Teasing Out The Human Factors
This is a carefully thought through analysis by the pilot of his response to an unusual incident in a rare vintage ex-military helicopter.
Dozens of people have asked what was going through my head when faced with my recent helicopter emergency. It was a very rare (unheard of!) mechanical failure for which there was no training, nothing in the technical manuals and no mention in the annual flight skills test. Was I frightened, did I panic, how did I cope?
As you can imagine, I’ve pondered over this quite a lot in the past month. It’s been cathartic to write out the events of the 23rd September 2016 for the obligatory reports as well as to keep the Westland Wasp Historic Flight followers and supporters informed by social media but, in all honesty, I can truly say that it really didn’t seem to be a big deal at the time. Even now I wonder why I chose to make a PAN calI rather than a MAYDAY! The truth was that I was so busy trying to work out what the problem was, and then what I could do to get safely back to mother Earth, that I had no time for panic or fear. I had to focus and not be distracted, there was a lot going on inside the helicopter and I also needed to keep my eyes outside because of the potential for harm to others on the ground and/or their property. After the event I was even concerned enough to ask Boscombe Zone about my radio calls, only to be told I sounded very calm and organised.
This article came about because I’ve looked retrospectively at the Human Factors (HF) associated with the event. A pure mechanical failure is definitely not caused by pilot error, but could HF have worsened the situation, or saved it?
My normal day job is as a doctor, and maybe there is crossover between what I do for a living and my attributes as a pilot. My medical skills are those of an anaesthetist and intensive care/emergency medicine specialist, familiar with managing the most complex and dangerous illnesses and injuries in extremely sick patients. Wearing that hat, I’ve spent many years dealing with emergency life-threatening events that need quick decision making followed by precise and rapid interventions. Speaking generically, pilots and anaesthetists share many common traits. They are both often type A personalities that thrive on
adrenaline, possess attention to detail, and are familiar with long periods of routine dullness interspersed with moments of hyperactivity, if not sheer excitement (sometime misnamed as ‘sheer terror’)! I learned long ago not to waste energy on panic or unfocused thoughts. Perhaps that helped me on the 23rd September but, having a crewperson who kept her cool and didn’t distract me, as well as an instant response from an excellent air traffic controller who let me get on with the flying, both must surely have helped enormously.
I found some more of interest in my CRM notes for the courses I’ve been teaching for many years. James Reason in his discussion of what he calls ‘heroic recoveries’ dismisses the possibility of finding one single most important contributing factor to a successful and survivable situation that may have seemed doomed to disaster. He talks of having the right people in the right place at the right time (which harks back to the origin of CRM – the ‘right stuff’ ). He adds that happy outcomes depend on acute situational awareness, personality, professionalism, teamwork and, sometimes, unexpected skills. Good fortune often also plays a part too, and yet the one certain barrier to effective recovery is a despairing lack of self-confidence.
In a discussion on ‘unlikely but possible hazards’ Reason explains that our heads are richly stocked with knowledge structures that are called to mind automatically by similarity-matching and frequency- gambling in response to situational-calling conditions. In other words, we tend to match what’s happening now to our previous experiences, and then use ‘tried and tested’ remedies to find a solution. Sometimes these unconscious search processes can lead us in to error, but it is more likely that what is called to mind in this way is going to be an appropriate response. Under pressure the mind defaults by remembering a solution that has been frequently used in this particular situation and, what is frequently used is, by implication, often very useful.
But what if we’ve never faced such a situation before and, even worse, we’ve never trained for it or even discussed it as a possibility? Here it helps to understand a bit more about the way the brain works, especially the science of decision making. Classical Decision Making (CDM) is traditionally described as being composed of a number of stages (figure 1):
Figure 1. Model of classical decision making
1 Identify a hazard or problem to be solved
2 Consider possible solutions
3 Assess the advantages and disadvantages of each
4 Select one option
5 Test its effectiveness
6 Return to point 1 if the desired effect has not occurred
However, real-life (so-called ‘Naturalistic’) Decision Making (NDM) in time-critical situations usually does not conform to the rational CDM approach and ‘experts’ with considerable past experience tend to follow a more recognition-primed decision model. As Reason described, experts match the current critical situation to similar past experiences, selecting an adequate course of action rather than comparing all possible solutions and selecting the one most optimal. However, modern CRM teaching embraces Jensen and Benel’s Aeronautical Decision Making (ADM) concept that sound judgement can also be taught and is not solely an attribute of very experienced experts. Their premise is based on the understanding of ‘hazardous attitudes’ (table 1) and the individual’s ability to modify their own behaviour.
