Measuring the interaction complexity of teleoperated systems

Abstract

Concerns control of telescience systems in space by ground-based scientists, as in the Canadian Advanced Protein Crystallization Apparatus (CAPCA). We describe in this report how the formal modelling and analysis of the operator-system interaction makes it possible to evaluate the "error-proneness" of the operator interface in terms of the definition and measurement of metrics related to the complexity of the operator decisions since reduced complexity often means improved reliability. The intent is to be able to compare alternative interfaces to the very same telescience infrastructure in terms of cognitive complexity (and therefore error-proneness) by carefully evaluating the operator decision-making interactions with the system in meaningful ways. As a result, our report concludes by defining candidate complexity metrics and by discussing how they might be validated by experimental work.