Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/drift-tech/drift-releases/llms.txt

Use this file to discover all available pages before exploring further.

Drift CLI makes robot creation intuitive by understanding your natural language descriptions and automatically generating the necessary ROS2 packages, URDF/SDF files, and configurations.

​
Robot creation workflow

Follow these steps to build a robot from scratch using Drift:
1

Start Drift CLI

Launch Drift in your terminal:
drift
2

Create a workspace (if needed)

If you don’t have a ROS2 workspace yet, create one:
drift> create a ROS workspace
Drift automatically sets up the proper ROS2 workspace structure with src, build, install, and log directories.
3

Describe your robot

Tell Drift what kind of robot you want to build using natural language:
drift> create a three arm manipulator with camera sensor
Drift will:
  • Create a new ROS2 package for your robot
  • Generate the URDF/SDF model files
  • Add the specified sensors and components
  • Set up proper joint configurations
4

Launch and test

Once created, launch your robot in simulation:
drift> launch my robot with the custom world file
Drift handles:
  • Launch file generation
  • Gazebo initialization
  • Process management

​
Adding sensors and components

You can modify your robot at any time by adding new sensors or components.

​
Adding a lidar sensor

drift> add a lidar sensor to my robot and rebuild
Drift will:
  1. Update your robot’s model file with the lidar configuration
  2. Rebuild the ROS2 workspace
  3. Validate the sensor integration
  4. Set up appropriate ROS2 topics for sensor data

​
Adding a camera

For robots that need vision capabilities: 
drift> create a three arm manipulator with camera sensor
The camera sensor includes:
  • Image topic publishers
  • Camera info publishers
  • Proper optical frame setup
  • Configurable resolution and FOV

​
Adding other components

You can add various components using natural language: 
drift> add a depth camera to the front of my robot
drift> add a gripper to the end effector
drift> add IMU sensor to the base
Drift understands the context of your robot and places components in appropriate locations based on your description.

​
Configuring robot properties

​
Mechanical properties

Modify physical characteristics of your robot: 
drift> change the wheel radius to 0.1 meters
drift> set the arm length to 0.5 meters
drift> increase the robot mass to 10 kg

​
Joint configuration

Configure joints for your robot: 
drift> set joint limits for the shoulder joint
drift> add a revolute joint between base and arm
drift> configure the gripper with prismatic joints

​
Sensor parameters

Adjust sensor settings: 
drift> set lidar range to 30 meters
drift> change camera resolution to 1920x1080
drift> set IMU update rate to 100 Hz

​
Real-world examples

# Create the base mobile robot
drift> create a mobile robot with differential drive

# Add a manipulator arm
drift> add a 6-DOF arm to the mobile base

# Add sensors for navigation and manipulation
drift> add a lidar sensor for navigation
drift> add a camera to the end effector

# Launch the complete system
drift> launch my robot in the warehouse world

# Create a quadcopter
drift> create a quadcopter drone

# Add sensor suite
drift> add a depth camera facing forward
drift> add IMU and GPS sensors
drift> add a downward-facing camera

# Test the configuration
drift> launch the drone in the outdoor environment

# Start with basic robot
drift> create a simple mobile robot

# Test it
drift> launch my robot

# Iterate based on testing
drift> add obstacle detection sensors
drift> increase the wheel size
drift> add a front bumper sensor

# Rebuild and relaunch
drift> rebuild the workspace
drift> launch my robot with the new configuration

​
Intelligent features

Drift’s AI-powered workflow includes:

​
Automatic package generation

  • Creates properly structured ROS2 packages
  • Generates package.xml and CMakeLists.txt
  • Sets up dependencies automatically
  • Follows ROS2 conventions

​
Launch file generation

  • Creates validated launch files
  • Includes proper parameter passing
  • Handles multi-robot setups
  • Configures Gazebo world loading

​
Adaptive planning

  • Real-time error recovery during builds
  • Intelligent problem-solving for common issues
  • Suggestions for configuration improvements
  • Automatic dependency resolution

​
Smart debugging

When something goes wrong, ask Drift: 
drift> why isn't my camera publishing images?
drift> check if all sensors are working
drift> analyze build errors
Drift provides:
  • AI-assisted troubleshooting
  • Root cause analysis
  • Step-by-step solutions
  • Configuration validation
Always test your robot in simulation before deploying to real hardware. Drift helps you iterate quickly in a safe virtual environment.

Build docs developers (and LLMs) love

Get started for freeTalk to us