AI tool development for cocoa disease diagnosis

Researchers from the State University of Santa Cruz (UESC) in Brazil, in collaboration with partners from Peru and the Netherlands, have developed an artificial intelligence-powered application to assist cocoa farmers. The tool, introduced during the ExpoCacau 2026 event in Ilhéus, Bahia, allows producers to submit text, audio, photos, and videos to receive preliminary diagnoses of pests, diseases, and nutritional deficiencies.

A primary function of the application is the detection of major threats such as ‘vassoura-de-bruxa’ (witches' broom) and moniliasis. The project aims to provide immediate guidance to small-scale and family farmers who may experience delays in receiving professional technical assistance, thereby reducing potential financial losses. While the tool offers rapid identification of symptoms, developers noted it is not intended to replace professional agronomists.

The launch occurred alongside discussions on sustainable development and resilient production systems, involving initiatives such as the ‘Parceiros da Mata’ and ‘Cacau+’ projects to better integrate family farming into the cocoa production chain.