Postgraduate Course: Automated Planning (Level 11) (INFR11080)
|School||School of Informatics
||College||College of Science and Engineering
|Credit level (Normal year taken)||SCQF Level 11 (Postgraduate)
||Availability||Not available to visiting students
|Summary||The aim of this course is to provide a solid grounding in artificial intelligence techniques for planning, with a comprehensive view of the wide spectrum of different problems and approaches, including their underlying theory and their applications.
* Introduction and overview: intuitions and motivations. Basic conceptual model for planning: state transition systems, classical assumptions. Overview of different planning problems and approaches.
* Classical planning: The classical planning problem. Situation Calculus and the Frame Problem. Classical representations and languages (e.g., STRIPS-like). Overview of State-Space Planning and Plan-Space Planning.
* Hierarchical Task Network Planning. Partial-Order Planners. Mixed-initiative Planners.
* Neoclassical Planning: Modern approaches to the classical planning problem: e.g., Planning-Graph techniques, SAT-based planning.
* Heuristics and Control Strategies: Heuristics (in state-space and plan-space planning). Hand-coded control rules and control strategies. Deductive planning and control strategies in deductive planning.
* Planning with Time and Resources: Basics of point and interval temporal algebra. Temporal constraints networks. Planning with temporal operators. Integrating planning and scheduling
* More advanced planning topics: Knowledge Engineering for Planning (including advanced representations), distributed multi-agent planning, and plan execution.
* Case Studies and Applications: A selection from robotics, manufacturing, assembly, emergency response, space exploration, games, planning for the web, etc.
Areas Covered by Self-Study and Literature Review
* Scheduling: Linear and Integer Programming. Dynamic Scheduling. Applications to real world scheduling problems. Design, development and implementation of scheduling systems.
* Planning under uncertainty: different sources of uncertainty (e.g., nondeterministic actions, partial observability). Extensions to classical approaches (e.g., plan-space, planning-graph and propositional satisfiability techniques). Planning based on Markov Decision Processes. Planning based on Model Checking.
* Other problems and approaches which are open to review and study: Case-Based Planning. Plan Merging and Plan Rewriting. Abstraction Hierarchies. Domain Analysis. Typed variables and state invariants. Other kinds of domain analysis. Planning and Learning. Planning and Acting, Situated Planning, Dynamic Planning. Plan Recognition. Learning in Planning. Mixed-Initiative Planning. Knowledge-based Planning.
Relevant QAA Computing Curriculum Sections: Artificial Intelligence
Entry Requirements (not applicable to Visiting Students)
|Prohibited Combinations|| Students MUST NOT also be taking
Automated Planning (Level 10) (INFR10045)
||Other requirements|| For Informatics PG and final year MInf students only, or by special permission of the School.
Course Delivery Information
|Not being delivered|
| 1 - Understand and formalize different planning problems.
2 - Discuss the theoretical and practical applicability of different approaches.
3 - Have the basic know how to design and implement planning systems.
4 - Ability to review planning literature relevant to an area covered in the course.
5 - Know how to use planning technology for projects in different application domains.
|"Automated Planning: Theory and Practice" by M. Ghallab, D. Nau, and P. Traverso (Elsevier, ISBN 1-55860-856-7) 2004.|
|Course organiser||Dr Michael Rovatsos
Tel: (0131 6)51 3263
|Course secretary||Miss Kate Weston
Tel: (0131 6)50 2692