Quick Start

Quick Start is designed forBeginner Learners. It provides problem descriptions and starter code for four major areas—perception, prediction, decision-making and planning, and control—balancing theory with practice. Learners can quickly understand the autonomous-driving technology stack, the role of each area, and its inputs and outputs, while also running code and completing hands-on tasks.

  • Task descriptions:To help beginners understand the autonomous-driving technology stack and the role and research direction of each module, Quick Start provides a task-description page for every area. Each page explains why the area matters, its related research tasks, and the inputs, outputs, key characteristics, and challenges of each task, helping learners build a thorough understanding of autonomous driving.

  • Get started with code:Quick Start selects representative tasks for each area and provides example models. Using foundational models such as YOLO, LSTM, and MPC for different autonomous-driving tasks, learners can quickly gain practical experience and deepen their understanding through hands-on work. The examples cover the full workflow from environment setup to model execution, while keeping complexity manageable for beginners.

项目式学习 · 交互实践

用一个交叉口，串起五个模块

调整参数、观察车辆通行，完成感知、预测、决策、规划与控制的学习任务。
无需安装环境，从浏览器里的第一次实验开始。

进入交叉口实验 →