in Proceedings of the 3 rd International Conference on Intelligent System Design & Applications, Springer Verlag, New York, 2003. What-if Planning for Military Logistics M. Afzal Upal 1 OpalRock Technologies 42741 Center St. Chantilly, VA 20152 afzal@opalrock.com Abstract. The decision makers involved in military logistics planning need tools that can answer hypothetical (“what-if”) questions such as how will a detailed logistics plan change if the high level operational plan is changed. Such tools must be able to generate alternative plans in response to such questions, while maintaining the original plan(s) for comparison. This paper reports on the work performed to add this capability to the multiagent planning and execution system called Cougaar. We state the what-if planning problem, describe the challenges that have to be addressed to solve it, discuss a solution that we designed, and describe the limitations of our approach. 1 Introduction & Background Conventional wisdom in the AI has been that the classical AI planning problem is hard enough without the additional complexity of generating and managing multiple plans. Military logistics planning in particular, is known to be a notoriously difficult problem because of (a) the millions of different object types and thousands of heterogeneous interacting organizational units involved, (b) the complex continual interplay between planning, execution, and replanning, and (c) stringent performance requirements [1, 2, 3]. Work on DARPA’s Advanced Logistics Project (ALP) and its successor Ultra*Log has produced a multi-agent, hierarchical planning, execution monitoring, and replanning system called Cougaar (Cognitive Agent Architecture) that aims to address this problem [3, 4]. Cougaar was designed to take advantage of the efficiencies available in the military logistics domain. The result is a robust system that can produce a level-5 (to the ‘eaches’ and ‘bumper numbers’) detail logistics plan for a given high level Operations Plan (Oplan) within minutes [3]. However, Cougaar does not allow the decision makers to specify hypothetical changes to the Oplan (such as “what if another tank unit needs to be moved from Fort Anderson to Fort Davis?”) and see the alternative plan(s) produced to satisfy the new Oplan. This paper reports on the work that IET’s Ultra*Log team performed to add this capability to Cougaar. 1 This work was performed during author’s tenure at Information Extraction & Transport Inc. 1901 North Fort Myer Drive, Arlington, VA 22209, under a DARPA contract.