Eco-Guard-Ramsar - Edge-AI Ecological Monitoring System

Solar-powered Edge-AI for wetland monitoring at Ramsar sites. Integrates physico-chemical sensors with AI-based biological indicator detection.

---

About This Project

Eco-Guard-Ramsar is an end-to-end framework for the PeachBot Eco autonomous monitoring units, specifically designed for deployment in protected ecosystems. Validated at Sasthamcotta Lake, India (Ramsar Site #1212), this system bridges the gap between manual water sampling and high-frequency digital environmental monitoring.

The framework combines Hardware Systems Engineering with Environmental Data Science—a critical capability for Singapore's "Green Plan 2030" initiatives and global wetland conservation.

Repository Topics

edge-ai iot environmental-monitoring sustainability smart-water raspberry-pi tensorflow-lite ramsar-wetlands conservation-tech

---

Research Foundation

This project implements the technical framework for peer-reviewed research:

"Dedicated Edge-AI Single-Board Computer Systems for Ecological Monitoring in Protected Wetlands: Evidence from a Ramsar Site in India."

Author: Swapin Vidya
Affiliation: PeachBot Technologies (Agastya Biologic Solutions)

---

Core Innovation

The system uniquely combines three engineering disciplines:

1. Energy-Aware Computing

• Optimized for 98% operational uptime on solar-limited power budgets
• Deterministic inference latency: 120–180 ms
• Local data buffering during network outages
• Adaptive sampling based on battery state-of-charge

2. Multi-Modal Sensing

Unlike standard environmental loggers, Eco-Guard-Ramsar processes:

• Chemical Telemetry: pH, Turbidity, Nitrogen levels via analog sensor integration
• Biological Telemetry: Computer vision-based identification of indicator species (Chaoborus spp., E. coli presence)
• Environmental Context: Real-time power generation monitoring

3. Field Resilience

• Low-bandwidth MQTT communication for intermittent connectivity
• Graceful degradation when sensors fail
• Automatic data recovery and timestamp consistency
• Designed for rural wetland environments with minimal infrastructure

---

System Architecture

Dual-Unit Deployment

• Unit A: Real-time monitoring of pH, Turbidity, and Nitrogen levels
• Unit B: AI-assisted image analysis for biological indicators (E. coli, Chaoborus spp.)

Deployment Specifications

Specification	Details
Operational Uptime	98% (Field Validated)
Inference Latency	120–180 ms (TFLite)
Deployment Site	Sasthamcotta Lake, Ramsar Site #1212
Primary Stack	Python / TFLite / MQTT / Solar-SBC
Power Source	Solar with battery backup
Connectivity	MQTT (low-bandwidth) with local fallback

---

Hardware Deployment Platforms

Eco-Guard-Ramsar supports dual-unit deployment with platform-specific optimizations:

Unit A: Physico-Chemical Monitoring (Raspberry Pi 4B)

For sensor data collection, MQTT publishing, and edge processing

• Processor: ARM Cortex-A72 (1.5 GHz quad-core)
• Power: 2.5–3.5W (ideal for solar deployment)
• Sensors: pH, Turbidity, Nitrogen via ADS1115 I2C ADC
• Communication: Ethernet or WiFi 5
• Reference Guide: raspberry_pi_deployment.md

Unit B: Biological Detection (Jetson Nano)

For AI-based species detection with GPU-accelerated inference

• Processor: NVIDIA Tegra X1 + 128-core Maxwell GPU
• Power: 5–10W (higher performance, GPU-optimized)
• Camera: Raspberry Pi v2 (8 MP) via CSI interface
• Inference: TensorFlow Lite with GPU delegate (120–150 ms latency)
• Reference Guide: jetson_nano_deployment.md

Quick Hardware Comparison

Feature	Jetson Nano	Raspberry Pi 4B
Best For	AI/computer vision detection	Sensor data + MQTT
Inference Speed	120–150 ms (GPU-accelerated)	180–250 ms (CPU-only)
Power Consumption	5W idle → 10W inference	2.5W idle → 5W active
GPIO/I2C Support	Limited	Full flexibility (27 GPIO)
Total Cost	~$290 (with sensors)	~$300 (with sensors)

👉 See hardware_configs/README.md for complete platform selection guide, hardware shopping list, and network topology.

