An Autonomous Agent for Supply Chain Management

Supply Chain Management involves planning for the procurement of materials, assembly of finished products from these materials, and distribution of products to customers. The Trading Agent Competition Supply Chain Management scenario (TAC SCM) provides a competitive benchmarking environment for developing and testing agent-based solutions to supply chain management. Autonomous software agents must perform the above tasks while competing against each other as computer manufacturers: each agent must purchase components such as memory and hard drives from suppliers, manage a factory where computers are assembled, and negotiate with customers to sell computers. In this chapter, we describe TacTex06, the winning agent in the 2006 TAC SCM competition. TacTex-06 operates by making predictions about the future of the economy, such as the prices that will be offered by component suppliers and the level of customer demand, and then planning its future actions in order to maximize profits. A key component of TacTex-06 is the ability to adapt these predictions based on the observed behavior of other agents. Although the agent is described in full, particular emphasis is given to agent components that differ from the previous year’s winner, TacTex-05, and the importance of these components is demonstrated through controlled experiments.