The design principles of flight deck automation and the occurrence of active failures in aviation

Abstract

The evolution of advanced technology systems in aviation has seen radically increased capabilities of aircraft, and equally radical changes in how aircraft are flown. Relieving flight crews of much of the manual workloads associated with flying, automation has brought about a shift in the dynamic on the flight deck as the role of crews - who are gradually being removed from direct control of the aircraft - moves towards that of supervisors and managers of the vast array of systems on-board. There is little doubt that automation has provided significant benefits in terms of increased performance, endurance and safety. Yet the sleekness and simplicity of the modern flight deck has proven deceptive. The complexities of aircraft systems, their dependencies and interdependencies, may mask interactions and inhibit the pilot's understanding of systems functionalities. Perhaps just as importantly, as automated systems assumed greater levels of autonomy and authority, the position of automation - and its relationship with those other key players in the cockpit - has not always been explicitly stated. Now managing and overseeing the aircraft’s systems, crews, whose exposure to manual flying has been reduced largely to the take-off and landing phases of flight, may be exposed to error causing conditions where they may not understand what the automation is doing.

The aim of this study was to examine the effects of latent conditions (pre-cursor faults) on the occurrence of decision errors, skill-based errors, perceptual errors and violations (active failures / unsafe acts). Based on the ASRS data analysis it was determined that while there was a significant number of automation pre-cursor faults associated with Airbus, Boeing aircraft were more likely to have mechanical related pre-cursor events.