1 INTRODUCTION

Standard Operating Procedures (SOPs) provide the “common ground” to conduct a safe flight (Cahill et al., 2017). They define the script for the interactions between operators (Degani et al., 1999), aircraft sys-tems (Mauro et al., 2012) and the operational envi-ronment (Barshi et al., 2016). The workflows de-fined by SOPs should be appropriate for the operation (FAA, 2003) so that tasks can be complet-ed in a logical sequence to avoid any overlaps or dis-ruptions (Kourdali & Sherry, 2016b). However, the human factors literature highlights that procedures frequently lack a clear operational logic (Degani & Wiener, 1998), are poorly designed (Funk et al., 2009), inherently incorrect (Degani & Wiener, 1997) or do not adequately address human factors limitations (Kemény & Popp, 2016). SOPs that are not practical to use can impose multiple task demands (Loukupoulos et al., 2003, 2009) forcing the flight crews to deviate from the procedure and to adapt the task sequence to the operational environ-ment (de Terssac & Chabaud, 1990; de Brito, 1998; BEA, 2013). These deviations are not necessarily re-lated to complacency (Cahill et al., 2014). Instead, cockpit operations do not always follow the work-flow linearly, and therefore not all operational con-ditions can be considered in the procedure (Pélegrin, 2007; Cahill et al., 2014; Maille, 2016). However, situation-dependent deviations from SOPs can vary from flight to flight and even from moment to mo-ment (Loukupoulos et al., 2009). They undercut the

execution of habitual tasks which increases the vul-nerability to human error (Loukupoulos et al., 2009). A good example in this respect is the Go-around (GA). It is a normal flight manoeuvre (GA Safety Forum, 2013) conducted by flight crews if an ap-proach cannot be completed to a landing (Manchan-da & Sikora, 2016). However, for most commercial pilots the GA is a rare event (Manchanda & Sikora, 2016): On the Airbus A320 fleet, the GA rate equals to 0.29 percent which corresponds to 1 GA per 340 approaches (Marconnet & Roland, 2014). Regular training and compliance to SOPs guarantee that the A320 GA procedure flows well if performed correct-ly (ATSB, 2014; Marconnet & Roland, 2014). Nonetheless, promoting the GA as a normal and scripted flight phase does not mean that there are no safety risks (GA Safety Forum, 2013): High work-load caused by the startle effect or unexpected task demands frequently contributed to a loss of state awareness during GA (BEA, 2013) in several safety-related events on the A320 fleet (e.g. Kingdom of Bahrain CAA, 2000; ATSB, 2007, 2014, BEA, 2000, 2007, 2009). Whatever the type of recent GA accident, investigative findings often point out an imperfect performance of SOPs (Dehais et al., 2017) and shortcoming in Crew Resource Management (CRM) (BEA, 2013). We argue that these deficien-cies are caused by the design of rigid crew function allocations. The industry has recently changed the nomenclature from a Pilot Not-Flying (PNF) to a Pi-lot Monitoring (PM) to encourage participation but without redesigning SOPs or providing the guide-lines on the required skills to be an effective PM

OneTeamCockpit – Enhancing the Flexibility of Flight Deck Procedures during the Go-around

Tim André Schmidt (Airbus Defence and Space GmbH), Dr Jim Nixon (Cranfield University), Houda Kerkoub Kourdali (George Mason University), Christof Kemény (Lufthansa Cityline), Dr Christian Popp (JetBlue)

ABSTRACT: The Go-around (GA) is a normal phase of flight strictly defined by standard operating proce-dures to ensure a safe, efficient and predictable flight operation. However, high workload caused by the startle effect and unexpected task demands frequently contributed to a loss of state awareness during GA. Whatever the type of recent GA accident, investigative findings often point out shortcomings in crew resource manage-ment and imperfect performance of procedures. These deficiencies are caused by the dynamics and complexi-ty of real-world operations that are not necessarily addressed in the design of flight crew function allocations. Therefore, we propose the OneTeamCockpit Concept that goes beyond linear workflow accounts by an equal and active engagement of both flight crew members. The concept was applied to the Airbus A320 GA proce-dure. The OneTeamCockpit procedure redistributes function allocations and implements an adaptive task management to improve the resilience of the design in non-normal GAs. Both procedures were tested in the Air Vehicle Simulator of the German Aerospace Center (DLR) together with airline and test pilots. The One-TeamCockpit procedure simplifies crew coordination while lowering crew workload. In particular, the adap-tive task management provides a promising concept to enhance the flexibility of flight deck procedures. Keywords: OneTeamCockpit , Airbus A320, Go-around, Standard Operating Procedure.

