Robot Representation and Reasoning with Knowledge from Reinforcement Learning

TL;DR

This paper presents a novel integration of logical-probabilistic knowledge representation with model-based reinforcement learning, enabling robots to reason with declarative knowledge and learn from interactions, leading to improved task performance.

Contribution

It introduces a unified framework combining KRR and RL, allowing agents to reason with declarative knowledge while learning from experience in dynamic environments.

Findings

Significant performance improvements over existing model-based RL methods.

Effective reasoning with declarative knowledge in robotic tasks.

Successful application in dialog, navigation, and delivery tasks.

Abstract

Reinforcement learning (RL) agents aim at learning by interacting with an environment, and are not designed for representing or reasoning with declarative knowledge. Knowledge representation and reasoning (KRR) paradigms are strong in declarative KRR tasks, but are ill-equipped to learn from such experiences. In this work, we integrate logical-probabilistic KRR with model-based RL, enabling agents to simultaneously reason with declarative knowledge and learn from interaction experiences. The knowledge from humans and RL is unified and used for dynamically computing task-specific planning models under potentially new environments. Experiments were conducted using a mobile robot working on dialog, navigation, and delivery tasks. Results show significant improvements, in comparison to existing model-based RL methods.