AI platform lets farmers name crop threats instead of showing photos

A new Ghanaian AI tool known as Ndua Analytics helps farmers name crop threats simply by describing symptoms. No photos are needed – farmers type words such as yellow spots on leaves or wilted stems. The system created by Nana Benyin Eboe Nyamekye Essel‑Biney of the University of Ghana gives a diagnosis and then offers treatment tips. It speaks in plain English and also in local languages so even those who cannot read English can understand. Speech options let users hear the advice aloud. Farmers can try the platform now on maize, cassava and tomato crops in a pilot run by the researcher. Ndua works with low quality or shaky observations because image analysis often fails under bad lighting or movement. The technology also overcomes problems caused by limited local language data sets since translation tools help the model learn. Experts say early detection matters more as changing weather brings new pest attacks such as armyworms that may appear at unexpected times. Professor Benjamin Lamptey of Leeds University stresses that AI can spot pest forecasts but only if that knowledge reaches farmers, extension officers and policy makers together. He urges the country to invest not just in young researchers but also in mid‑career scientists who can turn findings into real solutions. Nana Benyin hopes to add rice, ginger, garlic and regional maps as the platform grows. New language options and a login system are planned so more people can join without a simple link. The tool aims to close the gap between what a farmer sees in the field and the knowledge needed to protect a harvest before a small problem causes major loss.