William Cecil, Director for Business Development for Teledyne Controls’ Aircraft Data Services, considers how the practice of Flight Data Managemen has evolved and how effective it has been in improving airline flight safety and operational efficiency.
Flight Data Monitoring (FDM) solutions first came on the market to the airline community in the mid-1980s, at a time when DOS-based IBM Personal Computers were lacking the required horsepower.
Airlines who purchased them bought mini-computer based systems that had the ability to drive Data Retrieval Units to extract flight data from ¼ inch tape cartridges that had been removed from early aircraft quick access recorders (QAR). The data analysis software included the ability to identify, for safety purposes, exceptions within the data, or ‘events’. This required a set of algorithms, each crafted to identify a specific operational excursion (such as a ‘high Pitch Rate at Takeoff’, for example), together with an event ‘threshold’ that has previously been determined by the airline flight safety departments team (in this example, the value of pitch rate beyond which the airline flight safety team considers that risk is being introduced to aircraft operations).
The original FDM systems at this time were two-level event systems. They had a lower ‘Detect’ threshold for trending purposes, as well as being able to set higher ‘Alert’ thresholds to detect individual occurrences of significant operational risk.
Such a technique is still used in many FDM systems today, although three-level event detection is the preferred approach. The Special Event Search & Master Analysis (SESMA) program, a safety event set developed by the UK CAA and British Airways in the late 1970s, is still a benchmark for many airlines and is often the basis of a dependable FDM program within an airline.
The data-basing of these safety ‘events’, together with tools to plot and chart the stored time-series data in a meaningful way, started in the mid-1990s. The introduction of measurements, which allows normal operations and ‘outliers’ to be better understood, followed shortly afterwards. The introduction of 3D animation tools, designed to visually recreate the flight in more tangible ways, appeared around the early millennium. This ability to animate flight data gave flight safety departments new tools with which to interact more effectively with crews when communicating operational risk on a per-flight basis.
Today, modern wireless technologies on the aircraft that automatically transmit large batches of recorded data, such as Teledyne’s GroundLink® Comm+, are making traditional QAR systems obsolete; data capture for FDM is extremely dependable and can be 99% and above with data available every landing. On the ground, digitization and Application Programming Interfaces (APIs) allows many other data sources (sources such as weather data, flight schedules, crew rosters and air safety reporting systems) to be used in conjunction with QAR data and FDM results. This combination provides a richness of data that can be used to provide a significantly more insightful picture of airline operations than has previously been possible. To achieve this, visualization and mathematical tools such as Tableau and MATLAB, when exposed to this wealth of data, allows an airline analyst or data scientist to produce unparalleled views of the risks and levels of efficiency within airline operations.
So how successful have such FDM solutions been? Perhaps it is no coincidence that aircraft accidents have trended downwards in the last 30 years as FDM has been adopted by almost every airline. IATA’s statistics on accident rates show a fall (from a rate of 2.17 accidents per million departures in 2012) across the industry, including jets and turboprops, to a lowest-ever rate of 1.08 in 2017. This is despite a rise of passenger flights over the same period of 37.3%. Air travel is now so safe that there were no passenger jet crashes in 2017 anywhere in the world, according to the Airline Safety Network.
Complacency, however, can be a dangerous thing, and many of us are continuing to provide airline flight safety departments with enhanced safety analysis solutions to ensure that risks are even better understood and flight safety can be maintained.
In the next forty years the same flight data, used to improve flight safety over the last forty years, will drive airline operational efficiency and reliability improvements we can’t imagine today.