Table 1. Hazardous attitudes and their antidotes
The US Federal Aviation Administration has for many years included the management of hazardous attitudes in their summary of the elements of good decision making:
Identify hazardous attitudes
Learn methods of modifying behaviour
Learn to recognise and cope with stress
Develop skills in risk assessment
Use all the resources available to the crew
Evaluate the effectiveness of your decision making skills
As pilots know only too well, the task of flying is characterised by the need to share attention between four main types of subtasks which Wickens (Wickens, C.D., & Morphew, E. (1997). Predictive features of a cockpit traffic display: A workload assessment. University of Illinois Institute of Aviation Technical Report (ARL-97-6/NASA-97-3). Savoy, IL: Aviation Res Lab.) describes as Aviation, Navigation, Communication, and Systems Management, but it’s well known that even highly experienced pilots may perform these subtasks less than optimally under conditions of high workload such as in a time-critical emergency. But what features of the task are direct causes of high intensity workload? There are seven contributing factors, most of which are recognisable by all aircrew and many others who work in demanding professions (such as doctors):
1 ‘Difficulty’ of the task
2 Number of tasks running in parallel (concurrently)
3 Number of tasks in a series (switching from task to task)
4 The time available for the task (speed of task)
5 Duration of task
6 Level of arousal
Task difficulty is the most obvious factor, and is easy to appreciate, for instance the mental addition of two small numbers is a low workload task, whereas multiplication or division of large numbers is a high workload task. As tasks become more difficult the brain, just like a computer, has to do more processing, more buffering, more searching for learned rules and more interrogation of both short and long term memory.
A well-defined model of the brain (figure 2) (Martin TE. Using functional magnetic resonance imaging to understand the mechanisms of consciousness. Aviat Space Environ Med. 1998; 69(12):1146–1157) describes how it receives inputs from many senses, and these pass through processors such as the visual spatial sketchpad and the phonological loop which result in perceptions in the sense modality that enter the ‘working memory’. If a task requires continual visual information input (such as flying an aircraft) then the visual spatial sketchpad will be in constant use, picturing and perceiving information coming from the eyes. The same is true of the phonological loop during verbal processing tasks (such as communicating with air traffic services). Additionally, because the brain can be considered as having only one executive CPU, then it follows that only one of the tasks can fully use it at a time. Sharing that CPU can lead to degradation of task management. The most optimum process of task sharing happens when a visual/spatial/manual task is being performed concurrently with a verbal/auditory task as well as at least one well- learned ‘automated’ skill (hence not requiring the central CPU). Ideal task sharing only occurs if the combined sub- tasks rely on separate sense organs for input, such as eyes and ears, and separate forms of output such as arm movements and vocal sounds.
Clearly, the need to process a task more rapidly in a time-critical situation adds difficulty to that task, and the time-crunch may itself induce error as the distraction of ‘counting the seconds’ interferes with constructive information processing and decision making. Hence the very awareness of reduced time available can itself add an extra attentional demand and will certainly increase the level of arousal.
One potential consequence of overload is an increase in selective attention, i.e. too much focus on one element of the task. Faced with increased task demands, individuals may choose to act rapidly at the expense of accuracy, or even shed some subtasks completely. On the other hand, too much focus in one area is responsible for loss of overall situational awareness that can lead to an even worse deteriorating situation. Training has clearly shown that the automation of skilled behaviour produced by over-learning (continual practice after proficiency has been reached) enables crew to have spare mental capacity in situations of high task workload.
