A Stochastic Optimization Framework Integrating Real-Time Wave Forecasting, Probabilistic Cargo-Loss Assessment, and Performance-Linked Subsidy Allocation for Small-Scale Jukung Operators in Pulau Pisang, Pesisir Barat
Background: Small-scale jukung operators in Pulau Pisang experience frequent cargo losses due to stochastic wave conditions, yet no decision-support tool links real-time forecasts with financial incentives. Objectives: To develop a mixed-integer stochastic programming model that (i) quantifies loss probability as a function of significant wave height, peak period, cargo type and insulation quality, (ii) allocates a performance-linked subsidy under a fixed budget, and (iii) respects Indonesian civil-code provisions and customary hak ulayat. Methods: Empirical data from 120 operators, BMKG buoy observations, and BPS production statistics calibrated a logistic loss model and a Weibull wave-height distribution. The model minimizes expected loss minus weighted subsidies, subject to safety constraints and force-majeure thresholds (Hs>3 m or Tp>10 s). Monte-Carlo simulation (10 000 scenarios) and sensitivity analysis evaluated robustness. Results: A 10 % increase in subsidy improves the safety performance index from 78 % to 85 %, reducing expected monetary loss by 3.2 % while satisfying the budget (B=Rp 1.2 billion). Conclusion: Integrating real-time wave forecasts with probabilistic risk assessment and incentive-compatible subsidies yields measurable safety and economic benefits, providing a replicable template for coastal logistics policy.