Run Edge Impulse vision on a JetBot with Rubik Pi

This tutorial runs an Edge Impulse object detection model on a JetBot-class robot with a Qualcomm Rubik Pi, Ubuntu 24.04, and ROS 2 Jazzy. It combines the robot-specific jetbot_ros package with edgeimpulse_ros. The JetBot launch file starts a v4l2_camera driver that publishes camera frames. edgeimpulse_ros subscribes to them and publishes detections. jetbot_ros handles motor control, teleoperation, and navigation. Because perception is a separate node, you can reuse it on other robots.

JetBot Gazebo simulation with a camera stream — JetBot Gazebo development environment and simulated camera stream

What you’ll build

Object detection on the robot’s camera feed that publishes results to ROS 2, alongside motor control that a behavior node can drive.

Prerequisites

A Rubik Pi running Ubuntu 24.04 and ROS 2 Jazzy
A SparkFun JetBot AI Kit v2.0 or another JetBot-class chassis with a SparkFun or Waveshare motor controller
A USB camera connected to the Rubik Pi
A Linux .eim object detection model exported from Edge Impulse Studio
Local copies of ei_jetbot_ros_rubikpi and edgeimpulse-ros

Read Run Edge Impulse object detection with ROS 2 first. It covers the model export, runtime, and topics that this tutorial uses.

1. Provision the Rubik Pi

On the Rubik Pi, run the repository’s provisioner:

cd ~/jetbot_ros_rubikpi
sudo bash scripts/rubikpi_provision.sh

The script installs the ROS 2 Jazzy base packages, V4L2 camera support, vision_msgs, diagnostic_msgs, OpenCV, NumPy, PortAudio (for the SDK’s pyaudio dependency), and the JetBot hardware dependencies. If you use the SparkFun motor controller, check that the Qwiic driver is installed:

python3 -c "import qwiic; print(qwiic.__file__)"

2. Build a shared ROS 2 workspace

The JetBot setup script links both packages into one workspace:

cd ~/jetbot_ros_rubikpi
bash scripts/setup_edge_impulse_workspace.sh ~/jetbot_edgeimpulse_ws ~/jetbot_ros_rubikpi ~/edgeimpulse-ros

cd ~/jetbot_edgeimpulse_ws
source /opt/ros/jazzy/setup.bash
rosdep install --from-paths src --ignore-src -r -y
colcon build --symlink-install
source install/setup.bash

Install the Edge Impulse Linux SDK and the JetBot runtime requirements:

sudo apt install -y ros-jazzy-vision-msgs ros-jazzy-diagnostic-msgs \
                    ros-jazzy-v4l2-camera python3-opencv python3-numpy portaudio19-dev
sudo python3 -m pip install --break-system-packages \
  -r ~/jetbot_ros_rubikpi/requirements-edge-impulse.txt

These commands install the Edge Impulse runtime into the Ubuntu 24.04 system Python with --break-system-packages. Use this setup only on a dedicated robot image or a disposable development machine.

3. Start motor control and object detection

The integrated launch file starts the motor node, a v4l2_camera driver, and the edgeimpulse_ros detector. Give it the model path and the camera device:

cd ~/jetbot_edgeimpulse_ws
source /opt/ros/jazzy/setup.bash
source install/setup.bash

ros2 launch jetbot_ros jetbot_edge_impulse.launch.py \
  motor_controller:=motors_sparkfun \
  model_path:=/home/$USER/models/jetbot-detector.eim \
  video_device:=/dev/video0

For a Waveshare motor controller, install its driver package, then launch with motors_waveshare:

pip install Adafruit-MotorHAT

ros2 launch jetbot_ros jetbot_edge_impulse.launch.py \
  motor_controller:=motors_waveshare \
  model_path:=/home/$USER/models/jetbot-detector.eim \
  video_device:=/dev/video0

The launch file starts a v4l2_camera driver on video_device. Don’t start a second camera driver on the same /dev/video* device. If MJPG gives an empty encoding or a cv_bridge exception, use YUYV.

