Constraints and AI Planning

Abstract

Tackling real-world planning problems often requires considering various types of constraints, which can range from simple numerical comparators to complex resources. This article provides an overview of techniques to deal with such constraints by expressing planning within general constraint-solving frameworks. Our goal here is to explore the interplay of constraints and planning, highlighting the differences between propositional satisfiability (SAT), integer programming (IP), and constraint programming (CP), and discuss their potential in expressing and solving AI planning problems.