English | 中文 | 한국어 | 日本語

A high-performance autonomous driving system

Table of Contents

• Introduction
• Prerequisites
• Quick Start
• Copyright and License
• Connect with Us

---

Introduction

Apollo-Lite provides powerful modules and features for autonomous driving development. Before getting started, please ensure your environment meets the prerequisites and follow the installation instructions below.

For a deeper understanding, refer to the following documents:

• Design Document: [中文 | [English]]
• Integration Document: [中文 | [English]]
• Development Process: [中文 | [English]]
• Tools Document: [中文 | [English]]
• Product Manual: [中文 | [English]]
• Issues / FAQ: [中文 | [English]]
• WEP Proposal: [中文 | [English]]
• Company Introduction: [中文 | [English]]

---

Prerequisites

• Machine: Minimum 8-core CPU, 8GB RAM
• GPU: NVIDIA Turing GPU recommended for acceleration
• Operating System: Ubuntu 20.04 LTS

---

Quick Start

Note: For quick startup and verification, only the CPU-based image (simulation planning module) is provided. The full GPU-dependent tutorial will be released later. We recommend starting with the CPU image because the GPU image is large and has complex dependencies, which may not be suitable for beginners.

1. Install Deployment Tool

pip install whl-deploy

2. Setup Host Environment

Run the following scripts to prepare your host machine. These steps will:

1. Install Docker (checks if already installed)
2. [skip] Install NVIDIA Container Toolkit (checks if already installed, depends on Docker)
3. [skip] Perform host system configurations

whl-deploy setup docker

3. Start Docker Container

Download and start the Apollo container image (only required once):

bash docker/scripts/dev_start.sh -d testing

To enter the running container environment in subsequent sessions:

bash docker/scripts/dev_into.sh

Set environment variables:

source cyber/setup.bash

4. Build Apollo

To build the entire Apollo project:

./apollo.sh build_cpu

To build a specific module:

./apollo.sh build_cpu <module_name>
# Example:
./apollo.sh build_cpu planning

Notes and Troubleshooting

• Out of Memory (OOM) Issues: If the build process is terminated due to insufficient memory, try limiting the number of CPU threads used during the build:

  ./apollo.sh build_cpu dreamview --cpus=2

• Slow Download Speeds: If you experience slow downloads, you can manually download the required archive from the following link: Caiyun Cloud Drive (Extraction code: jfwu). After downloading, place the archive in the .cache/distdir directory within your codebase.

---

Copyright and License

Apollo-Lite is licensed under the Apache License 2.0. Please comply with the license terms when using or contributing to this project.

---

Connect with Us

• ⭐ Star and Fork to support the project!
• 💬 Join our community discussion group to chat with developers.
• 📧 For collaboration or business inquiries, contact: [email protected]

---

Thank you for being part of Apollo-Lite's journey towards autonomous driving innovation!