🌿 NeoPlant: A Gamified, AI-Driven Plant Care Network

Plant Sync – Platform-Based Centralised/Decentralised Plant Care Ecosystem

NeoPlant transforms everyday plant care into an interactive, community-powered game. Combining IoT sensors, machine learning, and Unreal Engine 5, our system fosters collaboration through data sharing, rewards, and real-time simulation of your plants—both locally and socially.

---

🎯 Objective

To create a technology-powered, gamified platform where users monitor, nurture, and gamify plant care using real-time sensor data. Through smart automation, AI prediction, and social features, we promote sustainability in a fun and rewarding way.

---

🌟 Key Features

• 📷 Plant Classification
  Upload a photo to classify your plant into one of 5 categories (succulent, pothos, vine, air plant, foliage).

• 🎮 Pixel Avatar Spawn
  Automatically generate and spawn your plant as a pixel-style avatar inside a virtual UE5 room.

• 📡 IoT Sensor Integration
  ESP32 + DHT11 sensors monitor real-time soil moisture, temperature, and humidity.

• 💧 Remote Watering System
  Trigger an Arduino-controlled pump via the app to hydrate your plant.

• 💚 Health Bar Visualisation
  Health bar updates dynamically based on sensor data to reflect plant well-being.

• 🪙 Gamified System
  Earn reward coins for timely watering. Spend them in seasonal shops to unlock XP boosts, rare skins, and limited quests.

• 🏡 Friend Garden Mode
  Visit friends' virtual gardens, help water their plants, and rank on a seasonal leaderboard.

• 🔗 Smart Contract Inspired Resource Sharing
  (Optional) Distributed logic enables decentralized care actions—like alerting nearby users to share excess water or nutrients.

• 🧠 AI-Powered Predictions
  Forecast future plant care needs based on trends and past data patterns.

---

🧠 Machine Learning Model

• Type: ResNet18 Convolutional Neural Network (CNN)
• Framework: PyTorch (trained via Google Colab)
• Classes:
  • succulent
  • pothos
  • vine
  • air_plant
  • foliage
• Model File: plant_cnn_model.pth
• Deployment: REST API via Flask (/predict)

---

🛠️ System Architecture

User Uploads Plant Photo
        │
   ▶ Flask API (/predict)
        │
   ▶ CNN Classifies Plant
        │
   ▶ UE5 Blueprint Parses Result
        │
   ▶ Pixel Avatar Spawned in Virtual Room
        │
   ▶ Avatar Connects to Sensor Data via Serial
        │
   ▶ Health Bar & Watering Icon Live Update
        │
   ▶ Arduino Triggers Pump if Needed'

📁 File Structure

Plant_api/
├── app.py                  # Flask API endpoint
├── plant_cnn_model.pth     # Trained PyTorch model
├── test.py                 # Classification test script

CommunityPlant/
├── Content/
│   ├── Sprites/            # Pixel-style plant avatars
│   └── Blueprints/         # Watering + Health UI logic

esp32/
├── moisture_logger.ino     # Reads moisture & humidity
├── pump_controller.ino     # Controls watering pump

🚀 Deployment Guide

✅ Requirements

Hardware:

• ESP32 board
• DHT11 sensor
• Arduino-compatible water pump

Software:

• Python 3.x
• Flask
• PyTorch
• Unreal Engine 5 (Blueprints enabled)
• Google Colab (for training the model)

---

▶️ Run the Flask API Locally

This API receives plant images, performs classification, and sends the result to UE5.

# Navigate to the API folder
cd plant_api

# Install required dependencies
pip install flask torch torchvision pillow

# Run the API server
python app.py

The server will run at http://localhost:5000/predict by default.

---

👥 Contributors

This project was created with love, curiosity, and lots of plant waterings by:

• 🪴 Yufei – ML Engineer and UI Designer
• 🌸 Evie – Game Designer and Software Engineer
• 🌿 Vullnet – Product Manager and UX Researcher

Special thanks to our pixel plants for never giving up on us 🌱