Learning Type-Generalized Actions for Symbolic Planning

Abstract

Symbolic planning is a powerful technique to solve complex tasks that require long sequences of actions and can equip an intelligent agent with complex behavior. The downside of this approach is the necessity for suitable symbolic representations describing the state of the environment as well as the actions that can change it. Traditionally such representations are carefully hand-designed by experts for distinct problem domains, which limits their transferability to different problems and environment complexities. In this paper, we propose a novel concept to generalize symbolic actions using a given entity hierarchy and observed similar behavior. In a simulated grid-based kitchen environment, we show that type-generalized actions can be learned from few observations and generalize to novel situations. Incorporating an additional on-the-fly generalization mechanism during planning, unseen task combinations, involving longer sequences, novel entities and unexpected environment behavior, can be solved.

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2023

References in corpus (8)

• PDDL2.1: An Extension to PDDL for Expressing Temporal Planning Domains
• Learning Symbolic Models of Stochastic Domains
• Symbol Emergence in Cognitive Developmental Systems: a Survey
• Ordered Landmarks in Planning
• From Machine Learning to Robotics: Challenges and Opportunities for Embodied Intelligence
• Explainable Human-Robot Training and Cooperation with Augmented Reality
• Evolutionary Training and Abstraction Yields Algorithmic Generalization of Neural Computers
• Teaching Robots Novel Objects by Pointing at Them

Cited by in corpus (2)

• CoPAL: Corrective Planning of Robot Actions with Large Language Models
• Neuro-Symbolic Imitation Learning: Discovering Symbolic Abstractions for Skill Learning