Michael Freed and Gregg Collins
Human agents typically evolve a set of standard routines for carrying out often-repeated tasks. These routines effectively compile knowledge about how to carry out sets of interacting tasks without causing harmful interference. By modifying its routines in response to observed failures, an agent can refine enlarge and refine its planning repertoire over time. We are constructing a software agent, RAFTER, that applies this incremental improvement strategy to the management of planning routines for use in the Truckworld simulator. Currently, we are building in the capacity to make a variety of repairs to the method used in a routine to coordinate constituent tasks; in the future, RAPTER will be extended to include methods for coordinating tasks from different routines.