The crew activity tracking system: leveraging flight data for aiding, training and analysis

Technological advances have made flight data a viable real-world data source for studies of human error and error prevention; hundreds of parameters are currently available for analysis. These data have enabled airlines to institute increasingly advanced Flight Operational Quality Assurance (FOQA) programs, which analyze flight data from line operations to detect "operational irregularities that can foreshadow accidents and incidents," and proactively disseminate this information to flight crews and maintenance personnel. This paper presents an intent inference technology, referred to as activity tracking, that in the future could also support flight-data-driven safety-enhancement efforts. A methodology for activity tracking has been implemented and validated in the Crew Activity Tracking System (CATS) As implemented for the flight deck, CATS uses knowledge about the pilot's task and the current operational context to predict nominal activities and interpret actual pilot actions. By analyzing pilot action data in conjunction with clearance constraints and other flight data parameters, CATS can help disambiguate errors from other causes of abnormal flight conditions, and characterize error-inducing contexts in operational terms.