Language-Facilitated Human-Robot Cooperation within a Human Cognitive Modeling Infrastructure: A Case in Space Exploration Task

Abstract

It is natural and efficient to use natural language for transferring knowledge from a human to a robot. The inconsistency of human-robot spatial cognitive style, the high frequency communication and low-cognition-level symbol matching control have greatly affected the operational efficiency in spatial-cognition-demanding tasks such as positioning and exploring. To fill the knowledge gap, this study applies ACT-R cognitive theory to establish a new way of knowledge representation and processing for the robots with a purpose to improve the flexibility of natural language facilitated humanrobot cooperation. This idea is specifically validated in the task of human-robot teaming space exploration.