Data analysis to identify operational patterns at a port terminal in Guayaquil: A case study (2024–2025).

Abstract

Operational congestion at the Port of Guayaquil, caused by the increase in vehicular flow, represents a challenge to logistical efficiency. This study aimed to identify patterns of operational behavior that contribute to improving port traffic management through data analysis techniques. The CRISP-DM model was applied to 29,067 records processed using the K-Means and Expectation Maximization (EM) clustering algorithms in WEKA, in order to segment the data and characterize different operational patterns. K-Means produced 2 groups (45%, 55%) and 3 groups (35%, 40%, 25%) that segment operations at a macro level, whereas EM identified 6 groups, with 2 dominant ones (56%, 25%) and 4 minority groups (5%, 7%, 3%, 4%) classified as critical. The data were integrated into business intelligence dashboards in Google Looker Studio to visualize congestion patterns and support resource planning. It was determined that the EM algorithm better captured the complexity of the data by identifying 6 groups with distinct characteristics. These operational data, transformed into useful analytical information, enabled improvements in logistics and evidence-based decision-making.