Autoware Documentation#
Autoware is the world's leading open-source autonomous driving framework. Autoware provides a comprehensive, production-ready software stack designed to accelerate the commercial deployment of autonomous vehicles across diverse platforms and use cases.
Get started#
-
Set up Autoware from source or with Docker.
-
Run Autoware in a planning, rosbag replay, scenario, or digital twin simulation.
-
Follow step-by-step guides, from simulation to your own vehicle.
-
Learn how Autoware is built: its components and interfaces.
-
Find out where to ask questions and report problems.
-
Help improve Autoware and this documentation.
More documentation#
-
The core packages of Autoware.
-
Community packages that extend Autoware Core.
-
Tools to develop and evaluate Autoware.
-
News and information from the Autoware Foundation.
About Autoware#
Vision, mission, and core values
What Autoware aims for, and the values behind it.
-
The long-term goals of the Autoware Foundation.
-
The logos, the colors, and the fonts of Autoware.
Capabilities#
Autoware handles the full autonomous driving stack, from sensing to control.
AI & learning
Support for End-to-End (E2E) driving models, ML-based perception, and data-driven trajectory prediction.
Perception
Multi-Sensor Fusion (Lidar, Camera, Radar), Traffic Light Recognition, and dynamic object tracking.
Localization
High-precision NDT matching combined with GNSS/IMU odometry for robust positioning in mapped environments.
Planning
Behavior Planning for intersections and lane changes, plus real-time Dynamic Obstacle Avoidance.
Control
Precise Trajectory Following and standardized vehicle interfaces for drive-by-wire platforms.
Simulation
Digital Twin Simulation (AWSIM), Rosbag Replay, and Scenario Simulation for validation.
Validated use cases#
Autoware is designed to be platform-agnostic. It powers applications such as these:
- Robo-taxis and urban mobility
- Cargo delivery and logistics
- Robo-bus shuttles
- Privately-owned vehicles (POV)