Optimalisasi Resource Allocation di Perbankan dengan Pendekatan Desain Eksperimen dan Orthogonal Array untuk Minimasi Skenario Simulasi

Abstract

This study optimizes resource allocation in Islamic banking using Design of Experiments (DOE) and Orthogonal Arrays (OA) to minimize simulation scenarios while ensuring statistical validity. Focusing on a West Aceh-based bank (4 tellers, 2 CS agents, 5 ATMs), the research addresses operational inefficiencies prolonged queues, resource misalignment conflicting with Sharia principles of efficiency (israf avoidance) and equity, exacerbated by seasonal demand fluctuations. A Full Factorial Design initially generated 40 scenarios (200 iterations), imposing high computational burdens. OA techniques reduced scenarios to 9 (77.5% fewer iterations, 45 total), maintaining a 95% confidence level (z-score=1.96) and 1-minute margin of error. Utility differences between methods were marginal (CS:1.3%, Teller:1.5%, ATM:2.3%) and statistically insignificant, confirming OA’s reliability. Critical interactions (e.g., CS=1 paired with all Teller levels) and balanced configurations (Teller=3 + ATM=3 “sweet spot”) were identified, despite minor imbalances (e.g., ATM=4 tested once). The approach advocates data-driven policies deploying ATMs during CS shortages and optimizing human-technology ratios aligning with Sharia ethics by reducing waste and enhancing equity. With 77.5% computational savings, this methodology offers scalable solutions for Islamic finance and dynamic sectors, bridging statistical rigor with ethical governance in resource optimization.