If you wish to contribute or participate in the discussions about articles you are invited to join SKYbrary as a registered user

Go-around Decision Making

From SKYbrary Wiki
Article Information
Category: Runway Excursion Runway Excursion
Content source: Flight Safety Foundation Flight Safety Foundation
Content control: EUROCONTROL EUROCONTROL

The Background

In 1998 a Flight Safety Foundation (FSF) Study concluded that the failure to recognise the need for and then execute a missed approach was an important contributor to approach-and-landing accidents, including but not limited to those involving controlled flight into terrain (CFIT)[1]. This Study found that only 17% of accident/incident flight crews had initiated go-arounds when conditions indicated that go-arounds should have been conducted. It was stated that the majority of go-arounds were made because of weather (forward visibility, ceiling (cloud base), wind velocity and turbulence).

This Study led to the creation of the Approach and Landing Accident Reduction Toolkit, in which Stabilised Approaches[2] and Being Prepared to Go-Around[3] were highlighted as two interrelated key elements for improving safety. Central to both them was the principle of decision-making becoming more closely guided by SOPs.

Since these recommendations were made, many aviation safety regulators have introduced a requirement for aircraft operators to define and apply Stabilised Approach procedures suitable for their operations, but this fairly recent development has been merely a recognition of best practice on stabilised approaches which had already begun to be adopted by leading operators outside the USA well before the FSF Study, especially in Europe.

In 2013 at the Go-Around Safety Forum (supported by SKYbrary)[4][5], more recent studies and analyses were cited by presenters. These studies indicate that:

  • Between 3 and 4% of all approaches are reported/recorded as unstablised
  • Yet, only 3% of these result in a go-around being flown
  • In other words, 97% of unstabilised approaches continue to be flown to a landing contrary to airline Standard Operating Procedures (SOPs)

Risk Management

Since current accident rates are so low (e.g. 1:2,000,000), the vast majority of these 97% unstabilised approaches self-evidently land without incident. However, perhaps not always considered by the pilots making these decisions to land (or not to go-around) is the substantial increase in risk when flying an unstabilised aircraft down the final approach.

This raises many questions, including:

  • Why are pilots deciding to ignore their stabilised approach procedures?
  • What do pilots typically consider when making decisions to continue an unstabilised approach?
  • How can compliance with stabilised approach procedures be increased?

These questions should be considered by aircraft operators against the picture painted by their own operational performance data, from both Flight Data Monitoring (OFDM) Programmes and pilot self-reporting. Using this data, and the presentations from the Go-Around Safety Forum flight operations departments are encouraged to open discussions on this topic, including:

  • are our stabilised approach procedures properly predicated on approach and landing risk management?
  • if and when this is so, can demonstration of it to pilots be used to improve compliance?
  • if compliance is going to lead to more go arounds, are we sure that they will be flown at least as competently as any other 'normal' flight phase?

Non-Compliance with Stabilised Approach Procedures

Procedural non-compliance usually precedes a violation and a violation often precedes an accident. So, why do pilots continue with unstabilised approaches instead of going around[6]? A number of explanations have been put forward:

It’s OK

The fact that so many (97%) of unstabilised approaches are flown to a safe landing can reinforce several “landing-minded” views, including: unstabilised approaches as defined are not necessarily all equally risky.

It’s the SOP

Since many pilots have discovered that continuing and recovering from what is defined as an unstabilised approach so that a normal landing follows is likely can reinforce both personal and more widespread view that the procedures are unwarranted as specified. They may then be considered either as 'flexible' or as too restrictive and impractical. Some pilots may view the procedures as merely guidance which doesn’t need to apply to me right here, right now, in this particular situation. It may also be possible that the extant procedure is too complex and therefore difficult to use.

We’ll decide later

Because flying an approach to land is a dynamic process of continuous correction and decision making, there can be an undue focus (and this may be discussed between pilots during an approach) that there is always the option to go-around all the way to touchdown - which of course is true with or without stabilised approach procedures.

