Case study · NUST SEECS · UKRI
Pakistan's
climate, live.
S2Cool (Super-efficient Sustainable Cooling Solution for All Applications) is a UKRI-sponsored research project developed by NUST's MachVIS lab with Northumbria University. As a full-stack developer at NUST SEECS, I built its public climate platform. I scraped weather data from across Pakistan, turned it into formulas and put it on a live map.
- Client
- NUST SEECS · MachVIS Lab
- Sponsor
- UKRI, with Northumbria University
- Role
- Full-stack developer & data engineer
- Stack
- Next.js · React · Selenium · Vercel
Live site audit · 2026-09-29
Measured on the live site with Google Lighthouse (desktop) and a crawl of its sitemap and structured data.
Extracted and structured from sources across Pakistan with Selenium pipelines.
Temperature, wind speed, humidity and dew point.
From Gilgit and Peshawar down to Karachi.
Readings and layers update every hour.
UKRI, NUST, Northumbria University and MachVIS.
Scraping, data modelling, formulas, front end and deployment.
01 · The data
Scraped from every corner.
Good cooling research needs good climate data, and Pakistan's is scattered across websites, PDFs and scanned reports. I engineered Selenium scraping pipelines that collected and structured 15 years of historical weather data from across the country. I also built PDF parsers and image-to-text converters for the reports that only existed as documents.
02 · The formulas
From raw numbers to insight.
I turned the raw readings into formulas: long-term means, standard deviations, anomalies and derived parameters that show how each region's climate behaves. Those formulas drive the map layers, so researchers see patterns rather than spreadsheets.
03 · The platform
A live map,
deployed on Vercel.
Live parameter layers
Switch between temperature, wind speed, humidity and dew point, each with its own heatmap and legend.
City stations
Pins across Pakistan show the current reading for each city at a glance.
Hourly auto-refresh
Data and layers refresh every hour, with a visible last-updated time.
Regional analytics hub
A dedicated Regional Analytics Hub for exploring the data region by region.
Light & dark themes
A research-friendly interface with theme switching and a centred map reset.
Next.js on Vercel
A React / Next.js front end deployed on Vercel for fast, global delivery.
many countries →