ABSTRACT
This chapter begins with the proposition that the thinking used in "artificial intelligence," particularly about information and control in animals, machines, and persons, may help in explicating "rational planning." Computer programs, algorithms and mathematical proofs, constructions, or derivations must be planned in advance. Planning for problem solving was discussed in one of the first review articles on artificial intelligence. The chapter discusses the cognitive elements of planning, the knowledge needed for planning, what can and cannot be planned, and finally, planning procedures. Within the constraints on rational planning it argue for more knowledge-intensive planning, especially at this third level of planning support. The chapter concludes with a proposed traffic information and control system that illustrates the dual role of information processing in planning: the use of knowledge as an input for planning, and knowledge as a critical part of what is being planned as an output.