I don't agree with the detail of these procedures

Stabilised Approach 'Gates' where specific criteria must be met, vary considerably between operators, sometimes for entirely justified reasons. However, pilots who change employer may then encounter difference which they find are not satisfactorily explained by the operator. Even those who don't may feel that they haven't been given a credible explanation of the connection between the applicable SOP and safety. Stabilised approach gates vary in their number, their rigidity, the height above touchdown they are set at and whether they are different in different visibility conditions or for different types of approach. When asked in a recall survey to explain their behaviour, many pilots stated that they had been quite sure that they were capable of continuing safely whilst out of compliance with their stabilised approach criteria.[7]. It has been suggested too that encountering a situation which can destabilise an aircraft carries more risk if the physical and mental attention of the pilot is already focussed on flying an unstabilised approach. Since the pilot has already predicted that a safe landing will be possible, any additional complication such as a sudden adverse change in wind velocity may significantly impact on their capacity to react appropriately. It may even compound existing errors. If such a situation occurs very close to touchdown, then the time available to react will be just a few seconds and only a fully vigilant and attentive pilot will be able to react adequately.

I can see the runway

Being visual with your “goal” is a strong motivating factor in influencing a decision to continue, and when correcting from an unstable condition, it can be used to judge predicted progress quite efficiently. Therefore, seeing the runway and judging progress visually is how most pilots learned to fly, without any thought to meeting parameters such as rate of descent, drift angle, and heading. In fact when flying visually, it may be only power settings and speed that a pilot checks his instruments for regularly. However, it is interesting to note that just as many unstable approaches occur in Visual Meteorological Conditions (VMC) as they do in Instrument Meteorological Conditions (IMC),[8].

It’s our job

Commercial pressure and professional pride always appear somewhere in a pilot’s thinking (unconsciously or consciously) and this can influence the decision to continue or to go-around. Fir some pilots, this is implied by the tightly-defined fuel loading policies which are widespread these days.

Unexpected and/or Unwilling

For many pilots, go-arounds are rare events and the conditions that require a go-around decision rarely encountered. If pilots are not confident that they can fly a go around just as well as any other normal flight phase' then they are unlikely to be prepared to make a decision to go-around until they see it as inevitable - which is likely to make the transition to it much more difficult than if they had taken the decision at the most appropriate time. And whilst a thorough approach briefing [3] can help timely decision making, it does not guarantee it.

Inadequate Go-Around Specific Training

Without effective training, both in the classroom and in the simulator, the chances of a pilot automatically selecting go-around when it is the right option will be reduced. Decision making is often heavily biased towards the first option that comes to mind which is usually a continuation of the approach being flown.

Semantics

It has been suggested that that the specific word “unstabilised” can have a subtle effect on pilot decision making. Simply in terms of definition, “unstable” means loss of equilibrium, or in layman’s terms “about to topple”. However, the majority of “unstable” approaches are anything but this as they pass the qualifying threshold. It may be difficult logically to associate “unstable” with a situation where, say the speed is 15 knots above Vref, the thrust is at flight idle, the rate of descent is 1000 fpm and or the land flap is still in transit as the aircraft passes 500 feet, especially if whichever of these apply are gradually trending to the criteria which will make the approach "stable".

Go-Around Decision Making Flexibility

Placing more emphasis on the option to go around above as well as below the formal challenges of the stabilised approach gate(s) may have at least three major beneficial impacts:

  • making the option of a go-around both natural and 'normal' at any time
  • supporting dynamic decision making
  • reducing the cognitive capacity required to make correct decisions when under a high workload and thereby increasing the ability to respond to unforeseen circumstances.
Options for go-around decision making
Options for go-around decision making

From the representative profile above, it can be seen that at the time and place a go-around decision is required, a dynamic (rapidly changing) situation is very likely. This in turn will occupy a large part of a pilot’s attention and cognitive capacity. In such situations, it has been proposed that potential decisions need to be “armed” ready to be to be triggered by previously-considered threats so that decision making becomes reactive and automatic.

Underlying each decision making phase, as shown, is an effective procedure which can be learnt and applied in training, discussed and refreshed during briefings and “triggered” automatically when required.

