A novel approach based on IoT and log-normal distribution for supplier lead time optimization in smart engineer-to-order supply chains

Background: In Engineer-to-Order (EtO) supply chains, managing supplier lead times is particularly challenging due to high customization and intensive customer involvement. This study addresses the critical need for more accurate and dynamic lead time prediction to enhance supply chain resilience and efficiency in EtO environments. Methods: We propose a novel approach that integrates Internet of Things (IoT) technologies with statistical modeling using the log-normal distribution to model and optimize supplier lead times, especially for customized raw materials. The model incorporates real-time data from IoT-enabled suppliers and considers long-term contractual relationships to reduce variability. Monte Carlo simulation is employed to validate the model's predictive capabilities. Results: The results demonstrate significant improvements in predicting supplier performance and reducing uncertainty. Simulation outputs reveal reductions in lead times and enhanced reliability. Statistical metrics such as the Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) confirm the robustness and accuracy of the predictions. Conclusions: The proposed methodology supports better decision-making in supplier selection and procurement planning by enabling effective risk management. It contributes to improved performance and greater resilience in Engineer-to-Order supply chains.