Updates (2)
• 6.6 OTA Update Plan
• 6.7 Intelligent Cockpit Configuration
• 6.8 Seating Configuration of L9, L8 and L7
• 6.9 Li AI
• 6.10 Li AI Interaction (1)
• 6.10 Li AI Interaction (2)
• 6.10 Li AI Interaction (3)
• 6.11 Example of Li AI Application: L9
• 6.12 Other Intelligent Cockpit Suppliers
• 6.13 Intelligent Cockpit R&D Planning
• 6.14 Intelligent Voice System
• 6.15 Automotive Application Ecology
• 6.16 Actual IoV Security
• 6.17 Dynamics of Cooperation in IoV