Daniel M. Gaines, Caroline C. Hayes, and Fernando Castano
Researchers attempting to deploy complex planners in practical applications have met with the high cost of maintaining and debugging a planner’s knowledge base. There has been some success in using knowledge acquisition tools and machine learning techniques to assist in development of operator schemas. However, it is often hard for the domain experts who are the users of the program to do their own planner maintenance unless they are trained to think in terms of AI planning concepts, since their own representations of the problem may or may not resemble typical operatorschema- based representations. Such automated planning tools may fall by the wayside in practical applications due to difficulty of retraining for maintenance. In this work we have addressed the knowledge acquisition problem by using an alternative representation of the planner’s knowledge, which better fits the users’ representations in particular types of complex physical domains in which the available effectors (i.e. tools and materials used in creating actions) change rapidly, but the way in which they are used does not. We call our representation for these problems and the technique for applying it Effector-based Operator: Construction. One of the important characteristics that distinguishes the effector-based approach from other approaches is that the representation separates the effector descriptions from the descriptions of how those effectors are used. Operator and action schema typically combine these two types of information in the representation. This separation is of importance in maintenance of complex physical domains because it makes it possible for domain experts who are non-programmers to make the changes required for planner maintenance by expressing the changes in concrete physical terms with which they are familiar.