Google has added millions of previously unmapped buildings across Africa to its Maps service, following a major artificial intelligence initiative aimed at closing digital infrastructure gaps on the continent.
Programme Manager for Google Research, Abdoulaye Diack, said the project involved using satellite imagery and machine learning models to detect structures in underserved and remote areas, many of which had never been formally documented.
‘When we started, we noticed that many neighbourhoods in Africa weren’t even visible in Google Maps. Our project helped fix that’, Diack told The PUNCH. ‘Almost half of the buildings you see on Google Maps in Africa came from our initiative’.
According to him, the team used European Sentinel satellite data, which is updated every five days, to train models capable of identifying building footprints. Over time, the project mapped more than 150 million buildings across the continent, including informal settlements and rural dwellings.
He added that the model went beyond identifying rooftops and was trained to estimate building heights, using a method that analyses the shadow length of structures along with the time of day the satellite image was taken.
‘We were able to detect height for over 1.8 billion buildings globally. That wasn’t the original goal, but it was one of the most fascinating outcomes’, he said, adding that the data is now aiding urban planners, emergency responders, and economists in understanding the physical layout of African cities.
The initiative, which began in Ghana, has since been scaled globally, with AI tools developed for Africa now being applied to other parts of the world.
Diack noted that a key driver of innovation on the continent is data scarcity, which has forced teams to think creatively: ‘Because high-resolution images were lacking, we had to use lower-quality satellite data, which pushed us to innovate on model efficiency’.
Beyond infrastructure mapping, the Google Research team has been involved in projects tackling local challenges through AI. One of the earliest, known as PlantVillage, helps farmers detect diseases in cassava plants using a smartphone camera.
The project, which began in collaboration with Makerere University in Uganda, involved collecting thousands of images of healthy and diseased leaves to train the model. ‘The AI can detect issues before they’re visible to the human eye’, Diack said.
He described data collection as a simple but critical process, often requiring on-the-ground efforts: ‘We send people to farms to take pictures of plants, and over time, the model learns by comparing patterns’.
In the area of language, Diack said Google is investing in building datasets for under-represented African languages, partnering with organisations like Masakhane and universities across the continent. He explained that for speech recognition, participants are asked to record themselves in their native languages or dialects, helping models understand both local languages and regional accents.
According to him, a major part of Google’s work in Africa also involves AI talent development. The company previously ran an AI residency programme, which took in early-career researchers and trained them over 18 months.
‘People with mathematics and physics backgrounds tend to do well’, Diack said. ‘We work with institutions like the African Institute for Mathematical Sciences and others to identify and support talent’.
He added that Google has also partnered with groups like AfricaToML to prepare software engineers with the right skills for AI careers, particularly in countries like Nigeria.
On what Africa needs to ensure full adoption of AI, Diack said local engagement and data accessibility are essential.
‘We need more grassroots efforts that train and build communities. And we must invest in datasets that reflect African languages and realities. If AI tools can’t understand you, they can’t serve you’, he said.
