DSN3099 GitHub Details, Stars and Alternatives | OpenRepoFinder
Varaxion / repository
DSN3099
Sahay - A predictive disaster intelligence and rapid emergency response platform. Features ML-driven river flood forecasting (ARIMA), CNN-based rainfall analysis, a lightning-fast blood bank locator, and interactive disaster SOPs. Built for DSN3099 Engineering Project in Community Service @ VIT
A transparent discovery signal based on current public GitHub metadata.
Recent activity35% weight
72
Community adoption25% weight
0
Maintenance state20% weight
100
License clarity10% weight
0
Project information10% weight
75
This score does not audit code, security, maintainers, documentation quality, or suitability. Verify the repository and its current documentation before adoption.
Advanced hydrological forecasting, rainfall analysis, blood bank locator, emergency guidelines, and incident-reporting simulation for public safety and preparedness.
[!NOTE]
Sahay was originally developed in 2025 as part of the academic curriculum for the EPICS course at VIT Bhopal University. It has now been completely overhauled into a modern, dynamic web application, upgrading its model speed, repository structure, and visual aesthetics to state-of-the-art glassmorphic standards.
Sahay is a comprehensive Flask-based disaster resilience platform that fuses machine-learning hydrological intelligence with practical, on-the-ground emergency support utilities. Designed with a stunning dark glassmorphic UI, it provides a unified hub to:
Forecast river flood risks
Analyze monthly subdivision rainfall patterns
Locate life-saving blood banks
Access direct-dial emergency helplines
Study interactive disaster Standard Operating Procedures (SOPs)
Simulate mock incident-reporting for training and demonstrations
(The active v1.0 application is housed inside app/, while the preserved v0 legacy prototype is kept under v0Archive/ to document the project's evolution.)
✨ Core Features
Flood Prediction (ARIMA + LDA): Forecasts water-level metrics for major rivers (Cauvery, Godavari, Krishna, Mahanadi, Son) and classifies real-time flood risk using Machine Learning.
Rainfall Analysis (1D-CNN + RF): Predicts April-December monthly rainfall for Indian meteorological subdivisions using either a cached Deep Learning CNN or a Random Forest Regressor.
Blood Assistance Portal: A lightning-fast, AJAX-powered locator to find local blood bank records by state and city, complete with copy actions, direct dialers, and donor eligibility guidelines.
Emergency Contacts: A quick-dial public safety directory for Police, Ambulance, Fire, NDRF, NDMA, and Women Helplines.
Interactive Disaster SOPs: Actionable, tabbed do's and don'ts checklists for Floods, Earthquakes, Cyclones, Landslides, Tsunamis, Wildfires, and Thunderstorms.
Incident Reporting Simulator: A visually engaging prototype workflow simulating emergency dispatch routing and coordinate generation.
User inputs River & Date ➔ hydrology.py reads historical river workbook ➔ Generates ARIMA future trend ➔ Extracts structural features ➔ Passes vector to LinearDiscriminantAnalysis ➔ Classifies Normal vs. High Risk ➔ Renders Matplotlib chart & metrics table.
2. Rainfall Analysis Pipeline
User inputs Subdivision, Year & Model ➔ rainfall.py parses IMD dataset ➔ If CNN, loads cached Keras .h5 model (or trains on-the-fly) ➔ If RF, trains RandomForestRegressor ➔ Predicts April–December mm ➔ Calculates MAE & Variance ➔ Renders comparative chart.
3. Blood Bank Search
routes.py loads blood_banks_india.csv globally on server start ➔ Normalizes strings (Title Case) ➔ Exposes /getCities and /searchBloodBanks JSON endpoints ➔ Client-side JS fetches and renders results instantly without page reloads.
🚀 Getting Started
1. Clone the Repository
git clone https://github.com/Varaxion/DSN3099.git
cd DSN3099
LDA Covariance: Flood risk classification assumes equal covariance matrices across normal and high-risk classes.
Rainfall Granularity: Both CNN and RF models operate on aggregated monthly subdivision data, not hyperlocal or daily weather metrics.
Incident Reporting: The reporting module is strictly a UI prototype/simulation. It does not dispatch or alert real-world NDRF agencies.
Medical Data: Blood bank contact numbers and addresses are sourced from open datasets and should be independently verified in a real emergency.
📚 Data Sources
This academic project relies on several historical and open-source datasets to power its machine learning models and locator utilities:
Hydrological River Data (Cauvery.xlsx, Godavari.xlsx, etc.): Historical water level and runoff metrics sourced from the Central Water Commission (CWC) and India-WRIS (Water Resources Information S
