A Cognitive Control Architecture for an Artificial Creature using Episodic Memory

Abstract

This paper describes a new cognitive control architecture incorporating episodic memory for the artificial creature Rity, the software robot component of the Ubibot, the ubiquitous robot system. The episodic memory is defined as a scalable structure that stores episodic perceptual snapshots as Rity's experience grows. The system also utilizes a temporally variant spatial map to store spatial information and a higher level procedural memory using Finite State Machines. The system is designed to enable Rity to be cognitive in its approach to task selection through the dual process of experiential and spatial learning. This is brought about through a multi-agent strategy based on six principal modules: perception module, internal state module, behavior selection module, interactive learning module, memory module, and motor module, to control its behavior considering its internal state. Experiments on completion of task and maintenance of ideal internal state are described. The results show that the artificial creature possesses the ability to improve its performance as its experience grows.

Towards Long-Term Memory for Social Robots: Proposing a New Challenge for the RoboCup@Home League

“…During the past decade there have been several approaches that implemented episodic and semantic long-term memory in artificial systems (e.g., [5,6,10,13,14,15,16,18,22]). We believe that these works addressed only partially some of the main challenges that poses the implementation of long-term memory for social robots, and that still some of the questions described in the next paragraphs need to be answered.…”

Deep Episodic Memory: Encoding, Recalling, and Predicting Episodic Experiences for Robot Action Execution

“…When restricted to a specific context, symbolic representations and pre-specified perceptual instances stored in an episodic memory can indeed be a powerful approach for enhancing the reasoning capabilities of a cognitive system, as shown in Soar [5]. However, most of the described approaches are customized to a specific problem domain and rely on pre-defined, problem specific representations [6], [10]. In complex real world scenarios transferring and generalizing knowledge persisted in the episodic memory is very limited when pre-defined symbolic representations are used.…”

An Episodic Memory Implementation for a Virtual Creature

“…The Episodic Memory architecture was mainly developed by researchers N. S. Kuppuswami and Se-Hyoung Cho, from Kim's team, in the middle of the decade in order to provide a cognitive task selection mechanism for Rity. The creature's architecture explores the advantages of a reactive architecture with the higher level planning offered by the Episodic Memory, in addition to provide a learning mechanism that evolves with time, since Rity's decision making process is more efficient as the creature's experience grows [12].…”

“…This is the main motivation of this work. Even though some related initiatives already started to appear in the literature [8,6,14,3,5,12,15,16,9], we are still very far from having this as a well known technology to be widely used in intelligent agents. In this work we report on our ongoing efforts to bring up such technology by building up a cognitive architecture where episodic memory is a central capability.…”