Coordinating Decisions in a Supply-Chain Trading Agent Wolfgang Ketter, John Collins † , and Maria Gini †⋆ Dept of DIS, RSM Erasmus University, Rotterdam, NL † Dept of CSE, University of Minnesota, Minneapolis, MN, USA Abstract. An autonomous trading agent is a complex piece of software that must operate in a competitive economic environment. We identify the problem of de- cision coordination as a crucial element in the design of an agent for TAC SCM, and we review the published literature on agent design to discover a wide variety of approaches to this problem. We believe that the existence of such variety is an indication that much is yet to be learned about designing such agents. 1 Introduction Supply-Chain Management is an especially challenging domain for a rational decision- maker. Such an agent must not only operate simultaneously in multiple markets (a cus- tomer market and a supplier market), but it must coordinate its market activities with each other and with internal processes such as production scheduling and inventory management in a way that maximizes its utility across an extended time horizon. Organized competitions can be an effective way to drive research and understanding in complex domains, free of the complexities and risks of operating in open, real-world environments. Artificial economic environments typically abstract certain interesting features of the real world, such as markets and competitors, demand-based prices and cost of capital, and omit others, such as human resources, secondary markets, taxes, and seasonal demand. The Trading Agent Competition for Supply-Chain Management [1] (TAC SCM) is based on an economic simulation in which competing autonomous agents operate in a simple supply-chain scenario, purchasing components, managing a factory and warehouse, and selling finished products to customers. TAC SCM has been an active competition since 2003, and the design of the game has been stable since 2005. More than 50 different teams have participated, and a num- ber of papers have been published that describe agent designs, agent and game analyses, and specific methods for modeling the markets and decision processes in the simulation. TAC SCM is an interesting challenge for a number of reasons. Different groups have approached the problem from a variety of perspectives, depending on the individ- ual interests and backgrounds of the participants. For example, a team that is primarily interested in developing and testing machine-learning techniques will have a very dif- ferent approach to the problem than a team that is primarily interested in developing methods to solve constrained optimization problems under uncertainty. To better un- derstand this variety, we conducted an informal survey of many of the active teams in ⋆ Supported in part by the National Science Foundation under award NSF/IIS-0414466.