Optimalisasi Strategi Monitoring Pengiriman Berbasis Fuzzy C-Means Clustering (FCM) pada Logistik Perdagangan Umum di Pulau Jawa: Optimalisasi Strategi Monitoring Pengiriman Berbasis Fuzzy C-Means Clustering (FCM) pada Logistik Perdagangan Umum di Pulau Jawa

Mohamad Iqbal (Department of Industrial Engineering, Faculty of Industrial Technology, Universitas Islam Indonesia)
Muhammad Ridwan Andi Purnomo (Department of Industrial Engineering, Faculty of Industrial Technology, Universitas Islam Indonesia)

Journal Science Innovation and Technology (SINTECH)
Vol. 6 No. 2 (2026): Vol. 5 No 2 (2026): SINTECH JURNAL BULAN MEI 2026

The rapid expansion of Indonesia’s logistics sector, particularly within the e-commerce domain, necessitates adaptive and data-driven shipment monitoring strategies. Operational assessments in one logistics case reveal two critical negative symptoms: customer complaints resulting from discrepancies between the Estimate Time Arrival (ETA) and Actual Time Arrival (ATA), and the presence of excessive or out-of-budget distribution costs on several delivery routes. Addressing these issues requires analytical approaches capable of capturing complex shipment behavior across diverse geographic areas. This study aims to optimize shipment monitoring strategies using the Fuzzy C-Means (FCM) clustering method applied to operational data from the general trade logistics sector on Java Island. A total of 1,076 shipment records were analyzed using seven key variables: Sub Area, Actual Time Delivery (ATD), Actual Time Arrival (ATA), Colly, CBM, and Weight. FCM clustering produced five distinct distribution clusters with varying operational characteristics. Cluster 3 demonstrated the highest efficiency, exhibiting the largest shipment volume (Colly = 1,690.86; CBM = 24.86; Weight = 18,400.65 kg) and the fastest ATD (0.78 hours), indicating strong performance in dense and accessible regions. In contrast, Clusters 1 and 4 recorded the highest Sub Area values (9.25 and 9.04) and slower ATD durations (>2.5 hours), highlighting significant logistical challenges in remote areas prone to delays and elevated costs. Membership analysis further showed that distribution points predominantly aligned with a single cluster while retaining cross-cluster behavioral flexibility. These findings offer a foundation for data-driven operational strategies, including fleet redistribution, adaptive scheduling, cost control, and last-mile delivery remapping. The proposed clustering approach enhances visibility into shipment patterns and supports more efficient, responsive, and adaptive decision-making in logistics supply chain management.