I don’t think I was the ‘right’ person at the right time, etc… nor do I consider myself an above average pilot. However, there is likely something about my personality type that allows me to thrive in adrenaline-producing situations hence why I am professionally attracted to the serious end of the medical specialty spectrum and why I choose to fly and display a single engine vintage jet helicopter. There is no doubt that an annual skills test in which I must demonstrate responses to a wide array of emergency in-flight scenarios is extremely valuable. Year on year, I have learned more about flying, more about my own aircraft, and more about myself when performing under pressure. I have learned to trust and respect my examiners and instructors and yet feel able to question them when needed. When the moment came, I heard my examiner’s voice in my head ‘Aviate, Navigate, Communicate’, and so I did.
As for advice, I can’t really add to the words of wisdom already discussed from professionals who make a living from interrogating the psyche to dig deep and unravel those complex human factors. However, I can, perhaps share my ten take- home messages for that occasion when the totally unexpected happens:
1 Don’t let that moment be the first time you’ve ever thought about the possibility of such an emergency and how you’d deal with it.
2 Don’t let panic or despair enter your mind.
3 Consider the problem and concentrate in trawling your memories for the best possible solution.
4 Fly the aircraft as best you can, but don’t be too distracted from maintaining situational awareness.
5 Make the Pan or Mayday call as soon as possible.
6 Utilise all the resources that you have available, whether they be in the aircraft or on the ground.
7 Be methodical but don’t be afraid to be innovative if you’re still searching for a solution.
8 Practise, practise, practise. All types of emergency.
9 Rehearse multi-tasking whenever possible.
10 Evaluate and debrief after practice drills and real-life incidents.
The account of the accident
Friday 23rd September 2016 was a beautiful late summer’s day. Visibility was unlimited, and the sky was blue and almost cloudless. I was flying a Westland Wasp helicopter in the cruise at 1000ft and in straight, level and balanced flight at a speed of 80knots. The flying controls had all been behaving normally and all temperature and pressure gauges were ‘in the green’.
A few miles southwest of Salisbury I felt a 2 to 3 second vibration in the collective lever and soon discovered I had a total collective pitch failure. Finding I could neither climb nor descend was a little disconcerting, as I was in hilly and rising ground. Fearing this event may be the first part of a greater mishap, I opted to turn away from the undulating hills ahead of me and prepare for a precautionary landing as soon as possible.
I made a ‘PAN’ call to the air traffic controller at Boscombe Zone to announce intention to land asap. A ‘PAN’ call is like a ‘MAYDAY’ call but less urgent, i.e. used for emergency situations that are not immediately life threatening, but do require assistance from someone on the ground. A serious aircraft system failure that requires an immediate route or altitude change falls in to this category.
I soon found what looked like a suitable flat site in the nearby Chalke Valley, a somewhat small sports field adjacent to a larger flat field with low lying crops and no visible fence or barrier between the two. The chosen site was opposite a village so I’d be sure of responders in the event that the landing didn’t go well, but far enough away to minimise risks to others. I noticed a small play area but there were no children anywhere to be seen.
There was time for two circuits and to do some slow speed handling checks. This was valuable time. As I cruised in a figure of 8 around my selected safe site, I had time to brief my crewperson who secured the cabin, whilst I focused on trying to figure out how I was going to descend and land safely without collective pitch control.
Of the options that were running through my head, autorotation was not an option with no collective pitch and I considered converting from governed to manual throttle but the main rotor RPM was tending to over-speed, so I decided initially to fly by attitude, i.e. to decay some rotor energy and forward speed. I found that decaying the NR and pulling the nose up would give me a gentle rate of descent and also slow my forward speed, so after a final check of temperatures and pressures, RRPM, height and airspeed, and with no alarms or attention-getters, I made another radio call to Boscombe Zone to ensure I was still on radar and to inform the controller that I was making my final approach.
Before committing to the approach, I checked that the selected site was still the best option and that there were no people, animals or other obstructions to a clear landing area. I eventually converted to manual throttle at about 500-300ft for the final approach to the field. In doing so, and taking account of a flare at the bottom of the descent, I underestimated my ‘glide’ range and landed about 5m short of the sports ground.