---

Data Samples & Field Validation

The data_samples/ directory contains representative datasets from field deployments at Sasthamcotta Lake:

Available Datasets

• sensor_readings_2026_02.csv: 30 days of hourly physico-chemical data

  • pH, Turbidity (NTU), Nitrogen (mg/L), Temperature, Battery SOC, Signal Strength
  • 720 measurements per month

• biological_detections_2026_02.csv: 30 days of AI classification results

  • Species detected (Healthy Algae, Chaoborus spp., E. coli)
  • Confidence scores (0.70–0.99 range)
  • Model version tracking

📊 Load sample data for analysis:

import pandas as pd

# Sensor data
sensors = pd.read_csv('data_samples/sensor_readings_2026_02.csv')
print(f"pH (mean): {sensors['ph'].mean():.2f}")
print(f"Turbidity (mean): {sensors['turbidity_ntu'].mean():.2f}")

# Biological detections
bio = pd.read_csv('data_samples/biological_detections_2026_02.csv')
high_conf = bio[bio['confidence_score'] > 0.90]
print(f"High-confidence detections: {len(high_conf)}/{len(bio)}")

👉 See data_samples/README.md for data interpretation guide and usage examples.

---

Tech Stack

Hardware

• Raspberry Pi / Low-power SBC
• Solar charge controller integration
• Environmental sensors (pH, turbidity, nitrogen probes)

Software & AI/ML

• Language: Python 3.13+
• Scientific Computing: NumPy, SciPy
• Deep Learning: TensorFlow Lite (edge inference)
• Data Processing: Pandas
• Communication: MQTT (low-bandwidth telemetry)
• Hardware Integration: RPi.GPIO, Adafruit CircuitPython

---

Project Structure

eco-guard-ramsar/
├── main.py                          # Entry point - field deployment loop
├── requirements.txt                 # Python dependencies
├── README.md                        # This file
├── src/
│   ├── __init__.py                 # Package initialization
│   └── bio_monitor.py              # Biological indicator AI engine
├── models/                          # TensorFlow Lite model files
├── data_samples/                   # Representative field data
│   ├── README.md                   # Data documentation
│   ├── sensor_readings_2026_02.csv # 30 days of physico-chemical data
│   └── biological_detections_2026_02.csv # 30 days of AI detection results
├── hardware_configs/               # Platform-specific deployment guides
│   ├── README.md                   # Hardware comparison & selection guide
│   ├── jetson_nano_deployment.md   # Unit B setup (AI/GPU optimization)
│   ├── raspberry_pi_deployment.md  # Unit A setup (sensors/GPIO)
│   ├── sensor_calibration.yaml     # Analog sensor calibration values
│   ├── mqtt_config.env             # MQTT broker credentials (template)
│   ├── tflite_model_config.yaml    # TensorFlow Lite inference config
│   └── power_management.yaml       # Solar charging & power budgets
└── .git/                           # Version control

Key Modules

src/bio_monitor.py

Implements the biological indicator detection system:

• BiologicalIndicatorAI class for species detection
• Statistical inference or TFLite model integration
• Support for E. coli and Chaoborus spp. identification
• Confidence scoring for detection results

main.py

Main field deployment loop:

• Collects physico-chemical parameters (pH, Turbidity, Nitrogen)
• Triggers biological AI analysis
• Logs all measurements with timestamps
• Runs continuously until user interruption (Ctrl+C)
• Comprehensive error handling for sensor/inference failures

---

Installation & Setup

Prerequisites

• Python 3.10 or higher
• pip package manager
• Virtual environment (recommended)

Step 1: Clone Repository

git clone <repository-url>
cd eco-guard-ramsar

Step 2: Create Virtual Environment

python -m venv .venv

Activate Virtual Environment:

• Windows: .venv\Scripts\activate
• macOS/Linux: source .venv/bin/activate

Step 3: Install Dependencies

pip install -r requirements.txt

For Development (includes testing tools):

pip install -r requirements.txt pytest black pylint

Step 4: Verify Installation

python -c "from src.bio_monitor import BiologicalIndicatorAI; print('Installation successful!')"

