Explainable AI Model for Predictive Maintenance in Smart Agricultural Environments

Smart Agricultural Facilities (SAF) Artificial Intelligence (AI) often suffers from obscureness, which prevents farmers from making the most of its benefits. To fill this void, the authors of the present study proposed a model that combines PdM with eXplainable Artificial Intelligence (XAI). Data, model, outcome, and end-user are the four cornerstones that the model aspires to provide anticipated insights and explanations for. This approach reshapes how these technologies are understood and used, marking a paradigm shift in agricultural AI. With XGBoost, the study achieves an accuracy boost of up to 10.54% and with Bidirectional--Long - Short-Term-Memory (LSTM), an increase of 6.81%. Compared to previous approaches, the study performs much better. In addition to shedding light on data quality and explainable methodologies, the study highlights the significance of explainability in PdM in SAF, going beyond conventional accuracy measurements and hypothetical situations. It promotes more study into HITL systems and multimodal data to resolve ethical concerns in agricultural AI, such as openness, responsibility, and justice.