Google estimates African solar partnership cut over 160,000 tonnes of CO2 equivalent
The company’s update with Ignite Energy Access connects satellite mapping and delivery planning to solar deployment. Detecting tiny installations from orbit remains research in progress.
Google and Ignite Energy Access are applying mapping, route-planning and supply-chain optimization to the challenge of delivering solar equipment to households without formal addresses. Their work also includes computer-vision detection of tiny rooftop panels and models for community microgrids that could support electric cooking, though both efforts remain in development. The partnership’s reported climate result is an estimate of more than 160,000 tonnes of CO2e avoided in the first half of 2026; that figure covers a different period from the 3.7 million devices delivered since 2023.
01
Satellite imagery, Google’s Open Buildings dataset and routing algorithms help locate homes, while geographic models add rural paths missing from standard maps.
02
Operations research models help optimize warehouse and hub locations, delivery routes, vehicle capacity, transport costs and inventory.
03
A typical rooftop solar panel can occupy barely one pixel in satellite imagery; models trained with Ignite drone images are being tested across thousands of code variations.
Off-grid solar delivery in sub-Saharan Africa helped avoid more than an estimated 160,000 tonnes of carbon dioxide equivalent in the first half of 2026, Google said on October 26. The company attributed that result to its three-year collaboration with Ignite Energy Access, which uses mapping and logistics tools to reach remote households.
Google linked the estimated reductions to 3.7 million solar devices delivered across the region since 2023. Those figures cover different periods: the delivery total spans several years, while the emissions estimate covers six months. The collaboration operates through Google’s Startups for Sustainable Development program and has involved more than 100 advisory sessions with researchers, engineers and data specialists.
The work addresses an electricity-access gap affecting more than 600 million people in sub-Saharan Africa, according to Google. Families without power rely on kerosene and candles for lighting. Getting solar equipment to them requires finding homes without formal addresses and navigating rural paths missing from standard maps.
Finding homes, then getting equipment there
Google and Ignite combined satellite imagery, Google’s Open Buildings dataset and routing algorithms to locate remote households. Finding a building, however, is only the first step. Google researchers are also developing geographic models that identify rural paths and incorporate them into delivery planning, accounting for the routes field teams can actually use across difficult terrain.
The planning extends beyond directions to individual homes. Google says its operations research models—mathematical tools for weighing logistical choices—help optimize Ignite’s supply chain. The system evaluates possible delivery networks and balances several practical constraints:
Storage locations: where to place central warehouses and smaller local hubs.
Transport: how delivery routes, transportation costs and vehicle capacities fit together.
Inventory: how the cost of holding equipment affects the network’s overall configuration.
A rooftop panel can be barely one pixel
A different mapping problem begins after solar equipment reaches a village: identifying where systems are already operating. Google says a typical micro-solar installation consists of one rooftop panel, which appears as barely one pixel in standard satellite imagery. That makes visual identification difficult at scale, even though an installation map could help teams identify underserved villages and plan future deployments.
Researchers are developing computer vision models—software that interprets images—using Gemini for Science. Ignite supplies drone imagery of actual installations to fine-tune the models against known examples. Google says the team is testing thousands of code variations to improve both how reliably it identifies panels and how many installations it finds. This detection work remains under development.
Cooking needs more than a lighting kit
The partners’ next area of work moves beyond lighting. Google and Ignite are modeling community-level microgrids intended to supply enough electricity for clean cooking. These shared local power systems are still a modeling effort, rather than a reported rollout of electric cooking infrastructure.
Electric cookstoves require far more power than an individual rooftop solar kit can generate. Google describes that capacity gap alongside the health and time costs of cooking with wood and charcoal: smoke inside homes and hours spent collecting fuel, a burden falling disproportionately on women and girls. The microgrid work aims to support cooking as well as household lighting.
Sources
blog.googlePowering sub-Saharan Africa beyond the grid
Reader comments
Newest comments first. Replies stay oldest first.