This project demonstrates the integration of Delsys Trigno sensors within a ROS2-based robotic control framework through an interactive drone game.

The demo showcases how wearable biosignal sensors can be used to control a virtual drone in real time:

• IMU sensor → controls the drone direction and movement.
• EMG sensor → controls speed boosts through real-time muscle activation.
• ROS2 handles the complete data acquisition and processing pipeline.
• Godot Engine provides the visualization and gameplay environment.

The goal is to provide a complete example of a Human-Machine Interface (HMI) based on physiological signals and ROS2 communication.

Delsys Trigno Sensors
          │
          ▼
delsys_connector_socket_ros.py
          │
          ├── TCP 9001 → imu_processing_node
          ├── TCP 9002 → emg_processing_node
          └── TCP 9003 → drone_controller_node
                          │
                          ▼
                     cmd_vel
                          │
                          ▼
                 ROSBridge WebSocket
                          │
                          ▼
                      Godot Game

The system uses three dedicated TCP ports:

Required ROS nodes:

The Delsys bridge communicates through sockets with the ROS nodes:

• imu_processing_node
• emg_processing_node

These nodes publish ROS topics.

The drone_controller_node subscribes to those topics, processes the information, and generates a cmd_vel command.

The command is sent through ROSBridge and consumed by Godot to control the virtual drone.

Linux PC:

• Ubuntu 22.04

• ROS2 Humble

• Python 3.10.12

Install the following packages:

sudo apt-get update

sudo apt-get install -y wmctrl xdotool

sudo apt install -y ros-humble-rosbridge-server

pip install PyQt5

This project was developed and tested with:

Godot_v4.3-stable_linux.x86_64

The executable must be placed inside:

delsys_bridge/

Example:

delsys_bridge/
├── Godot_v4.3-stable_linux.x86_64
├── AeroPy/
├── delsys_connector_socket_ros.py
└── ...

Before running the project, you must insert your Delsys license information.

Edit:

delsys_bridge/AeroPy/TrignoBase.py

and add your:

• License Key
• License Information

To obtain a valid license contact Delsys:

Europe:

https://delsyseurope.com/

USA:

https://delsys.com/

Before using your own Trigno sensors, you must update the hardware mapping.

Edit:

delsys_bridge/delsys_connector_socket_ros.py

Locate the mapping dictionary near the beginning of the file.

The mapping associates:

Hardware GUID  -> Logical Sticker ID

Example:

sensor_map = {
    "GUID_SENSOR_A": 1,
    "GUID_SENSOR_B": 2,
}

Update the GUIDs with your own sensors.

This mapping is fundamental because the entire ROS pipeline identifies sensors using the logical sticker numbers.

The entire system can be launched through:

python3 drone_launcher.py

The launcher automatically:

1. Builds the ROS workspace.
2. Starts ROSBridge.
3. Starts EMG processing.
4. Starts IMU processing.
5. Starts the Drone Controller.
6. Starts the Delsys bridge.
7. Opens the Godot simulation.
8. Displays the integrated monitoring GUI.

The demo follows a simple convention:

Example:

Sensors should be configured as:

EMG RMS (148 Hz, 100 ms window)
Skin Check (74 Hz)
Accelerometer ±2g (296 Hz)
Gyroscope ±250 dps (296 Hz)
EMG Range ±11 mV
Bandpass 20–450 Hz

EMG RMS (148 Hz, 100 ms window)

provides several advantages:

• RMS already computed onboard by the sensor.
• Direct measurement of muscle activation intensity.
• No software downsampling required.
• Reduced CPU load.
• Smooth signal ready for real-time control.

This makes it ideal for mapping muscle contraction intensity to drone speed.

Advantages:

• More robust.

• Handles strong contractions without saturation.

• Easier MVC calibration.

• Recommended for:

  • Biceps
  • Quadriceps
  • Forearm muscles

Disadvantages:

• Lower sensitivity for weak contractions.

The accelerometer range is expressed in multiples of gravity:

1g = 9.81 m/s²

Maximum measurable acceleration:

±19.6 m/s²

Advantages:

• Highest resolution.
• Ideal for normal arm movements.

The gyroscope range is expressed in:

Degrees Per Second (dps)

Advantages:

• Highest angular resolution.
• Best for controlled arm movements.

If needed, the system can also be started manually.

From:

ros2_ws

run:

colcon build

source install/setup.bash

source /opt/ros/humble/setup.bash
source install/setup.bash

ros2 run emg_processing emg_processing_node

Open a second terminal:

source /opt/ros/humble/setup.bash
source install/setup.bash

ros2 run imu_processing imu_processing_node

Open a third terminal:

source /opt/ros/humble/setup.bash
source install/setup.bash

ros2 run drone_controller drone_controller_node

Inside:

delsys_bridge

run:

source /opt/ros/humble/setup.bash

source ~/delsys-unimore/ros2_ws/install/setup.bash

python3 hud.py

Inside:

delsys_bridge

become root:

sudo su

then run:

python3 delsys_connector_socket_ros.py

The bridge will:

• Scan sensors.
• Assign IMU and EMG roles.
• Perform IMU calibration.
• Perform EMG calibration.
• Start/stop recording sessions.

From:

ros2_ws

run:

source /opt/ros/humble/setup.bash

source install/setup.bash

ros2 launch rosbridge_server rosbridge_websocket_launch.xml

Inside:

delsys_bridge

run:

./Godot_v4.3-stable_linux.x86_64

The project uses:

godot_ros

to subscribe to ROS topics.

The drone receives movement commands generated from:

IMU → ROS2 → drone_controller → cmd_vel → Godot

while EMG activity is used to modulate speed and boost behaviour.

University of Modena and Reggio Emilia (UNIMORE) - Alessandra Fava

Human–Machine Interaction Demo using Delsys Trigno Sensors, ROS2 Humble and Godot Engine.