Human-machine interaction in the cockpit and applicable design changes towards better collaboration

Abstract

Current flight deck automation has improved the safety and efficiency of commercial aviation but a broad consensus has developed over the last 20 years that this technology is deficient in some areas. It has been developed in an ad hoc manner and without a human centered approach; leading to problems regarding the human/machine interaction and adversely impacting decision making throughout the flight. Current procedures and design do not give automation liability although it has great authority and autonomy during most phases of flight. Cockpit automation has not been designed in such a way to provide adequate and unambiguous feedback to the human operator as to its current and intended actions. More or different training is the most common response to this problem but has failed to fully compensate for the design flaws in current automated systems. Accidents that cite pilot error do not always acknowledge how difficult it is for human operators to overcome fundamental, system level, flaws in the design of the machines they work with. This paper proposes some changes in cockpit automation design that will improve the vigilance of the pilots and therefore create better decision-making. Numerous accident and incident reports have been cited by regulatory authorities when making changes in automated flight operation regulations. This reflects a “reactive” approach to FAA automated flight safety guidelines and highlights the need for an improved governance system in the cockpit. This paper also provides a literature review for current studies on human-machine interaction related to the cockpit.