Mobile Interfaced Crops Diagnosis Expert System (MICDES): a case for rural Kenyan farmers

Abstract

This paper describes a prototype case-based reasoning expert system that enables farmers to get assistance with crop diagnostics and other extension services through the web via a dynamic website or through mobile phones. It is also voice enabled to support visually impaired people. Mobile Interfaced Crops Diagnosis Expert System (MICDES) was motivated by the lack of extension services to farmers in the rural areas of Kenya, a situation compounded by poor communication infrastructure. The system conceptualisation, design and development were based on contextual inquiry putting the user and culture at the centre. The design, implementation and testing was done based on studies carried out on two rural agricultural self-help groups and schools in Kenya. The diagnoses currently performed are major tropical diseases and pests affecting rice, tomatoes, bananas and sweet potatoes. The system automatically alerts a human expert via SMS in cases where it fails to diagnose a case presented by the user.