An Approach to Modeling Error in Air-Midas Using Contextual Control Model

Abstract

Predictability of the possible avenues of human error is important for designing error recovery mechanisms in systems' interfaces. Our approach for modeling errors during the taxiing and pre-flight phase, involves using Hollnagel's (1 993) Contextual Control Model (CoCoM). Hollnagel postulates that changes in the human's operational context make her switch control modes or cognitive strategies. Previously, Verma & Corker (2002) encoded rules to implement CoCoM for Air Traffic Controllers (ATCo) in the Air-MIDAS (Man-Machine Integrated Design and Analysis System) architecture. They defined the operational context as the number of simultaneous goals and the event horizon time required to successfully complete those goals, and also encoded behavior associated with three types of cognitive strategies-unplanned, tactical, and strategic. In this study, new sets of rules are being coded for the pilots, such that changes in the context will cause the simulated operator to change cognitive strategies. The dependent variables of interest are cognitive strategies used by pilot-agents, workload, status of tasks (completed, aborted, working), and consequences of errors performed.