Introduction to AI Planning
Marco Aiello, Ilche Georgievski
TL;DR
The notes introduce AI Planning by outlining a state-space, deterministic, and static state model and then surveying major planning paradigms. They connect classical planning with STRIPS/PDDL representations, analyze worst-case complexity (e.g., PlanEx and PlanLen being PSPACE-complete) and algorithmic approaches (forward/backward search with heuristics), and show how planning can be framed as CSPs with bounded plan lengths. The HTN planning section presents hierarchical knowledge, task decomposition, plan/state spaces, and constraint posting, highlighting how rich domain knowledge (e.g., SHOP) yields speedups at the cost of requiring well-crafted domain models, alongside a complexity landscape that ranges from NP-complete to EXPSPACE in unrestricted settings. The Bonus PDDL coverage anchors the practical language used to encode domains, problems, types, and a STRIPS subset, illustrating how domain/problem descriptions are separated and instantiated for real-world planning problems. Overall, the notes provide a cohesive, technically grounded overview of foundational planning formalisms, their computational properties, and practical representation standards used in AI planning today.
Abstract
These are notes for lectures presented at the University of Stuttgart that provide an introduction to key concepts and techniques in AI Planning. Artificial Intelligence Planning, also known as Automated Planning, emerged somewhere in 1966 from the need to give autonomy to a wheeled robot. Since then, it has evolved into a flourishing research and development discipline, often associated with scheduling. Over the decades, various approaches to planning have been developed with characteristics that make them appropriate for specific tasks and applications. Most approaches represent the world as a state within a state transition system; then the planning problem becomes that of searching a path in the state space from the current state to one which satisfies the goals of the user. The notes begin by introducing the state model and move on to exploring classical planning, the foundational form of planning, and present fundamental algorithms for solving such problems. Subsequently, we examine planning as a constraint satisfaction problem, outlining the mapping process and describing an approach to solve such problems. The most extensive section is dedicated to Hierarchical Task Network (HTN) planning, one of the most widely used and powerful planning techniques in the field. The lecture notes end with a bonus chapter on the Planning Domain Definition (PDDL) Language, the de facto standard syntax for representing non-hierarchical planning problems.