The Community for Technology Leaders
RSS Icon
Issue No.06 - December (1996 vol.11)
pp: 4-12
<p>Planning systems generate partially ordered sequences of actions (or plans) that solve a goal. They start from a specification of the valid actions (also called operators), which includes both the conditions under which an action applies (the preconditions) and the expected outcome of applying that action (the effects). The general problem is quite hard both because of the potentially enormous search space and the difficulty in fully and accurately representing real-world problems. Approaches to planning include operator-based planning, hierarchical task-network planning, case-based planning, reactive planning, and many more. Early planning work focused largely on "toy" problems (for example, the blocks world). More recently, there has been a big push toward applying planning systems to real-world applications. While planning systems have not yet achieved the level of commercial success enjoyed by some other areas of artificial intelligence—neural nets, for example—a number of successful applications of planning technology to real-world problems have recently emerged.</p><p>This installment of "Trends & Controversies" highlights five such applications. I have asked the developers of these systems to describe the application domain and the planning technology used to solve the problems. These systems all use some form of hierarchical task-network planning (in some cases combined with other techniques). HTN planning provides a way of specifying, as part of the operator definition, how to hierarchically expand actions into partially ordered sequences (task networks) of actions. This approach succeeds, in part, because it provides a natural way of limiting the possibly very large search spaces. See <it>Readings in Planning</it> (Morgan Kaufmann, 1990) or <it>Artificial Intelligence: A Modern Approach</it> (Prentice Hall, 1995) for more details on various planning techniques.</p><p>In the first article, Stephen Smith, Dana Nau, and Thomas Throop describe their use of planning technology to build a system for declarer play in contract bridge. The system can beat the best commercially available program and is currently being incorporated into a commercial product. Second, John Mark Agosta and David Wilkins describe how the SIPE-2 planner helps evaluate the US Coast Guard's ability to respond to marine oil spills. This system, which automates a problem that is currently done by hand, is undergoing evaluation by the Coast Guard. Third, Austin Tate describes a planning application, in use by the European Space Agency, for the project management of spacecraft assembly, integration, and verification. Fourth, Steve Chien and his colleagues describe their use of a planning system to automate the operations of NASA's Deep Space Network communication antennas. This system is currently being integrated into a new system that will become operational in 1997. Finally, Thomas Lee and David Wilkins describe their use of SIPE-2 in producing military air campaign plans. Their planner is part of a demonstration system that is fully integrated with the other software modules currently used for solving parts of this problem.</p><p><bi>—Craig Knoblock</bi></p>
"AI planning systems in the real world", IEEE Intelligent Systems, vol.11, no. 6, pp. 4-12, December 1996, doi:10.1109/MIS.1996.10033
6 ms
(Ver 2.0)

Marketing Automation Platform Marketing Automation Tool