Stabilised Approach and go-around procedures and policies can be seen as the ultimate in pre-planning for decision making and van be seen as containing when X then Y, and, if A then B clauses. Such an approach can also encourage effective cross-monitoring and task-sharing, by reinforcing the concept that go-around decision making is a collective responsibility resulting in a “command”.

Threats and Go-Around Decisions

Central to effective go-around decision making is a comprehensive knowledge and understanding of situations, threats and circumstances that should, or may, require a go-around to be flown. Whilst a stabilised approach procedure may highlight the symptoms, it is the threats to flying a stabilised approach, threats preventing a safe landing and the threats to decision making capacity that need to be fully understood. Typical threats include:

  • lack of preparation – “rushed” approach[9]
  • a late runway or approach procedure change
  • an inadequate approach briefing
  • challenging prevailing wind velocity
  • inappropriate energy management
  • inadequate traffic spacing
  • unfamiliar approach - maybe a straight in non-precision or circling
  • inappropriate aircraft configuration
  • runway surface condition
  • a predicted late touchdown point
  • unexpected runway occupancy after clearance to land
  • degraded aircraft systems status
  • the effect of fatigue
  • the effect of commercial and personal pressure (stress)

Each of these threats presents a different decision-making scenario depending where the aircraft is on the approach (see decision making profile above). For example, some wind-shear, or a runway incursion, whilst the aircraft is still at 800ft may demand a different response than if they occur at 100ft. Knowledge and understanding of potential threats must include all stages of an approach, especially the part below DA/DH where, the pilot mind-set is that a landing will now be possible.

Improving Go-Around Decision Making

As shown in the go-around decision making profile above, improvements to go-around decision making begin made well in advance of flying an approach and a lot of improvement can probably be made pre-flight! But pilots need the right kind of support and the following points need to be seriously considered:

  • Operation of a non-punitive policy for go-arounds and diversions is essential. This goes beyond not taking any action against pilots for going-around, but also requires management not to show displeasure! Care needs to be taken in how any de-briefing after such events is conducted.
  • In return for making the effort to properly understand the prevailing fuel loading policy, the pilot-in-command must not perceive undue interference with their final decision on fuel to be carried on a specific flight. It would serve no useful purpose, for example, for an airline to publish 'league tables' of fuel burn per route per pilot!
  • It would be prejudicial to safety to predicate personal-performance bonus schemes on flight punctuality
  • The stabilised approach procedure must be able to be justified to the pilots required to comply with it as proportionate risk management.
  • Procedures that clarify the role and responsibilities of any supernumerary pilots on the flight deck should be available and should clearly state when and how they might intervene of the operating crew are observed to be breaching significant SOPs.
  • Realistic and regular Go-around Training must be provided and must adequately cover the decision to go around from both specified decision points such as stabilised approach gates and instrument approach decision altitudes and the ad hoc decision.
  • Approach briefing procedures must require appropriate reference to the circumstances which might require a go around and the way it would the be flown.
  • The potential benefits of the Monitored Approach system for cross monitoring should be carefully considered.

Having decided to go-around is not the end of decision making! Executing a go-around places the pilots and the aircraft in a new situation, perhaps one that is rarely practiced and which contains new threats. Over 60% of go-arounds introduce increased risk[10], this increases to over 70% where the pilots had a problem on approach.

Go-Around Safety Forum (2013) Findings and Conclusions

The Findings and Conclusions from the June 2013 Go-Around Safety Forum held in Brussels contain many useful ideas for consideration including Strategies and Conclusions for various stakeholders. These conclusions are not necessarily recommendations, but are valuable for airlines to use as starting points to improve go-around decision making amongst their pilots. Some of these are summarised below.

Strategies to Ensure Go-Around Decision Making

  • Strategy 1 – Enhance crew dynamic situational awareness.
  • Strategy 2 – Refine the go-around Policy (stable approach parameters and stable approach height).
  • Strategy 3 – Minimise the subjectivity of go-around decision making.

