NoPawZone

🐾 NoPawZone – Real-Time Pet Detection Signage with Grove Vision AI & ESP32-S3

We love floofs, but rules are rules.

ssss

It all began at MakerChat0x2D — a month-long build series hosted by MakerGram, focused on the theme of Interactive Digital Signage. Every Thursday, makers gathered at TinkerSpace to brainstorm ideas, build prototypes, and share progress in a collaborative, hands-on environment.

During one of these Thursdays, I designed a digital signage concept called Motion Mirror - a playful interactive display where an animated character on screen mimics your head movements.

That project sparked a curiosity - What else Ican built?

Inspired by that, I began building NoPawZone over the following weeks—a fun and functional real-time pet detection signage system. It uses the Grove Vision AI Module V2 with a pretrained animal detection model to identify cats or dogs through a live video feed. Powered by the XIAO ESP32-S3 Sense, the detection data is streamed directly to a browser using the Serial Web API, triggering alerts.

🐶 What Is NoPawZone?

NoPawZone is a digital signage system that detects if a cat or dog appears in the camera’s view. If a furry visitor is spotted, it instantly displays a playful but firm message:

"We spotted a floof! As much as we love pets, they gotta wait outside."

"A whiskered visitor detected! Sorry kitty, this space is pet-free."

It combines real-time detection, serial communication, and smooth frontend animation to make signage feel alive and responsive.

🔧 How It Works

The detection is powered by the Grove Vision AI Module V2 running a pre-trained model from Seeed Studio that identifies dogs and cats. The module connects to a XIAO ESP32-S3, which acts as a serial bridge, sending detection results to the browser using the Serial Web API.

The browser reads the serial input, and the React-based frontend updates the screen in real-time with animations and context-aware messages.

🛠 Hardware

• Grove Vision AI V2
• XIAO ESP32S3
• USB-C

📟 XIAO ESP32S3 Arduino Code

Here’s the Arduino code that runs on the ESP32-S3 and prints the pet detection status via serial:

#include <Seeed_Arduino_SSCMA.h>
SSCMA Infer;

void setup()
{
    Serial.begin(9600);
    Infer.begin();
}

void loop()
{
    if (!Infer.invoke())  // If inference was successful
    {
        auto boxes = Infer.boxes();

        bool catDetected = false;
        bool dogDetected = false;

        for (int i = 0; i < boxes.size(); i++)
        {
            int confidence = boxes[i].score;
            int label = boxes[i].target;

            if (confidence >= 60)
            {
                if (label == 0)
                    catDetected = true;
                else if (label == 1)
                    dogDetected = true;
            }
        }

        if (catDetected && dogDetected)
            Serial.println("both");
        else if (catDetected)
            Serial.println("cat");
        else if (dogDetected)
            Serial.println("dog");
        else
            Serial.println("nothing");
    }
    else
    {
        Serial.println("nothing");
    }

    delay(100);
}

🎯 Why NoPawZone?

Instead of putting up static, easy-to-ignore “No Pets” signs, NoPawZone reacts to the presence of pets in real-time. The system enforces rules—but with a little heart and humor.

It’s perfect for:

• Cafés and food areas
• Clean rooms or labs
• Clinics and hospitals
• Co-working spaces with mixed policies

It makes public spaces more intelligent, expressive

Huge thanks to MakerGram ❤️‍🔥