Agent-based refinery supply chain management

The refinery business involves tasks spanning across several departments and requiring close coordination and collaboration among them. Present-day business decision support systems are tuned to individual departments and their work processes and are incapable of exploiting the intricate interdependencies among them. There is a need for a decision support system that provides a structured way to make decisions at the enterprise level. In this paper, we address this critical need by developing an agent-based decision support system for refinery supply chain management. Software agents are used to emulate the internal departments of the refinery. These agents gather relevant information from across the enterprise, make decisions based on the embedded business policies and mimic the different business processes of the modeled enterprise. Uncertainties in decision-making are captured by stochastic elements embedded in the agents. Through this, the dynamics in the supply chain is emulated and the performance of the supply chain measured through department-specific key performance indicators.