Urban Mapping for Resilience: Where It Started
Spatial Collective / World Bank / 2020 to 2024
This is where my journey started — from field-level mapper to building the GIS validation tools used by the entire team. I helped digitize 47,000+ building footprints, capture 6,000+ Street View images, and build the data quality pipeline that made all of it trustworthy enough for government use.
Situation & Task
Nairobi's population grew tenfold in 50 years, and informal settlements expanded along with it. The government simply didn't have up-to-date spatial data to plan infrastructure or respond to disasters in these areas. I started at Spatial Collective as a field mapper in 2020, working on the ground in the Nairobi region, collecting and validating geographic data using HOTOSM tools and Java-based validation scripts. I saw firsthand what bad data looked like — and what it cost communities when planning decisions were based on it.
Objective: Over time my role evolved from field data collection to building the technical tools that ensured data quality. I was responsible for developing geographic data validation pipelines, transitioning from field-level mapping to creating the integrity checking tools used by the entire mapping team. When the World Bank's KISIP II pilot launched in 2022, I supported the technical deployment of the mapping management platform across three settlements.
Action: Technical Solution
I built GIS data validation tools using Java and HOTOSM frameworks that caught geometry errors, attribute gaps, and satellite cross-reference mismatches before data reached the production database. I worked with JOSM for building digitization, ODK Central and KoboToolbox for mobile surveys, and helped design a microtasking consensus pipeline for AI-based data validation. I set up temporary technology centres in each settlement with desktop computers and internet access, and provided on-site technical support to mappers. The quality assurance system I helped design used a tiered pay model — base pay, quality bonus, and overtime — to incentivize accurate data from 282 youth workers.
System Architecture
Result: Measured Impact
47,000+ building footprints were digitized and contributed to OpenStreetMap across three Nairobi settlements. 6,000+ terrestrial Street View images were captured, providing ground-level coverage where none existed before. The data validation pipeline I built caught errors that would have made the datasets unusable for urban planning. The resulting GIS layers now support government decisions in solid waste management, disaster risk, and infrastructure investment. The World Bank has referenced these datasets in its urban resilience assessments, and the pilot model I helped prove is now being replicated in Nakuru and Kisumu counties.
47,000+
Buildings Digitized
6,000+
Street View Images
3 settlements
Settlements Covered
47K+ features
OSM Contributions