Get Free Synopsis PDF
We will prepare a detailed synopsis for your college submission. Includes problem statement, objectives, DFD, ER diagram, and more.
A data science project that collects air quality data from public APIs and IoT sensors, visualizes pollution patterns on interactive maps, and predicts AQI levels using machine learning. The system helps citizens make informed decisions about outdoor activities and health precautions.
Problem Statement
Air pollution is a critical health crisis in Indian cities, but most citizens lack access to hyperlocal air quality information. Government monitoring stations are sparse, and existing AQI apps show city-level averages that do not reflect neighborhood-level variations. A system that combines official data with low-cost IoT sensors and predictive modeling can provide more granular, actionable air quality information.
Objectives
Collect and aggregate air quality data from multiple sources
Build interactive pollution heatmaps and trend visualizations
Train ML models for 24-48 hour AQI prediction
Create health advisory recommendations based on AQI levels
Design a public-facing dashboard and API
Modules
Data Collection
Fetch data from OpenAQ, CPCB APIs, and optionally from DIY PM2.5 sensors built with SDS011 and ESP32.
Data Processing
Clean, interpolate, and normalize air quality data with outlier detection and gap filling.
Prediction Model
LSTM neural network for time-series AQI forecasting incorporating weather data as features.
Visualization Dashboard
Interactive maps showing pollution heatmaps, station-wise trends, and prediction confidence intervals.
Health Advisory
Context-aware health recommendations based on predicted AQI, user health profile (asthma, elderly), and planned activities.
Dataset
OpenAQ Platform (open-source air quality data), CPCB (Central Pollution Control Board) real-time data API.
Expected Output
A dashboard displaying current and predicted air quality across city locations on an interactive map, with historical trend analysis, health advisories, and the ability to explore data by pollutant type (PM2.5, PM10, NO2, O3). The prediction model achieves at least 80% accuracy for next-day AQI forecasting.
Future Scope
Community-driven sensor network expansion, integration with smart home systems (auto-close windows), correlation analysis with health data, policy recommendation engine for city planners, and a mobile app with personalized pollution exposure tracking.