Google has successfully added millions of previously unmapped buildings across Africa to its Maps service, leveraging a major artificial intelligence initiative aimed at addressing digital infrastructure gaps on the continent.
Abdoulaye Diack, Programme Manager for Google Research, stated that the project utilized satellite imagery and machine learning models to identify structures in underserved and remote areas. “Almost half of the buildings you see on Google Maps in Africa came from our initiative,” Diack explained.
The team employed European Sentinel satellite data, updated every five days, to train models capable of detecting building footprints, resulting in the mapping of over 150 million buildings, including informal settlements and rural dwellings. The AI models were also able to estimate building heights by analyzing shadow lengths from satellite images.
This initiative began in Ghana and has since expanded globally, showcasing the innovative responses to data scarcity in African regions. Diack emphasized the importance of local engagement and data accessibility for the successful adoption of AI technologies across the continent.
In addition to infrastructure mapping, Google is also focused on other local challenges, such as helping farmers detect diseases in crops through initiatives like PlantVillage. The company is actively investing in AI talent development and building datasets for under-represented African languages to enhance understanding and representation.
Diack concluded by highlighting the need for grassroots efforts to train communities and ensure that AI tools reflect African languages and realities.
