🐟 Smart Fish Feeder v2.0

An IoT-based automated fish feeding system powered by ESP32, featuring real-time water quality monitoring, AI-based anomaly detection, cloud sync via Firebase, and an OLED dashboard — all managed through a modular production-ready codebase.

---

✨ Features

• 🕒 Scheduled Auto-Feeding — DS3231 RTC-based precise feeding schedule
• 🌡️ Temperature Monitoring — DS18B20 waterproof temperature sensor with Kalman filtering
• 🧪 pH Monitoring — Calibrated analog pH sensor with temperature compensation
• 💧 Turbidity Monitoring — Real-time water clarity tracking in NTU
• 🤖 AI / Predictive Analytics — Water quality scoring and anomaly detection algorithm
• ☁️ Firebase Cloud Sync — Real-time database updates every 5 seconds
• 📟 OLED Dashboard — SH1106 display showing live sensor readings
• 🔔 Buzzer Alerts — Critical alerts triggered on anomaly detection
• 🐾 Watchdog & Low Power Manager — System reliability and power efficiency
• 📋 Serial Logger — Heartbeat logging and event tracking via Serial Monitor

---

🛠️ Hardware Requirements

Component	Description
ESP32	Main microcontroller (Wi-Fi + BLE)
DS3231 RTC	Real-time clock module (I2C)
DS18B20	Waterproof temperature sensor (OneWire)
pH Sensor	Analog pH probe + signal board
Turbidity Sensor	Analog turbidity sensor
SG90 / MG995 Servo	Servo motor for feeder mechanism
SH1106 OLED	1.3" I2C OLED display
Buzzer	Active or passive buzzer

---

📁 Project Structure

FishFeeder/
├── FishFeeder.ino          # Main sketch — entry point
├── data/                   # SPIFFS data files (if any)
└── src/
    ├── Core/
    │   ├── SystemWatchdog.h        # Hardware watchdog handler
    │   └── LowPowerManager.h      # Power management logic
    ├── Timekeeping/
    │   ├── DS3231_RTC.h            # RTC initialization and time reading
    │   └── FeedScheduler.h        # Feeding schedule management
    ├── Sensors/
    │   ├── DS18B20.h               # Temperature sensor with filtering
    │   ├── pH_Sensor.h             # Calibrated pH reading
    │   └── Turbidity.h            # Turbidity with Kalman filter
    ├── Actuators/
    │   ├── ServoFeeder.h           # Servo motor control for food dispensing
    │   └── Buzzer.h               # Alert tones and patterns
    ├── Displays/
    │   ├── OLED_SH1106.h          # OLED dashboard rendering
    │   └── SerialMonitor.h        # Serial logging and heartbeat
    └── Cloud/
        ├── FirebaseClient.h        # Firebase RTDB read/write
        └── PredictiveAlgo.h       # Water quality score + anomaly detection

---

⚙️ Setup & Configuration

1. Install Arduino Libraries

Install the following via Arduino Library Manager or manually:

• Firebase ESP Client by Mobizt — v4.4.17
• ESP32Servo
• Wire (built-in)
• WiFi (built-in for ESP32)

2. Configure Credentials

Open FishFeeder.ino and edit these 4 lines:

const char* WIFI_SSID     = "YOUR_WIFI_SSID";
const char* WIFI_PASSWORD = "YOUR_WIFI_PASSWORD";
const char* FIREBASE_API_KEY = "YOUR_FIREBASE_WEB_API_KEY";
const char* DATABASE_URL  = "https://your-project-default-rtdb.region.firebasedatabase.app/";

⚠️ Never commit real credentials to GitHub. Use a .gitignore or environment-based config for production.

3. Firebase Setup

1. Go to Firebase Console
2. Create a new project → Enable Realtime Database
3. Copy the Web API Key from Project Settings
4. Copy the Database URL from Realtime Database settings
5. Set database rules to allow read/write (or configure auth)

4. pH Sensor Calibration

Calibration values are set in setup():

pHSensor.calibrate(1.75, 2.05);  // pH4=1.75V, pH7=2.05V

Adjust these values based on your actual buffer solution measurements.

5. Upload

• Select board: ESP32 Dev Module
• Upload speed: 115200
• Flash FishFeeder.ino via Arduino IDE

---

☁️ Firebase Data Structure

Data is pushed to /sensors/latest every 5 seconds:

{
  "device_id": "fish_feeder_001",
  "timestamp": 12345678,
  "uptime_s": 3600,
  "temp_c": 27.5,
  "ph": 7.2,
  "turbidity_ntu": 45.3,
  "quality_index": 85,
  "feeding_active": false,
  "anomaly_detected": false
}

---

🔄 System Loop (1-second cycle)

Each second, the system executes:

1. Feed Watchdog — keeps the system alive
2. Read Sensors — temperature, pH, turbidity
3. AI Processing — quality score + anomaly detection
4. Actuation — check and execute feeding schedule
5. Update Display — refresh OLED + Serial log
6. Cloud Sync — push JSON to Firebase every 5s
7. Critical Alerts — trigger buzzer if anomaly detected

---

📡 I2C Pin Mapping (ESP32)

Signal	ESP32 Pin
SDA	GPIO 21
SCL	GPIO 22

---

🧠 AI / Quality Index

The PredictiveAlgo module computes a water quality index (0–100) based on temperature, pH, and turbidity readings. If values fall outside safe thresholds, anomaly_detected is set to true, triggering buzzer alerts and a Firebase log event.

---

📜 License

This project is licensed under the terms found in the LICENSE file.

---

👤 Author

CJ ChANu — SLIIT IoT Project
GitHub Profile