We rely on simulators to teach and rehearse the procedures and skills needed to handle serious emergency situations. Unavoidably, this practice sacrifices realism. The interaction of operational conditions make it extremely difficult to recreate real-world complexity. From the perspective of the typical pilot, each line-flying emergency feels unique while the simulator training profiles tend to follow familiar, canned scenarios. This is not a failure of simulator training. Instead, it is a byproduct of our current training, certification, and evaluation system. As an unintended consequence, crews demonstrate less success with handling line-flying non-normal events than they do in simulator scenarios.
Consider the ubiquitous V1-cut training event (trains the decision to reject or continue the takeoff at the critical decision point – the calculated V1 speed). Typically, it starts with an engine failure, fire, windshear, or a “bang”. The average pilot will experience and practice V1-cut training events well over 100 times during their career. They become quite competent at handling them. Let’s examine the reasons why. First, crews expect these failure events in the simulator, so they already are primed and ready to handle them. This reduces surprise and startle. Second, their training simulators events are scheduled well in advance. Crews review the procedures and often practice them in procedural trainers ahead of time. This elevates the procedural steps to the forefront of their working memory. Third, the onset conditions typically follow a limited number of canned profiles. Year-after-year, we practice these same V1-cut scenarios. This gives us the skills to complete them smoothly, accurately, and quickly.
Real-world V1 events prove to be far more nuanced and complex. Consider the following summary of 15 reported V1-related, non-normal events from the NASA ASRS database from a one-year period (November 2019 – November 2020). Note: This list is copied from my book, Master Airline Pilot (pages 438-439):
- The ATC Tower controller queried the crew during takeoff prior to V1 (crew had initiated takeoff without clearance). The Captain elected to continue the takeoff. The FO was returning from a 7-month absence (Report #1769610).
- A highly experienced Captain rejected their takeoff due to a “sudden veer” to the right. The Captain neglected to make a required “Reject” callout. Both the FO and the jumpseater were confused, but monitored and supported the Captain’s actions (Report #1761771).
- Both pilots noticed a flock of birds on the runway approaching V1. They elected to continue. They hit about 30 birds which inflicted damage to both engines. They returned for an emergency landing (Report #1759404).
- The crew experienced numerous engine anomalies during takeoff roll (second incident for the same aircraft on the same day). With too many indications to analyze, they decided to reject at V1 due to the accumulation of unknowns (Report # 1758495).
- The pilots became confused during taxi-out and made an intersection take-off when full length was planned. Approaching V1, they detected minimal runway remaining, but continued their takeoff. They would not have been able to stop if they had rejected at V1 (Report #1751568).
- The EECs (Electronic Engine Control computers) reverted to ALTN during taxi-out. The crew coordinated with maintenance, reset the EECs, and were cleared to continue. The autothrottles disengaged during takeoff as the EECs again reverted to ALTN. The crew reported startle effect, but continued their takeoff. When airborne, they experienced airspeed indication problems (Report # 1748317).
- The crew received a windshear warning at V1. They successfully rejected the takeoff (Report #1746586).
- The crew experienced anomalous airspeed discrepancies near V1 and rejected their takeoff. Maintenance discovered mud dauber wasp blockage in pitot static system (Report #1740194).
- The Captain/PM became tunnel-focused on possible engine underperformance and missed both the V1 and VR (rotation speed) callouts. The FO/PF made his own callouts and rotated. The Captain, who was heads-down, called for a rejected takeoff. The FO informed, “Negative, we are passed V1.” The Captain pushed engines to full thrust and continued the takeoff (Report #1739089).
- The crew reported that multiple avionics blanked during takeoff and rejected 50 knots below V1 (Report # 1720363).
- One turboprop engine rolled back, but didn’t auto-feather after V1. The crew continued their takeoff and returned for an emergency landing (Report # 1715079).
- The crew experienced multiple anomalies during rotation. While returning for an emergency landing, they had unsafe gear indications and multiple confusing electrical system anomalies (Report # 1704467).
- A spoiler warning activated at VR. The crew performed a high-speed reject. The same spoiler warning from the previous flight had been signed off (Report #1702333).
