This contribution introduces a data acquisition and modelling framework for the prediction of banana pests’ incidence. An IoT sensors-based system collects weather and micro-climate variables, such as temperature, relative humidity, and wind speed, which are uploaded in real time to a cloud storage space. The incidence of the red rust thrips (Chaetanaphothrips signipennis) is collected “manually” by periodic inspection. The mathematical model is adapted from population growth functions and a model of insect species development and allows predictions to be made at various time intervals with an accuracy greater than 80%, improving decision-making capacity for agro-producers and enabling the improvement of pest management actions.