Flightcrew alerting: history, research, regulation, and successes

Flightcrew’s responses to alerts save lives multiple times a day. The proper design of the alert ensures correct detection, interpretation, and timely response. This paper explores the many factors of successful flightcrew alerting through the history, human factors research, regulations, and successes. By the 1970s transport aircraft had become complex, resulting in an increase in alert states (e.g., Boeing 747 with 455 alerts). Research revealed that the lack of prioritization, differentiation, and aggregation of flight deck alerts was a safety issue. The qualities of effective alerting were known: quickly orient, explain action needed, convey priority, minimize false positives/negatives, and indicate adequacy of resolution. In 1981, based on their human factors research, three large transport aircraft manufacturers compiled voluntary standards for alerting systems for the next generation of transport aircraft. Regulations and advisory materials followed for effective and standardized alerting (e.g., EASA CS 25.1322). These standards define the priority and appearance of warnings, cautions, and advisories. To comply, aircraft must have integrated alerting systems with complex logic, centralized data busses, sensors, and displays. Alerts for external hazards such as terrain, windshear, and traffic must also be integrated. These regulations continue to be harmonized across regulatory agencies. Guided by standardization, many human factors issues apply, including appropriate sensory modality, task saturation, crew coordination, and basic user-interface principles. Human factors science advises on what to alert, when to alert, where to alert, and how to alert effectively. All these factors are explained here. The relative criticality balanced with urgency leads to the priority of any alert being properly activated, given the current conditions. Even critical alerts are inhibited during high-workload moments in the flight. The flight deck systems must promote accurate flight crew response and immediate feedback when the non-normal condition no longer exists. The ability to reduce distraction by suppressing an alert is also important but must be accompanied by a salient indication that it has been suppressed. Naturally, nuisance occurrences reduce the effectiveness of alerts and must be minimized. Proper alerting requires continual improvement. Our human factors work is not complete. Flight data and incident/accident reports are an important source of alerting successes and failures. These data show us that alerts fail to be properly activated, properly responded to, and properly trained. Two areas that show promise in this area are training for response to the startle effect and to loss of control.