4. Inspect inference output

View the detections and diagnostics:

source ~/jetbot_edgeimpulse_ws/install/setup.bash
ros2 topic echo /edgeimpulse_detector/detections
ros2 run rqt_robot_monitor rqt_robot_monitor

Topic	Purpose
`/edgeimpulse_detector/detections`	`vision_msgs/Detection2DArray` messages with the accepted bounding boxes.
`/edgeimpulse_detector/debug_image`	Annotated `sensor_msgs/Image` when `publish_debug_image:=true`.
`/diagnostics`	`diagnostic_msgs/DiagnosticArray` with FPS, latency, and model health.

Don’t let detections drive the robot until they’re stable. When you add a behavior node, base each action on the confidence, label, and robot state rather than on every frame.

Motor and camera checks

Before you connect detections to motion, test the motors on their own. Start hardware control:

source ~/jetbot_edgeimpulse_ws/install/setup.bash
ros2 launch jetbot_ros jetbot_cpu.launch.py motor_controller:=motors_sparkfun

Then send a short forward command and a stop:

ros2 topic pub -1 /cmd_vel geometry_msgs/msg/Twist "{linear: {x: 0.05}, angular: {z: 0.0}}"
ros2 topic pub -1 /cmd_vel geometry_msgs/msg/Twist "{linear: {x: 0.0}, angular: {z: 0.0}}"

Simulation notes

The repository includes Gazebo Classic worlds for navigation and data collection. Run them on an Ubuntu or WSL development machine with Gazebo Classic installed:

ros2 launch jetbot_ros gazebo_world.launch.py

You can point the detector’s image_topic at the Gazebo camera stream, but the repository only tests the detector with a real V4L2 camera.

Troubleshooting

Symptom	Resolution
`package 'jetbot_ros' not found`	Source `/opt/ros/jazzy/setup.bash` and `~/jetbot_edgeimpulse_ws/install/setup.bash`.
Camera device is busy	Stop any other camera client using the same `/dev/video*` device before starting the launch.
Qwiic module cannot be imported	Re-run the provisioner, or install the SparkFun Qwiic packages into the system Python ROS 2 uses.
Waveshare motor import fails	Install `Adafruit-MotorHAT` in the Python environment used by the motor node.
No detections	Check the `.eim` path, `video_device`, model labels, and `confidence_threshold`. Check that the camera is publishing with `ros2 topic hz /jetbot/camera/image_raw`, and check `/diagnostics` for inference activity.
Robot moves but detections are unstable	Turn off autonomous behavior, collect images from the mounted camera, retrain the model, and raise `confidence_threshold` before you turn behavior back on.

Next steps

Add a behavior node that subscribes to /edgeimpulse_detector/detections and publishes /cmd_vel commands with safety checks.
Watch the FPS and latency on /diagnostics under load.
Collect more training data from the camera’s final position.
Test autonomous behavior at low speed before you increase it.

OVERVIEW

END-TO-END

TOPICS

TOOLS

INTEGRATIONS

Run Edge Impulse vision on a JetBot with Rubik Pi

What you’ll build

Prerequisites

1. Provision the Rubik Pi

2. Build a shared ROS 2 workspace

3. Start motor control and object detection

4. Inspect inference output

Motor and camera checks

Simulation notes

Troubleshooting

Next steps

​What you’ll build

​Prerequisites

​1. Provision the Rubik Pi

​2. Build a shared ROS 2 workspace

​3. Start motor control and object detection

​4. Inspect inference output

​Motor and camera checks

​Simulation notes

​Troubleshooting

​Next steps

What you’ll build

Prerequisites

1. Provision the Rubik Pi

2. Build a shared ROS 2 workspace

3. Start motor control and object detection

4. Inspect inference output

Motor and camera checks

Simulation notes

Troubleshooting

Next steps