An Air, Water, and Soil Quality Monitoring and Forecasting System
Using Wireless Sensor Networks and Random Forest Algorithm

System Overview

0

TOTAL NODES

0

ACTIVE NODES

0

TOTAL SENSORS

0

AIR SENSORS

0

WATER SENSORS

0

SOIL SENSORS

About EcoSense

EcoSense is an integrated environmental monitoring and forecasting system developed to address the growing challenges of air, water, and soil quality degradation. The system utilizes dynamic wireless sensor networks, machine learning-based forecasting, and a web-based dashboard to provide real-time environmental quality assessment and predictive insights.

Unlike existing systems that focus on a single environmental domain, EcoSense unifies Air Quality Index (AQI), Water Quality Index (WQI), and Soil Quality Index (SQI) into one modular and portable platform, improving environmental coverage, adaptability, and decision support.

Problem Addressed

Environmental degradation remains a major threat to public health, ecosystems, and sustainable development.
However, most existing monitoring systems:

Focus on only one environmental domain
(air or water or soil)

Use static sensor
nodes with limited spatial coverage

Lack real-time forecasting capabilities

Provide poor or fragmented data visualization

Objectives of the
System

EcoSense was developed to:

Monitor air, water, and soil quality using a single
adaptive sensor node

Forecast environmental quality levels using the Random Forest
algorithm

Maintain reliable data transmission through a wireless sensor
network

Present real-time and forecasted data through an interactive web
dashboard

Improve user understanding and decision-making through clear
visualization

System Features

Adaptive Sensor Node

  • Switches between air, water, and soil monitoring modes
  • Portable and modular design
  • Measures key environmental parameters:
    • Air: PM2.5 / PM10, gas concentration
    • Water: pH level, turbidity
    • Soil: pH level, soil moisture

Wireless Sensor Network (WSN)

  • Ensures real-time data transmission
  • Supports sensor mobility and repositioning
  • Evaluated based on:
    • Throughput
    • Packet delivery ratio
    • Packet loss
    • Latency and reliability

Machine Learning Forecasting

  • Uses the Random Forest algorithm
  • Trained on raw sensor data and engineered features
  • Forecasts daily environmental quality levels
  • Model performance evaluated using:
    • RMSE
    • MAE
    • R² score

Web-Based Dashboard

  • Displays real-time and historical data
  • Visual comparison of actual vs forecasted values
  • Integrates AQI, WQI, and SQI in one interface
  • Evaluated using:
    • Perceived Usefulness
    • Perceived Ease of Use (Technology Acceptance - Model TAM)

How EcoSense Works

Data
Collection

Sensor nodes collect air, water, or soil quality parameters.

Edge Processing

Noise filtering, outlier detection, and data tagging are performed before transmission.

Wireless Transmission

Data is sent to a central server via a wireless sensor network.

Forecasting

The Random Forest model predicts future environmental quality levels.

Visualization

Data and forecasts are displayed on the web-based dashboard.

Researchers: Ram Vincent M. Escoto Jacob Lawrence D. Alfante Neil Christian A. Aringo

Research Adviser: Engr. Ricrey E. Marquez, PCpE

EcoSense is a Bachelor of Science in Computer Engineering thesis project developed at:
Colegio de Muntinlupa - Computer Engineering Department