Fall armyworm (FAW) early detection system in Tanzania: Leveraging data and artificial intelligence to save livelihoods in Africa

Abstract

This project will leverage agronomic data, real-time detection and artificial intelligence to create an early warning system for Fall armyworm control in Tanzania.

Description

Fall armyworm (FAW), Spodoptera frugiperda (J. E. Smith) (Lepidoptera: Noctuidae), is an insect pest native to tropical and subtropical regions of the Americas. The species is a globally important economic agricultural pest and is considered to be one of the most highly destructive herbivorous insects in agro-ecological environments. While FAW is a generalist herbivore, known to feed on over 353 plant species, it is known to impose particularly high damage in maize, a staple crop globally .

This study aims to identify risk factors associated with FAW outbreaks and to create a model of spatially explicit risk.  There is a need to increase the resolution of monitoring for this pest in time and space, and to this end a "smart" moth trap using deep learning detection and classification of FAW will be developed incorporating artificial intelligence and data communication networks to alert farmers of risk in real-time.