Edge AI for Agricultural Advisory Challenge

Over 300 million African smallholder farmers make critical decisions with no access to agricultural expertise, losing up to 40% of crops to preventable pests, diseases, and poor timing while extension services reach less than 10% of farmers. Farmers plant based on tradition rather than weather forecasts, cannot identify diseases until crops are destroyed, apply fertilizers blindly without soil knowledge, and sell at harvest when prices are lowest. A maize farmer in Malawi loses entire fields to fall armyworm undetected, a tomato grower in Tanzania cannot identify bacterial wilt until too late, and cassava farmers across West Africa lack early warning for mosaic virus. Developers must create edge AI systems that provide farming advice through weather analysis, pest/disease identification from images, soil management guidance, and market intelligence, all working offline on basic smartphones to transform every phone into an agricultural extension officer.

The Edge AI for Agricultural Advisory Challenge seeks innovative solutions that bring precision agriculture capabilities to smallholder farmers through AI that works entirely on basic smartphones. This challenge addresses the critical need for timely, localized agricultural advice that can improve yields and reduce losses for the 60% of Africans who depend on farming.

Participants will develop on-device AI systems that provide comprehensive farming advice by analyzing weather patterns, soil conditions, and crop images without requiring internet connectivity. The system must identify pests and diseases from photos, recommend planting times based on local weather data, suggest fertilizer applications based on soil and crop conditions, and provide market price information when occasionally connected.

Successful solutions will implement computer vision for disease detection, weather prediction models, soil analysis algorithms, and decision support systems optimized for edge deployment. The system should understand local farming practices, provide advice in local languages, consider resource constraints of smallholder farmers, and learn from collective farmer experiences when synchronized.

We particularly value solutions that incorporate indigenous farming knowledge, support diverse African crops often ignored by global systems, provide early warning for pest outbreaks and weather extremes, and enable farmer-to-farmer knowledge sharing. The platform should help farmers increase yields, reduce losses, adapt to climate change, and access markets, transforming subsistence farming into profitable agriculture.