Using temporal logics for planning and control

Abstract

Traditionally, planning work in artificial intelligence has focused on primitive actions and instantaneous states. Innovative work has been done on composing primitive actions so as to bring about desired final states. The thesis advanced in this paper is that instead of simply focusing on primitive actions, it is also useful to use representation and reasoning tools whose primitive objects are sequences of actions (or the associated sequences of states that they generate). Temporal logics are a useful tool for representing and reasoning about action sequences. The author has examined some different applications of temporal logics to problems of planning and control.