Greetings from Arduino UNO Q

Servo Motor SG90 180 degree

USB Webcam (PK-910H)

3D Printer (Optional)

Arduino App Lab

Arduino® UNO™ Q 2GB

USB Hub, 4 Port

<html><body>Imagine walking up to your desk and being greeted by a friendly wave. This project turns your Arduino UNO Q into a polite digital companion that adds a touch of personality to your workspace. Instead of just sitting idly, your board actively looks out for you and reacts to your presence. By combining a USB webcam with advanced face recognition models, the system continuously monitors the environment in real-time. As soon as you approach your desk, the webcam detects your face and triggers a servo motor to wave a 3D-printed hand, while the on-board LED matrix lights up with a smile. This application runs entirely on the Arduino UNO Q. The face detection model is processed locally on the board, ensuring zero latency and complete privacy. <h3>Build the Greeter</h3>The Arduino UNO Q acts as the central hub, powering both the AI processing and the physical actuators. Here is the full hardware lineup: Here is the full hardware lineup: <ol><li>Vision Input → A standard USB Webcam, connected via the USB-C Hub. It acts as the board’s eyes, feeding video frames to the detection model. </li><li>Waving Mechanism → A Servo Motor connected to digital pin D9 and powered directly by the board's 5V output. It drives the 3D-printed hand to perform the greeting animation. </li><li>Structure → A Modulino Base Stand (or custom 3D printed mounts) holds the board, the servo, and the hand in place for a stable setup. </li></ol> <figure><img src="https://projects.arduinocontent.cc/ec95d028-e01f-4d69-82d4-eac5691378ea.png"/><figcaption></figcaption></figure><h3>The Dashboard</h3>The application includes a built-in dashboard that lets you see exactly what the system "sees" and track its activity in real time. Since the web interface is hosted directly on the Arduino UNO Q, you can access it from any device on the same local network It features four main areas: <ol><li>Video Face Detection: The main view shows the live webcam feed. When the AI recognizes a face, a red bounding box appears around it with a live confidence score (e.g., 0.94), confirming exactly who is being tracked. </li><li>Confidence Threshold: A slider allows you to adjust the sensitivity. You can set a minimum confidence level (0-1) required to trigger the greeting, filtering out false positives or background noise. </li><li>Feedback Status: When a face is detected, it displays a waving hand icon 👋 and a random greeting message, mirroring the physical action of the servo. </li><li>Recent Detections: To keep a history of interactions, this list logs the last 5 detection events with their confidence percentage and exact timestamp. </li></ol> <figure><img src="https://projects.arduinocontent.cc/5357d67f-a560-477a-af86-25beb7d0df23.png"/><figcaption></figcaption></figure><h3>The Vision model</h3>The face detection capability is powered by a machine learning model developed using <a href="https://docs.edgeimpulse.com/hardware/boards/arduino-uno-q">Edge Impulse </a>. Specifically, the project utilizes FOMO (Faster Objects, More Objects), a ground-breaking object detection algorithm designed for constrained devices. Unlike traditional heavy neural networks, FOMO is lightweight and incredibly fast. This efficiency allows the Arduino UNO Q to process the video feed locally in milliseconds, ensuring the system reacts instantly when a face is spotted. <h3>Software Architecture Overview</h3>The software is developed entirely within the Arduino App Lab, leveraging the board's dual-processor architecture to separate high-level processing from real-time hardware control. The responsibilities are split as follows: <ol><li>Processor: Running Python, this core handles the intelligence of the system. It manages the Video Object Detection Brick, which processes the video stream to identify faces using the FOMO model. It also hosts the logic for the WebUI, updating the dashboard and deciding when a detection is confident enough to trigger an event. </li><li>MCU: Running the Arduino Sketch, this core is dedicated to physical actuation. It directly controls the Servo Motor and the LED Matrix. </li><li>Communication (RPC): The two cores communicate via RPC (Remote Procedure Call) using the <a href="https://github.com/arduino-libraries/Arduino_RouterBridge">Arduino_RouterBridge </a>library. </li></ol> <figure><img src="https://projects.arduinocontent.cc/a0b043ee-403c-48d4-81fb-97be82b85abe.png"/><figcaption></figcaption></figure><h3>A quick look into the Code</h3>In this section, we’ll focus on the flow of: Detecting a face and Waving. The Processor: Detecting a Face On the Processor side (main.py), the logic is event-driven. We use the detection_stream object to listen specifically for the label "face". When a face is detected, the script uses the Bridge to send a command to the microcontroller: <pre># Define what happens when a face is seen </pre><pre>def face_detected(): </pre><pre>	 # Trigger the C++ function named 'greet' via RPC </pre><pre>	 Bridge.call("greet") </pre><pre>	 print("Face detected!") </pre><pre># Link the event to the specific label "face" </pre><pre>detection_stream.on_detect("face", face_detected) </pre>The MCU: Waving the hand On the Microcontroller side (sketch.ino), the board acts as the listener. In the setup(), we register the function so it can be called remotely: <pre>Bridge.provide("greet", greet); </pre>The greet() function handles the physical animation. It updates the LED Matrix to a "Happy" expression and performs a sweep with the servo motor: <pre>void greet() { </pre><pre>	 digitalWrite(LED_BUILTIN, HIGH); </pre><pre>	 matrix.loadFrame(hello_face); // Switch to happy face </pre><pre>	 // Sweep the hand 3 times </pre><pre>	 for (int i = 0; i &lt; 3; i++) { </pre><pre>	 // Move hand Up </pre><pre>	 for (int j=idle_position; j&lt;max_position; j++){ </pre><pre>	 servo.write(j); </pre><pre>	 delay(5); </pre><pre>	 } </pre><pre>	 // Move hand Down </pre><pre>	 for (int j=max_position; j&gt;min_position; j--){ </pre><pre>	 servo.write(j); </pre><pre>	 delay(5); </pre><pre>	 } </pre><pre>	 // ... return to center </pre><pre>	 } </pre><pre>	 matrix.loadFrame(idle_face); // Return to idle face </pre><pre>} </pre>For the complete code, including the 3D printed parts, see the Code section below. </body></html>

Greetings from Arduino UNO Q

Transform your Arduino UNO Q into a polite digital companion that recognizes you and waves hello whenever you approach your desk, using face recognition models and a servo motor.