Applying Bayesian belief networks as a tool for structuring and evaluating the planning of naval operations

Bayesian belief networks and influence diagrams are excellent methodologies for structuring complex planning problems under conditions of uncertainty and were applied to the process of operational planning of naval campaigns. This article describes how influence diagrams were used in the operations planning process of the South African Navy and the successes achieved. Influence diagrams enable the planning team to determine decisions and factors influencing the pending campaign in a structured manner. Conditional probability tables quantify the interrelationships between factors and decisions, and in the process a probabilistic model of the campaign is built. It allows the team to concentrate on small subsets of the campaign without losing sight of the whole. The network also serves as a tool for answering what-if questions.