- The crew struck a large (10′ wingspan) bird approaching V1 causing a very loud bang. They rejected the takeoff. All tires subsequently deflated (the proper functioning of a tire overheat protection feature designed to prevent tire explosion) (Report #1700045).
- The FO lost attitude indications during takeoff into a late afternoon sun. They estimated normal pitch until airborne, then transferred control to Captain who had normal indications (Report # 1699712).
As I analyze these fifteen events, only three followed typically indications/progressions as we see then in the simulator. The remaining twelve crews encountered complex/startling/untrained indications that fell outside of anything they had probably ever practiced before. Of these, five crews failed to follow procedures or made significant errors. There might have been additional mistakes that were not documented in these self-reported summaries. While none of these events resulted in accidents, at least a third of them would be classified as failed procedures.
We conclude that when crews experienced real-world events that were similar to the V1-cuts that they practiced in the simulator, they accurately followed procedures. The training worked. When scenarios strayed from canned profiles, their decisions and actions became less consistent. Despite how often they had practiced V1-cuts in the simulator, too many of these crews made significant procedural errors, experienced startle/surprise, or became confused by the indications.
Another problem is with scope. V1-cut training in the simulator centers on the moments immediately before and after reaching V1. Focusing just on engine failures, the real world reality is that most events occur well outside of the few seconds near V1. Engine failures outside of this time window require procedural modification. For example, an engine failure that occurs during climbout passing 500’ requires that we modify the typical V1-cut takeoff profile. So, while simulator training provides excellent practice, real-world events add more dimensions that challenge our expectations and complicate our decision making.
Another consideration is the distraction generated by operational complexity. Non-normal events require focused attention and crew discussion to accurately diagnose and solve the problem. This means that we’ll need to hold our train of thought as we diagnose and remedy the problem. In simulator practice, instructors typically reduce distractions or place the simulator on “freeze” to create an ideal environment for crew coordination. While flying, operational tasks and outside disruptions constantly interrupt us. ATC asks us questions or the FAs call up wanting to know what is going on.
How skills do we need to teach pilots for handling real world emergencies? The current system of simulator practice unintentionally leads pilots to hold different mindsets between simulator training and line flying. The most obvious split is that we expect to always experience non-normal/emergency events in the simulator, while we almost never experience them in the aircraft. In the simulator, expecting that something will go wrong primes us to mentally prepare. Before the event begins, we visualize the emergency, the indications that we should detect, and the steps we need to follow. When the instructor initiates the event, our mental preparation helps us to quickly make the mental switch from routine, normal flying to exceptional, non-normal event handling. Conversely, everyday flying promotes a mindset that subconsciously assumes everything to go normally. When something does go wrong, we often experience startle, surprise, and debilitating biases.
So, what can we do to solve this mismatch? What is the Master Class skillset? We start with the realization that we handle emergencies rather well in the simulator. We just need to find a way to carry our simulator world competence into the real world. When that rare emergency happens, we need to quickly switch from our routine, normal-flying mindset to our non-normal, event-handling mindset. This switch is guided by three distinct skills. We start with recognition. We need to detect and acknowledge that something exceptional has occurred. Often, mishap crews spend too much time wondering what happened, questioning how they misdiagnosed the problem, or downplaying its severity. Next, we must accept that that the non-normal event will upset our established game plan. This means that we will need to reduce our attachment to our original game plan to prevent succumbing to plan continuation bias. Even if we are established on final approach with the runway in sight, our best course of action might be to go around, sort out the problem, run checklists, and come back around for another approach. Third, we need to reorient our mental processes (detecting indications, recognizing patterns, applying meaning, and decision making) from our everyday flying mode to our emergency event handling mode. Our everyday habits and decision making may not work and may impede reaching a successful outcome. I devote several chapters to this process in my book.
I theorize that most crews that mishandle line-flying, non-normal events fail to make this switch in their mindset either quickly enough or accurately enough – either individually and as a crew. An additional complication is that aviation does not follow a simple dichotomy between normal flying mindset and serious emergency mindset. It is actually a continuum that requires us to master a range of responses. We will unwrap this concept in the next discussion topic.
