EFFECTS OF EXTENDED-RANGE FORECASTING IN AUTOMATED SHIP ROUTING

Abstract

Development of the Navy Earth System Prediction Capability (ESPC) has improved the accuracy of extended-range environmental forecasting. However, the effects of using this capability in ship routing are not known. The Naval Research Laboratory (NRL) has developed a tool, the Tool for Multi-objective PLanning and Asset Routing (TMPLAR), to optimize ship routes while considering objectives including time, fuel costs, and environmental conditions. This thesis reports a simulation experiment using recent ESPC forecasts and forecasts from a legacy forecasting system in TMPLAR. ESPC impact on ship routes and their fuel consumption, overall transit times, and exposure to waves is explored. Our findings are inconclusive regarding ESPC performance against the legacy and climatology forecasting systems. However, large-scale simulation has drastically improved TMPLAR robustness and provided useful insight into overall TMPLAR behavior at a significantly reduced cost relative to any real-world experiment.