Proactive Flight Operations Monitoring and Safety Management: a System Informatics Approach

Project: Research

View graph of relations


Hong Kong’s stature as a leading international hub for transportation is closely linked to itaviation safety record. Despite continuous efforts to improve safety, accidents are stilloccurring such as Finnair Flight 070 or Cathay Pacific Airways Flight 780. Unlike decadesearlier, safety incidents have been traced to the operation of the aircraft systems, or tohuman errors in particular, which continue to occur even as the development of hardwarehas advanced.To improve pilot operations, airlines have begun an extensive monitoring of flightoperations, thanks to an abundance of data recorded by the digital Flight Data Recorder (theBlack Box). However, traditional data analytics methods are quickly becoming obsolete andlargely irrelevant for proactive safety management. For instance, Exceedance Detection,widely used by the airline industry, can only detect hazardous behaviors from a pre-definedlist comprised of “known issues of safety concerns”; it cannot respond to emerging,previously unidentified issues that are yet potentially dangerous. Another flaw is that this listis only updated after accidents, if at all, usually at the cost of lives and airline reputations.In light of abundantly available datasets and recent advances in data analytics, there is acritical need to revamp the analytics toolset to enable the monitoring of flight operations toachieve proactive safety management. Yet, research advances in this area have been sparsedue to the fact that current approaches are ill-equipped to deal with the emerging “systems”perspective of flight operations. Existing model-based or rule-based techniques focus onmonitoring equipment performance but not other critical aspects of flights such as the pilot,weather, or aircraft systems. New holistic informatics approaches that treat flight operationsfrom a systems perspective are urgently called for.In response, this research proposes a system informatics-based approach to proactivelymonitor flight operations using Black Box data. With the rapid development of informationtechnologies and the big data collected by ubiquitous sensors, system informatics aretransforming how data is used for the sensing, learning, monitoring, diagnosis and prognosisof the structures as well as dynamics of systems. The objective of this research is to offertools that can identify deficiencies in flight operations, which requires three research tasks:(1) advancing data-mining methods for pattern identification and anomaly detection in flightoperations via a system informatics approach; (2) validating the above methods using expertreviews, case studies, and cross-checking with existing tools; and (3) applying thesemethods at airlines for safety management. The PI has recently conducted a pilot study withCathay Pacific Airways, with results showing that the new breed of system informaticsdriven approach can identify ‘common patterns’ as well as anomalies, allowing airlines toexamine the consistency of current operations while focusing efforts on investigatingunusual behaviors to look for latent risks. The academic and pragmatics experience fromthis pilot study will be further developed in the proposed effort.The research will contribute to literature in system informatics related to human factors, andthe methods will be valuable for use in a variety of human-in-the-loop systems goingbeyond flight operations (e.g. health monitoring, computer network security). From theapplication side, the research aimed to enhancing Hong Kong’s airline safety managementby allowing operators to detect latent safety degradations and adjust pilot trainingaccordingly.


Project number9048075
Grant typeECS
Effective start/end date1/09/1625/02/21

    Research areas

  • Airline Safety , Data Analytics , Flight Operations , System Informatics ,