EnviroSim Guide

Complete Documentation

Comprehensive guide covering ML models, backend APIs, frontend architecture, and everything you need to use, integrate with, or contribute to EnviroSim.

Overview

EnviroSim is a comprehensive environmental risk monitoring and prediction platform for Bangalore. It combines machine learning models with interactive visualization to help users understand environmental risks across 26+ neighborhoods.

The platform processes real-time environmental data and generates predictions for:

  • Air Pollution (PM2.5) - Next-day air quality predictions
  • Flood Risk Probability - Flood likelihood based on rainfall and area topology
  • Temperature Predictions - Heat stress index forecasts

System Architecture

EnviroSim consists of three main components working in harmony:

Frontend (Next.js :3000)
    ↓
Backend (Express :6969)
    ↓
InferenceService (FastAPI :8000)
    ↓
ML Models + Real Data

Frontend

Next.js React app with interactive maps and controls for scenario simulation

Backend

Express server for validation, routing, and API gateway between frontend and inference

Inference

FastAPI service running ML models for real-time predictions

Getting Started

For Users:

  1. Visit the Dashboard to view current environmental risks across Bangalore
  2. Use the Simulation tool to test what-if scenarios
  3. Adjust pollution, rainfall, temperature, and vegetation levels
  4. View risk predictions on the interactive heatmap
  5. Compare scenarios to understand environmental impacts