#SDGInsights challenge winner uses machine learning and geodata to identify water services

Water scarcity affects more than 40 per cent of the world’s population. Ensuring universal access to safe and affordable drinking water for all by 2030 is an essential element of Sustainable Development Goal 6. Locating communities and individuals still using unimproved water services in a country is an important step in designing drinking water infrastructure to deliver safe and sustainable water services to those that need it.

The United Nations announced this week that a team from the University of Southampton, United Kingdom, has won the Unite Ideas #SDGInsights challenge for its methodology providing a framework to combine multiple data sources based on geographic location, and to apply advanced machine learning algorithms to the data to discover previously hidden patterns.  This will go a long way to improving the tools available for designing sustainable development policies.

The winning team includes Mr. Weiyu Yu, a postgraduate research student within the Geography and Environment department; Dr. Jim A. Wright, Associate Professor in Geographic Information Systems; and Dr. Nicola A. Wardrop, Visiting Research Fellow within Geography and Environment. The University of Southampton is a public research university located in Southampton. The team applied its methodology to identify drinking water service in Liberia.

“Currently, disadvantaged drinking water services are often not reported [with precision],” said Mr. Yu. “However, as new datasets become available, predicting the potential spatial distribution of drinking water becomes possible. Our predictive model can fill the gaps of spatial coverage, give an indication of the potential distribution of specific water supplies and provide valuable policy insights.”

#SDGInsights was launched by the Office of Information and Communications Technology in collaboration with UN DESA’s Development Policy and Analysis Division (DPAD) and the United States Agency for International Development’s Demographic and Health Surveys. The challenge asked participants to use the Surveys’ data and other georeferenced datasets to better visualize progress on the Sustainable Development Goals, especially regarding health and family-related indicators.

The advisers for the #SDGInsights challenge noted that the winning solution matched various datasets with ease, and combined them in meaningful and innovative ways. The judges also recognized Dwitipriya Sanyal for her submission on “Service Gap Identification”, as well as David Jeronimo Giraldo Atehortua for his solution called “World Data Viewer”.

“The submissions for the Unite Ideas #SDGInsights challenge demonstrate the potential of using open-source technology to exchange ideas and identify paradigm‑shifting solutions to critical global issues,” said Mr. Salem Avan, Director of the Global Services Division in the Office of Information and Communications Technology.

For those interested in contributing to building open-source tools for achieving the Sustainable Development Goals globally, they can create an account on to receive notifications of upcoming challenges.

Follow Us