The landing wasn’t particularly heavy but, without ability to cushion and a very low forward speed, the first wheel to touch caused a significant ‘oleo-bounce’ and there followed a couple of seconds of bouncing between all four corners as I shut down. The large oleos on the Wasp were designed to be useful when landing on heaving decks in very rough seas. The Wasp also therefore has a ‘negative’ pitch (actually just a lower pitch called ‘superfine’) used to pin it to the flight deck whilst the flight deck crew lash/lock the helicopter down.
On the one hand the oleos clearly mop up some of the kinetic energy of the landing, but they also caused the bouncing which seemed to resonate and grow remarkably quickly. This may have been exacerbated by the rear wheels being ‘toed inwards’ which is designed to prevent the Wasp rolling out of control in a roll-on landing.
With the marvels of modern GPS technology, my onboard iPad, fitted with Airbox Runway HD, plotted the whole thing. By exporting the file to Google Earth, you can see an enormous amount of detail recorded every 5 seconds of the flight. From these data, I’ve calculated that my descent rate was only 170ft/min, hence the landing wasn’t particularly heavy but, of course, without collective pitch, I couldn’t cushion the landing and I certainly couldn’t take advantage of superfine pitch to counter the bouncing.
The data also showed that my forward speed was 11kt. That’s only about 12.6mph but, with the rear wheels toed in and dense vegetation on the field, the Wasp couldn’t roll forward to dissipate the energy, which probably exacerbated the bounce.
Terry Martin's Wasp following its landing in the crop field.
Unfortunately, after what should have been a good landing, the bounce immediately tipped the nose up and the tail rotor dug in to the ground. The subsequent rapid yaw felt like being in a human centrifuge (I know, I’ve done that in my RAF days) and almost caused a roll- over, so I was fighting the instability to keep us upright for a few more seconds whilst I throttled down and shut off the fuel (HP cock). As the rotors wound down, all focus was on neutralising cyclic control to keep the main rotor tips from contact with the ground.
Quite quickly everything settled, we were down on all four wheels and nobody was injured. I completed the shut-down checks properly before we safely made our exit.
It retrospect, there are other possible ways this emergency may have been resolved, but hindsight is always 20/20 vision. Unfortunately, since foresight is always preferable, this wasn’t really possible as complete collective pitch failure is one emergency that doesn’t feature in the Wasp training manual, operating procedures or even in the Annual Skills Test. It’s that rare!
Dr Terry Martin is a Consultant and Associate Clinical Professor in Anaesthesia and Intensive Care Medicine and is the Medical Director at Capital Air Ambulance who provide a rapid response to requests for medical assistance and repatriation flights 24 hours a day, 7 days a week. He is also the owner and display pilot of a rare Westland Wasp helicopter.
The Human Factors section of this article was originally published in the Capital Air Ambulance flight safety newsletter Hangar 68 in December 2016.
human centrifuge (I know, I’ve
done that in my RAF days) and almost caused a roll- over, so I was fighting the
instability to keep us upright for a few more seconds whilst I throttled down
and shut off the fuel (HP cock). As the rotors wound down, all focus was on
neutralising cyclic control to keep the main rotor tips from contact with the
everything settled, we were down on all four wheels and nobody was injured. I
completed the shut-down checks properly before we safely made our exit.
there are other possible ways this emergency may
have been resolved, but
hindsight is always 20/20 vision. Unfortunately, since foresight is always preferable, this wasn’t really
possible as complete collective pitch failure is one emergency that doesn’t feature
in the Wasp training manual, operating procedures or even in the Annual Skills Test. It’s that rare!
Dr Terry Martin is a
Consultant and Associate Clinical Professor in Anaesthesia and Intensive
Care Medicine and is the Medical Director at Capital Air Ambulance who provide
a rapid response to requests for medical assistance and repatriation flights 24 hours a day, 7 days a week. He is
also the owner and display pilot of
a rare Westland Wasp helicopter.
The Human Factors section of this article was originally published in
the Capital Air Ambulance flight safety newsletter Hangar 68 in December 2016.