The title of this article sounds rather obvious to most readers, especially when thinking about faults that are significant enough to ground an aircraft. However, even relatively minor faults can influence flight safety.
A recent New York Times article, by William Langewiesche, spoke about the safety implications of a declining level of aircraft systems knowledge among new pilots. The article said, in part, that airlines are meeting the growing demand for pilots by reducing the experience qualifications for employment, and that handling failed or partially failed systems is a secondary concern when it comes to selecting new hires. As a result, a deep understanding of aircraft systems operation is becoming a less common trait among pilots.
I passed the article on to an industry colleague in the flight safety organization at a major North American airline. He told me that “flying a modern aircraft in both routine and non-routine flight profiles is challenging enough when the aircraft is 100% serviceable; dealing with degraded systems complicates that challenge by adding distraction from the task of flying.”
He further explained…
“Think of it like this. An extension cord laying on a floor is a hazard. The risk is tripping over it. The hazard is easily navigated under normal circumstances, but the risk becomes significant when you are distracted. So it is in the cockpit. You can fly with one of four brakes inoperative, but a lot changes – things like the rejected take-off procedure, your available runways, and even your alternate airports. Now add further distraction to that, like low weather in icing conditions at destination, and that relatively benign technical fault raises the flight safety risk significantly.”
He further reminded me that cockpit distraction has been a factor in some recent aircraft flight incidents and accidents. An extreme example he noted was the AirAsia Flight 8501 accident, where a recurring rudder trim limiter fault started a chain of events. This recurring issue took the pilots into a situation where they mishandled the fault and ultimately lost control of the aircraft – but even something as benign as the brake example can be significant.”
“Some forms of cockpit distraction are difficult to control – confusion with ATC, adverse weather, and personal stress, for example. But reducing distractions caused by technical issues with aircraft systems is something that is within reach”, he told me. “Furthermore, minimizing the time that faults are present on aircraft both improves flight safety and reduces delays and cancellations. Double benefit!”
We all want to reduce the total time that faults are present on an aircraft for operational efficiency, but there is an automatic benefit also to flight safety. Essentially, faults that are fixed quickly or pre-empted through early detection are not able to distract pilots in the cockpit.
Note that faults that are significant enough to ground an aircraft are not the topic of this article, as grounded aircraft obviously pose no flight safety risk. That said, when a grounding fault first appears during flight, it creates a distraction, so that remains a concern.
There are two broad categories of failures which do not ground the aircraft: Faults with which the aircraft can still fly, and Recurrent faults that elude attempts to repair them.
Aircraft are permitted to fly with certain functions inoperative, provided that conditions specified in a document called the Minimum Equipment List (MEL) are satisfied. The MEL is what may permit an aircraft to fly with an inoperative brake, as mentioned above. An aircraft flying under MEL is flying with the failure on board along with any associated cockpit distraction. The faster the failure is fixed, the fewer hours the aircraft flies with that cockpit distraction. As such, fast remediation is called for and first-time fix is the ideal outcome.
With recurrent failures, a problem has been evaluated with the possibility of a positive action being taken, perhaps a part replacement, resulting in the aircraft being declared serviceable. Sometimes the “fix” is a reset, whereupon the symptom vanishes, and the aircraft is declared serviceable. However, the bottom line is that the repair attempts have been unsuccessful, and the fault returns on subsequent flights.
Intermittent faults are notorious because they cannot be reproduced on demand, making conventional troubleshooting difficult. A well-experienced expert will be able to recognize the possible causes of an intermittent fault based on data captured during a flight and from behaviour observed by the pilots when the fault occurs – in flight, notably.
This is when “global field experience” comes into play. Most faults that happen on a particular aircraft have been seen and solved before – by someone, somewhere.
A guided troubleshooting tool like SpotLight® can pose insightful questions that lead to a solution that has been effective in the past. Alternatively, some clues may be found in the recent maintenance history of the aircraft, or across the fleet, using a text-mining tool optimized for aircraft maintenance like ChronicX®.
The above is what we call “Reactive Maintenance” because we are reacting to the appearance of a fault. Ideally, we would like to prevent faults from ever being noticed in the cockpit. To do so, it is necessary to see the failure coming and take action before it occurs. This requires us to detect signs of deteriorating health and to correlate that to the responsible component. Luckily, there are tools and methodologies that can be applied to such a task.
Condition-Based Maintenance (“CBM”) strives to detect components that are in the process of failing. Those signs are, generally speaking, low-level error message patterns and conventional monitoring for adverse trends in temperatures, pressures, vibration characteristics, etc. – that indicate signs of failure are starting to appear. The fact that the aircraft is producing early signs of failure puts this into the category of Condition-Based Monitoring, even if the cockpit is not yet aware.
Predictive Maintenance (“PdM”), on the other hand, strives to go one step further. It attempts to predict the future failure of a component on a perfectly serviceable aircraft when there are no signs of failure present. PdM detects future failures by monitoring the stream of system parameters on normally operating equipment. It is looking for “leading indicators” that have been established beforehand, through rigorous analytics focused on a given failure. The search for each leading indicator is, in itself, a sizeable scientific endeavour.
In summary, the goal with both CBM and PdM is to pre-emptively apply a repair before any effect is noticed in the cockpit. However, when failures do occur, software tools like SpotLight and ChronicX serve to minimize their impact in the cockpit by finding the cause in a short period of time. The less time that failures are affecting the cockpit, the lower the risk to flight safety.