Metrics development for the visualisation and prediction of material delivery schedule variations in supply chains

Journal article, 2026

The study proposes metrics for visualising and predicting delivery schedule variations in supply chains. This includes exploring patterns of schedule variations and accuracies and how intra-organisational features explain schedule variations in a predictive forecasting model of future schedule volumes. We employ quantitative analysis based on multiple-year delivery schedule data from four European automotive industry suppliers. The study proposes the MAPE profile and predictive volume metrics to complement established metrics in assessing and interpreting delivery schedule variations. The proposed metrics provide descriptions of schedule variations and change/dynamics of schedule accuracy, as well as prediction of future schedule volumes using objective data transactions and master data as features. Our research contributes to the forecasting literature by adapting forecast metrics to the delivery schedule context and assessing features in predictive forecasting using machine learning, and initiates a discussion about the metrics mechanism role in managing and absorbing supply chain complexity and contributing delivery schedule utility.