Geowrangler provides scientists with consistent access to high-quality humanitarian geospatial data and the tools to analyze it
Every data scientist knows the frustration of spending countless hours cleaning and preparing messy datasets—a task that consumes up to 80 per cent of project time and derails many promising initiatives. UNICEF’s East Asia Frontier Data Lab (FDL), in partnership with Thinking Machines, tackled this challenge head-on with Geowrangler. This innovative, open-source tool streamlines geospatial data workflows, transforming a major bottleneck in data science into a manageable challenge for development and humanitarian projects worldwide.
Some of the biggest advancements in social impact data science are practical tools that improve efficiency across projects. Geowrangler exemplifies this by providing data scientists consistent access to high-quality humanitarian geodata, along with the tools to analyze it—capabilities once reserved for big-tech companies like Uber, Google, and Meta.
Humanitarian data science often stalls in the ideation phase due to the overwhelming effort required to wrangle messy datasets into a usable form. To confront this persistent challenge, FDL and Thinking Machines— a Philippines-based company supported by the UNICEF Venture Fund—introduced Geowrangler under the AI for Development (AI4D) initiative.
Geowrangler simplifies the process of preparing geospatial data—location-based information—for analysis, making it quick and easy for users. It automates key tasks like downloading, preprocessing, and extracting features from both vector data (e.g., OpenStreetMap points of interest) and raster data (e.g., nighttime satellite imagery). By reducing time spent on these logistical hurdles, analysts can focus on more advanced tasks, such as statistical modeling and policy formulation. In practice, this means data scientists are better equipped to prepare geospatial data for analysis quickly and accurately, accelerating their ability to generate valuable insights.
Applications in Humanitarian Work
Geowrangler is already delivering significant results. The tool enabled the Frontier Data Lab to estimate internet connectivity for every school and health centre across 28 countries in the East Asia Pacific region. In Cambodia, UNICEF and the National Institute of Statistics used it to produce high-resolution estimates of zero-dose children, directly supporting vaccination campaigns.
It has also played an important role in relative wealth mapping. By combining it with Ookla’s internet speed data, UNICEF generated insights into wealth and poverty levels across nine Southeast Asian countries. These findings were visualized through OpenConnect, a platform that provides real-time connectivity estimates for schools and health facilities, helping policymakers identify areas most in need of digital inclusion.
In Thailand, Geowrangler has been instrumental in air quality monitoring where it was integrated with machine learning models to predict PM2.5 — particle pollution — levels. There, it used ground sensor data and satellite imagery to monitor pollution at a one-square-kilometre resolution. In Lao PDR, it was used to identify schools with sufficient internet connectivity for deploying low-cost air quality sensors.
“Geowrangler has become our go-to tool for gaining insights from big geospatial data,” said Anthony Mockler, Data Scientist, UNICEF EAPRO. “From estimating school internet connectivity to understanding the relationship between climate and child health, Geowrangler makes the process easier, faster, and more accessible.”
Driving Open-Source Innovation
As an open-source software, Geowrangler is constantly evolving. Its codebase can be modified to meet different user needs, and the Frontier Data Lab has already used it to improve the speed of aggregating disparate satellite-derived datasets. A growing community of users is contributing to its development, helping expand its capabilities and enrich the broader open-source ecosystem for other humanitarian data science practitioners.
Geowrangler has emerged as a vital tool for improving the efficiency of data-driven development projects across the region. As new applications and use cases continue to emerge, it will play an even larger role in transforming geospatial data into actionable insights that drive global development.
Learn More:
- Geowrangler: Explore the open-source codebase on GitHub.
- UNICEF Venture Fund: Supporting open-source solutions for children.
- UNICEF Frontier Data Lab: Driving innovation through data-driven solutions.