Grid-friendly charging: How e-cars help with the energy transition
24. October 2023
7.11. | Sustainable Development as a Task for Humanity
7. November 2023
Grid-friendly charging: How e-cars help with the energy transition
24. October 2023
7.11. | Sustainable Development as a Task for Humanity
7. November 2023

AI4EA – Artificial Intelligence for Energy Access

The team of the "Artificial Intelligence for Energy Access" (AI4EA) project creates a high-resolution map of the electricity demand profiles of Niger, Nigeria, Benin, Togo and Ghana. To do this, the researchers use extensive data sets on electricity demand in the West African Countries where the data is processed with the latest methods of artificial intelligence. Rural electrification agencies, ministries and local communities receive the resulting map with the intention to fill the electricity demand data gap in rural areas.

Estimating power needs

Reliable and sustainable access to electricity is important to have functioning communities, businesses and industries. Especially now, since various climate adaptation measures envisage the conversion to electricity as the primary energy source in the future. To achieve reliable electricity access, however, the electricity requirement must be estimable. This is the only way to ensure, plan and implement supply-side decisions, investments and processes.

Closing the data gap

However, the biggest challenge in many countries is the lack of data on current demand profiles: How are communities using electricity today? What level of consumption do they have over time? For what purposes is the electricity used? These questions make it difficult to forecast future demand profiles or structural changes. Estimates could provide answers: Realistic electricity demand estimates are derived from local influencing factors such as household or company wealth, demographics, socio-economic factors, climate events and other dynamics. In addition, conclusions can be transferred from environments where data is available to environments where data is scarce.

Model training for reliable predictions

Data in the AI4EA project was collected by RLI in 2021 in three geopolitical zones of Nigeria, in parallel with remote sensing data from Atlas AI and regional datasets collected by WASCAL for the focus countries. The researchers use these data sets for model training, validation and ground truthing to ensure that the models are robust and the forecasts reliable. The project partners in the AI4EA project are developing a standardized electricity demand profile for typical households and small and medium-sized companies in Nigeria, Niger, Ghana, Benin and Togo. The resulting demand forecasts and simulation methods are made openly accessible for broad use. In the AI4EA project, the Lacuna Fund, the Reiner Lemoine Institute (Germany), Atlas AI P.B.C. (USA), Université Abdou Moumouni (Niger) and the Clean Technology Hub (Nigeria).

Project duration: October 2023 - April 2025

The RLI assumes the following tasks:

  • Leading partner in the project
  • Coordinating the collection of relevant datasets
  • Supporting the development of machine learning model that predicts electricity-related socio-economic aspects beyond surveyed areas.
  • Stochastic modeling of demand profiles
  • Arranging the publication details of outputs
  • Designing the workshops to involve local stakeholders


Catherina Cader

Project leader

WordPress Appliance - Powered by TurnKey Linux