Workshop on Knowledge Engineering for Planning and Scheduling (KEPS)

1. Overview

Despite the progress in automated planning and scheduling systems, these systems still need to be fed by carefully engineered domain and problem description and they need to be fine-tuned for particular domains and problems. Knowledge engineering for AI planning and scheduling deals with the acquisition, design, validation and maintenance of domain models, and the selection and optimization of appropriate machinery to work on them. These processes impact directly on the success of real-world planning and scheduling applications. The importance of knowledge engineering techniques is clearly demonstrated by a performance gap between domain-independent planners and planners exploiting domain dependent knowledge.

The workshop shall continue the tradition of several International Competitions on Knowledge Engineering for Planning and Scheduling (ICKEPS) and KEPS workshops. Rather than focusing only on software tools and domain encoding techniques –which are topics of ICKEPS– the workshop will cover all aspects of knowledge engineering for AI planning and scheduling.

We seek original papers ranging from experience reports to the description of new technology in the following areas:

• formulation of domains and problem descriptions
• methods and tools for the acquisition of domain knowledge
• pre- and post-processing techniques for planners and schedulers
• acquisition and refinement of control knowledge
• formal languages for domain description
• re-use of domain knowledge
• translators from other application-area-specific languages to solver-ready domain models (such as PDDL)
• formats for specification of heuristics, parameters and control knowledge for solvers
• import of domain knowledge from general ontologies
• ontologies for describing the capabilities of planners and schedulers
• automated reformulation of problems
• automated knowledge extraction processes
• domain model, problem and plan validation
• visualization methods for domain models, search spaces and plans
• mapping domain properties and planning techniques
• plan representation and reuse
• knowledge engineering aspects of plan analysis

In particular, this year we encourage papers with a specific focus on the task of model acquisition, more concretely characterized by the three following categories:

1. Solicitation: The manual process of specifying a model.
2. Induction: The fully automatic synthesis of a model from data.
3. Hybrid: A mix of partial-model induction and solicitation

In addition to the areas listed above, we particularly welcome works on:

• Data-driven learning approaches for model induction.
• User interfaces for model solicitation.
• Analysis techniques for model debugging.
• Hybrid methods for model induction & model solicitation.
• Patterns and anti-patterns for effective models.
• Challenges and open problems in model acquisition.

1. Accepted papers

Check out the papers at: https://openreview.net/group?id=icaps-conference.org/ICAPS/2019/Workshop/KEPS . The following is the list of accepted papers:

• Reprocess: Process Refinement for Improving Accuracy in Hybrid Planning Domain Models · Alan Lindsay, Santiago Franco, Rubiya Reba, Thomas L. McCluskey
• Domain Modeling for Multi-Payload Planning of Experimental Satellite · Peng Wu
• Commonsense Reasoning and Knowledge Acquisition to Guide Deep Learning on Robots · Tiago Mota, Mohan Sridharan
• Object-Centric Knowledge Representations in Hierarchical Planning for Decentralized Environments · Ugur Kuter
• Goal-constrained planning domain model formal verification of safety properties · Anas Shrinah, Kerstin Eder
• The International Competition on Knowledge Engineering for Planning and Scheduling: Food for Thoughts (and Call to Action) · Mauro Vallati, Lukas Chrpa
• Incremental Learning of Action Models for Planning · Jun Hao Alvin Ng, Ronald P. A. Petrick
• One-shot learning: From domain knowledge to action models · Diego Aineto, Sergio Jiménez, Eva Onaindia
• Planning for Implicit Coordination using FOND · Thorsten Engesser, Tim Miller
• Incremental Learning of Discrete Planning Domains from Continuous Perceptions · Luciano Serafini, Paolo Traverso

1. Program Details

Time	Program
09:00-09:15	Workshop Open
09:15-10:15	Invited Talk: Scott Sanner
10:15-11:00	Coffee Break
11:00-12:15	Session 1 Goal-constrained planning domain model formal verification of safety properties Object-Centric Knowledge Representations in Hierarchical Planning for Decentralized Environments Domain Modeling for Multi-Payload Planning of Experimental Satellite Planning for Implicit Coordination using FOND The International Competition on Knowledge Engineering for Planning and Scheduling: Food for Thoughts (and Call to Action)
12:30-14:15	Lunch Break
14:15-15:30	Session 2 Reprocess: Process Refinement for Improving Accuracy in Hybrid Planning Domain Models Commonsense Reasoning and Knowledge Acquisition to Guide Deep Learning on Robots One-shot learning: From domain knowledge to action models Incremental Learning of Action Models for Planning Incremental Learning of Discrete Planning Domains from Continuous Perceptions
15:30-16:00	Coffee Break
16:00-17:30	Tutorial: From Teaching the PDDL Novice to Empowering the Planning Solution Integrator

1. Organizing Team

• Mauro Vallati, University of Huddersfield
• Lukas Chrpa, Czech Technical University
• Ron Petrick, Heriot-Watt University
• Tiago Vaquero, NASA JPL
• Christian Muise, IBM Research AI (Cambridge, MA)
• Tathagata Chakraborti, IBM Research AI (Cambridge, MA)

1. Program Committee

• Mauro Vallati
• Lukas Chrpa
• Ron Petrick
• Tiago Vaquero
• Christian Muise
• Susana Fernández
• AndreA Orlandini
• Amedeo Cesta
• Roman Barták
• Simone Fratini
• Alan Lindsay