---

Running the Monitoring System

Quick Start

python main.py

Sample Output

[2026-02-17 10:52:41] INFO: [POWER] Solar-Powered Monitoring Unit Active: Sasthamcotta Lake (Ramsar Site)
[2026-02-17 10:52:41] INFO: [UPTIME] Target: 98% operational availability
[2026-02-17 10:52:41] INFO: [INIT] Biological Indicator Engine Initialized...
[DATA #1] pH: 8.42 | Turbidity: 4.18 NTU | Nitrogen: 1.64 mg/L
[AI]   Detection: Healthy Algae (confidence: 92.72%)
----------------------------------------------------------------------
[DATA #2] pH: 6.68 | Turbidity: 2.88 NTU | Nitrogen: 1.24 mg/L
[AI]   Detection: Healthy Algae (confidence: 91.87%)
----------------------------------------------------------------------

Stop Monitoring

Press Ctrl+C to safely shutdown the unit.

---

Deployment Configuration

Environment Variables (optional)

Create a .env file in the root directory:

# MQTT Configuration
MQTT_BROKER=<broker-address>
MQTT_PORT=1883
MQTT_TOPIC=ramsar/sasthamcotta/telemetry

# Sensor Calibration
PH_CALIBRATION_OFFSET=0.0
TURBIDITY_SCALE=1.0
NITROGEN_SCALE=1.0

# Inference Settings
CONFIDENCE_THRESHOLD=0.85

---

Troubleshooting

Import Error: ModuleNotFoundError: No module named 'src'

Solution: Ensure you're running from the project root directory:

cd eco-guard-ramsar
python main.py

Missing Dependencies

Solution: Install all requirements:

pip install -r requirements.txt

Sensor Data Not Reading

Check:

1. GPIO connections are properly secured
2. Sensor firmware is up to date
3. I2C/SPI interfaces enabled (Raspberry Pi): sudo raspi-config

AI Detection Issues

Check:

1. TensorFlow Lite installed: pip install tensorflow-lite
2. Model file path is correct if using bio_classifier.tflite

---

Development Workflow

Code Quality

# Format code
black src/ main.py

# Check for issues
pylint src/

# Run tests (if tests directory exists)
pytest tests/

Contributing

1. Create feature branch: git checkout -b feature/your-feature
2. Make changes and test locally
3. Commit with clear messages: git commit -m "Add feature description"
4. Push to repository and submit pull request

---

Dependencies Overview

Package	Version	Purpose
numpy	≥1.24.0	Scientific computing
scipy	≥1.10.0	Advanced math/statistics
tensorflow-lite	≥2.13.0	Edge AI inference
pandas	≥2.0.0	Data handling
paho-mqtt	≥1.6.1	IoT communication
RPi.GPIO	≥0.7.0	Raspberry Pi GPIO
Pillow	≥9.5.0	Image processing

See requirements.txt for complete dependency list with version specifications.

---

Future Enhancements

• Real-time data dashboard with web UI
• Cloud synchronization with local fallback storage
• Multi-model inference pipeline
• Battery health monitoring and predictive maintenance
• Mobile app for remote monitoring and alerts

---

GitHub Repository Configuration

To maximize discoverability and showcase this project's unique impact, configure your GitHub repository as follows:

1. Repository Description (160 characters)

Copy and paste into the "About" section (via the ⚙️ gear icon):

Solar-powered Edge-AI for wetland monitoring at Ramsar sites. Integrates physico-chemical sensors with AI-based biological indicator detection. 🌿🔋

2. Add Repository Topics

In the repository settings, add these topic tags under Topics:

• edge-ai
• iot
• environmental-monitoring
• sustainability
• smart-water
• raspberry-pi
• tensorflow-lite
• ramsar-wetlands
• conservation-tech

These tags improve discoverability for:

• IoT and hardware recruiters
• Environmental science teams
• Sustainability-focused organizations
• Green technology initiatives (e.g., Singapore's Green Plan 2030)

3. Steps to Update

1. Navigate to your eco-guard-ramsar repository on GitHub
2. Click the Settings tab
3. In the left sidebar, scroll to "About" section
4. Click the ⚙️ cog icon to edit repository details
5. Paste the 160-character description
6. Add the topic tags listed above
7. Set visibility to Public (for portfolio visibility)
8. Click Save changes

---

License & Citation

If using this work in research, please cite:

@software{vidya2026ecoguard,
  author = {Vidya, Swapin},
  title = {Eco-Guard-Ramsar: Edge-AI Systems for Ecological Monitoring},
  year = {2026},
  organization = {PeachBot Technologies (Agastya Biologic Solutions)}
}

---

References

• TensorFlow Lite Guide
• Raspberry Pi Documentation
• MQTT Specification
• Ramsar Convention on Wetlands

---

Last Updated: February 17, 2026
Project Status: Production Deployment
Python Version: 3.10+