Decision support system for Western Flower Thrips management in roses production

The objective of this study is to develop an innovative system to assess the risk of pests using a fuzzy logic approach. The system is designed to provide farmers with an index representing an estimate of the risk of presence of Western Flower Thrips (WFT), Frankliniella occidentalis in a roses greenhouse. For this purpose, a modular knowledge-based decision support system has been designed. The major findings of our research are summarized in four points. First of all, the model is based on variables measured automatically via sensors and do not require human activity (damaged area of a leaf, sex ratio). Secondly, as the system is not only oriented toward experimentation and research centers but also farmers, the phenomenon of manual counting could be replaced by a predicted value. In addition, the novelty associated with the system is that it supplies a daily rather than a weekly estimate of WFT risk level. In so doing, the farmers could stay aware of the influence of daily weather conditions on its evolution. Finally, this study could be beneficial to help reduce the utilization of pesticides and decrease the percentage of production loss, due to continuous monitoring of the risk level in the greenhouse. Because the development of F. occidentalis is highly sensitive to climate change, and in order to enhance the assessment of pest risk, an approach, which combines data related to the type of rose, the duration of sunlight and meteorological conditions, was followed. Simulation results are displayed at the end to validate our approach.