Context-Aware Dynamic Asset Allocation for Maritime Interdiction Operations

This paper validates two approximate dynamic programming approaches on a maritime interdiction problem involving the allocation of multiple heterogeneous assets over a large area of responsibility to interdict multiple drug smugglers using heterogeneous types of transportation on the sea with varying contraband weights. The asset allocation is based on a probability of activity surface, which represents spatio-temporal target activity obtained by integrating intelligence data on drug smuggler whereabouts/waypoints for contraband transportation, behavior models, and meteorological and oceanographic information. We validate the proposed architectural and algorithmic concepts via several realistic mission scenarios. We conduct sensitivity analyses to quantify the robustness and proactivity of our approach, as well as to measure the value of information used in the allocation process. The contributions of this paper have been transitioned to and are currently being tested by Joint Interagency Task Force—South, an organization tasked with providing the initial line of defense against drug trafficking in the East Pacific and Caribbean Oceans.