Applying Entropy Analysis to Optimize Cargo Dwelling Time at Container Terminals

This study applies entropy analysis to examine the quality, variability, and operational consistency of event log data documenting cargo dwelling time at a container terminal. In this context, dwelling time refers to the total duration between a container’s arrival and its departure from the terminal, serving as a critical performance indicator in port logistics. To quantify process unpredictability, Shannon entropy was employed as the primary measure. The research analyzed a dataset containing 952,069 records, from which a systematic 10% sample of 95,207 entries was drawn to facilitate computation. While this sampling approach enhanced efficiency, it also introduced certain limitations in capturing less frequent operational scenarios.Entropy calculations were performed using frequency-probability distributions derived from the event logs. The analysis yielded a Shannon entropy value of 5.18 bits, indicating significant variability and a lack of consistent patterns in container handling processes. This level of uncertainty suggests that dwelling time is highly unpredictable, which can complicate planning, resource allocation, and performance monitoring efforts. To address these challenges, the study recommends incorporating additional entropy measures, such as Renyi and Tsallis entropy, to deepen understanding of process complexity.Moreover, specific preprocessing strategies—including event name standardization, filtering of non-relevant records, and aggregation of low-frequency events—are proposed to reduce noise and improve data reliability. Overall, this research underscores the practical relevance of entropy-based metrics for evaluating operational performance in container terminals and highlights their potential to support more informed decision-making in global supply chains.