AI predicts African staple crop yields
A new tool, which uses cutting-edge machine learning techniques and satellite remotely sensed data predicts agricultural yields for nine key crops across Africa, the pan-African non-profit research organisation AKADEMIYA2063 announced in May 2023.
The web-based Africa Agriculture Watch (AAgWa) platform has provided predictions for agricultural yields in 47 African countries, across many of the most important crops for African food production, including maize, cassava and sorghum. The tool is designed to support farmers, policymakers and local communities in crisis management, monitoring and mitigation.
This new phase of the AAgWa program will support African countries in using emerging technologies like AI and advanced remote sensing in order to achieve their development objectives and broader economic growth, including the African Union’s (AU) Agenda 2063 and Digital Transformation Strategy for Africa (2020-2030).
“As we know, the African agricultural sector is facing a number of threats – from supply chain disruptions to extreme climate events and health crises. What all these crises have in common is the need for good planning and preparedness to reduce uncertainties in decision-making,” said Dr. Racine Ly, Director of the Department of Data Management, Digital Products and Technology at AKADEMIYA2063.
“Relying on conventional analytic techniques alone will not deliver the effective decision-making we need to meet these challenges. At AAgWa, we are improving decision-makers’ access to the data they need in order to take better actions”.
Harnessing the power of emerging technologies will prove vital in improving the productivity and resilience of African food production systems to future shocks, AKADEMIYA2063 states. This is particularly vital given that African agricultural yields, on average, are still just one-fifth the size of yields in the US, even though the population of sub-Saharan Africa is expected to double by 2050
Visit the Africa Agriculture Watch (AAgWa) platform