Urban Equitability Index — Hyderabad
A full-stack spatial platform that scores and visualizes equity across all GHMC wards in Hyderabad — measuring access to schools, healthcare, and transit alongside opportunity, environment, and governance indicators into a single composite UEI score.
Interactive map — scroll and click to explore.
Context
Hyderabad's GHMC wards vary enormously in access to government schools, hospitals, bus stops, metro stations, and green space — but no unified, ward-level framework existed to compare and rank these disparities. Evidence-based prioritization requires a composite index that is reproducible and spatially explicit.
Problem Statement
Which GHMC wards are the most and least equitable across access, opportunity, environment, and governance dimensions? How can indicator data from disparate GeoJSON sources be standardized, weighted without subjective bias, and presented in an interactive dashboard for planners?
Methodology
Point-in-polygon counts computed for each GHMC ward across 4 domains: ACCESS (affordable schools, government hospitals + PHCs, bus stops + metro + MMTS stations), OPPORTUNITY (commercial density, Fair Price Shops), ENVIRONMENT (parks, noise pollution), GOVERNANCE (Area Sabhas, Ward Committees). Indicators normalized using Min-Max scaling per ward area. Entropy weighting applied within each domain to derive objective, data-driven weights. Domain scores aggregated into a composite UEI as a simple average.
Analysis
Data engine built in Python (GeoPandas, Pandas, NumPy, Shapely). Spatial analytics: Global Moran's I for spatial autocorrelation, Z-score hotspot/coldspot detection, PCA dimensionality reduction + K-Means clustering to classify wards into 4 typologies (A–D). Backend: FastAPI + PostgreSQL/PostGIS serving ward scores via REST API. Frontend: Next.js + Mapbox GL JS for ward-level choropleth and Recharts for domain breakdowns. Fully Dockerized with docker-compose.
Insights
Entropy weighting revealed transit access as the highest-variance indicator, dominating ACCESS domain scores. Ward typology clustering identified four distinct groups: high-equity central wards (Type A), transit-rich but green-poor inner wards (Type B), opportunity-poor peripheral wards (Type C), and compound-disadvantage outer wards (Type D). Governance indicators (Area Sabha density) showed the weakest spatial clustering, suggesting more uniform distribution than service access.
Outcome
Deployed full-stack platform with interactive ward-level UEI dashboard, hotspot/coldspot maps, typology clustering visualization, and GeoJSON/CSV export. Data engine is modular — each indicator is a separate computation step that can be updated independently as new data becomes available.