Air Operator Conclusions

  • AO1 / AO2 – Develop SOPs to discuss instability threat factors during approach briefings prior to descent, and briefly throughout the approach.
  • AO3 – Develop ‘active’ communications procedures similar in concept to EGPWS or TCAS systems.
  • AO4 – Ensure unstabilised approach and go-around policies are clear, concise and unambiguous, including follow up procedures for non-compliance.
  • AO5 – Avoid directive or suggestive calls that may compromise on-going decision making, e.g., announcing, “Landing” at minimums.
  • AO6 – Re-define the stable approach criteria and stable approach height(s). In redefinition there is a valid argument to separate the profile (vertical and lateral) from the other stable approach criteria.
  • AO7 – Provide ongoing training to enhance awareness of psychosocial and management pressures that contribute to non-compliance during the approach phase.
  • AO8 – Cross monitoring effectiveness must recognise the importance of integrating low experience pilots into effective contribution to go around decision and 
execution.
  • AO9 – Pilots and their employers should understand that approach minima violation, is unacceptable because the evidence indicates that if a go-around 
then has to be made, the chances of a successful transition are reduced.
  • AO10 – The incidence of go-arounds should be continually tracked by Aircraft Operators based on a requirement for all PICs to file on the day of occurrence reports which explain the circumstances of the go-around. This will provide context to triggered OFDM events.
  • AO11 – Operations Manuals must contain a strongly worded policy statement which 
shows that, provided a full explanatory report is provided on any go-around made, no punitive action will follow. In addition, any ‘feedback’ will be 
provided in writing and be incapable of interpretation by a dispassionate expert observer as prejudicial to future operational safety.
  • AO12 – Pilots must be able to demonstrate that they are able to safely execute go-arounds commenced from high energy and low energy states at the point 
where the go-around decision is indicated.
  • AO13 – Pilots must be able to exercise tactical judgment as well as procedural 
compliance when deciding to go-around below the mandatory stabilised approach gate so that safe execution is not prejudiced by an inappropriate 
delay in the decision. Validation of this must be achieved by realistic training scenarios.
  • AO14 – Go-Around training should include a range of operational scenarios, including go-arounds from positions other than DA/MDA and the designated Stabilised Approach Gate. Scenarios should involve realistic simulation of surprise, typical landing weights and full power go-arounds.
  • AO15 – include lessons learned from operational events/incidents into go-around training.
  • AO16 – Clear guidance should be provided to pilots on how to act in respect of the three stages of cross-monitoring during approach, landing and go around i.e. - noticing/alerting/taking control. Observing members of augmented crews should have a clear understanding of their monitoring role.
  • AO17 – Pilot training to execute GA in automatic modes should be explicitly included and Aircraft Operator automation policy should address the go-around procedure.
  • AO18 – Pilots should have a clear understanding of how the pitch control system works on the aircraft type they fly. This should be validated by both theoretical testing and suitable simulator exercises conducted with full rather than reduced power/thrust available at typical landing weights.

Related Articles

Further Reading

References

  1. ^ FSF Safety Digest – Killers in Aviation: Facts about approach and landing and CFIT accidents
  2. ^ FSF ALAR Briefing Note - Stabilised Approaches
  3. ^ a b FSF ALAR Briefing Note 6.1 - Being Prepared to Go-Around
  4. ^ Go-Around Safety Forum – Presentations
  5. ^ Go-Around Safety Forum – Videos
  6. ^ Learning about unstabilised approaches through animations. Presentation by Joris Wenderich. 2013. Transavia.
  7. ^ Why are go-around policies ineffective. Presentation by Dr M Smith and Capt. Bill Curtis. The Presage Group Inc.
  8. ^ STEADES – Go-Arounds: in-depth analysis. Presentation by Giancarlo Buono. IATA.
  9. ^ Descent and Approach Profile management. Airbus Flight Operations briefing Note
  10. ^ Go-Around Accident and Incident Report Review. Captain Ed Pooley. 2013. The Air Safety Consultancy and FSF European Advisory Committee.


Back to the Go-Around Safety portal