Operator Assistant Systems: An Experimental Approach Using a Telerobotics Application*

Abstract

This article presents a knowledge‐based system methodology for developing Operator Assistant (OA) systems in dynamic and interactive environments. Recent results in human error studies show that automation appears to be the most critical factor in human‐machine interaction problems. This is a problem both of training and design, which is the subject of this article. Design includes both design of the system to be controlled and design of procedures for operating this system. A specific knowledge representation is proposed for representing the corresponding system and operational knowledge. This representation is based on the situation recognition and analytical reasoning paradigm (SRAR). It tries to make explicit common factors involved in both human and machine intelligence, including perception and reasoning. An OA system based on this representation has been developed for space telerobotics. Simulations have been carried out with astronauts and the resulting protocols have been analyzed. Results show the relevance of the approach and have been used for improving the knowledge representation and the OA architecture. © 1993 John Wiley & Sons, Inc.