(Kemény & Popp, 2016). Consequently, current GA procedures often induce workload imbalances by designating the PM as a mostly passive participant (Kemény & Popp, 2016) who monitors the flight path and performs tasks on command while the Pilot Flying (PF) operates the aircraft, manages automa-tion and conducts the necessary callouts (Dehais et al., 2017). Flight path monitoring and mode aware-ness are safety-critical (Reynal et al., 2016), but re-cent studies (Dehais et al., 2017; BEA, 2013) have indicated that monitoring during GA is often insuffi-cient due to high workload leading the PM ill-equipped to handle sudden onsets of concurrent ac-tions. Therefore, insufficient CRM notably with re-gard to contribution of the PM is still a key factor in most GA accidents and incidents (BEA, 2013). Thus, we suggest to go beyond linear workflow ac-counts (Cahill et al., 2014) by defining the “ideal-ised” notion of a crew composition that adequately addresses human factors limitations while enhancing the flexibility of the design (Degani & Wiener, 1997). To achieve this objective, we have explored new function allocations by applying the One-TeamCockpit Concept developed by Kemény & Popp (2016). The hallmark of the OneTeamCockpit is based on the idea that crew performance can be maximised when tasks demands are balanced within the crew capabilities and resources as outlined by Dornreich et al. (2017). Therefore, the One-TeamCockpit aims at an enhanced equal and active involvement of both crew members to handle the dynamics and complexity of the GA.

2 METHOD

2.1.1 Procedure Review The Airbus A320 GA procedure was reviewed from a human factors and pilot’s point of view to under-stand potential deficiencies associated with the de-sign of crew function allocations. The analysis was based on the review of the A320 GA procedure (Airbus, 2017) including flight guidance systems de-scriptions (ATG, 1999; Norden, 2010; Granger & Jeanpierre, 2011; Marconnet & Roland, 2014; Fernandez et al., 2017). We have analysed the pro-cedure in three prototype scenarios at or below min-imum altitude. Prototype scenarios are known situa-tions where distraction or interruptions are likely or where tasks may not be performed in the prescribed order (Loukupoulos et al., 2009). The baseline is the GA at minimum altitude due to a loss of the visual reference to the runway. After that, two scenarios including a rejected landing and a revised missed approach by air traffic control were developed. The scenarios are operationally relevant since they are ei-ther part of airline training or were already applied in other GA studies (e.g. Dehais et al., 2017; Nie-

dermeier & Buch, 2016; BEA, 2013). We visualised the operational sequence in a Human-Machine Inter-face Sequence Diagram (Kourdali & Sherry, 2016b). The analysis was validated in informal interviews together with A320 line pilots (N=6, flight experi-ence between 1700 and 9500 hours) and by observa-tions of various GAs in a full-flight simulator.

2.1.2 Procedure Redesign

The redesign builds on the review of the Airbus A320 GA procedure. The purpose of this study is to demonstrate that the OneTeamCockpit Concept pro-vides a complementary approach to enhance the flexibility of crew function allocations during the GA.

2.1.3 Testing

Table 1. Participants and Total Flight Experience ______________________________________________ Test FO CPT Line/Test Pilots ______________________________________________ Test 1 2500 h Designer (300 h) A320 Airline Pilots Test 2 1450 h 9500 h A320 Airline Pilots Test 3 2900 h 7000 h A320 Test Pilots ______________________________________________ Three tests were conducted in cooperation with the DLR using the AVES A320 full-flight research sim-ulator (Richter, 2014) to analyse the effect of certain design modifications on workload. The six partici-pants as outlined in Table 1 performed three scenar-ios with both procedures. All crews manually con-trolled the aircraft at GA initiation and engaged automated modes as required which represents one of the most applied GA profiles in line operation (see Maille, 2016). We have altered the scenario se-quence and the order of the involved procedures in each test to counterbalance the training effect. Train-ing scenarios before the test were conducted to fa-miliarise with the simulator and workload question-naire as proposed by Harris et al. (1992). No workload ratings were recorded for the SOP design-er of this study due to the familiarity with the re-search setting. The workload was assessed after each scenario by applying the Bedford Workload Rating Scale (BFRS) (Roscoe & Ellis, 1990). The scale is “easy to use” (Roscoe, 1984) and provides a valid and reliable measure of workload (Corwin et al., 1989). Ratings specifically referred to primary tasks which were modified in the redesign. These tasks were grouped together into three ratings for each pi-lot. We have outlined the piloting and non-piloting tasks which resulted in two different workload ques-tionnaires for the PF and PM. Also, workload scores by two Human Factors Specialists were recorded af-ter each scenario. These ratings referred to the over-all workload of each flight crew member for a given scenario and were used to support the comparison between both procedures and to improve the impati-

Figure 1. Airbus A320 GA Procedure lity of the evaluation (Wainwright, 1987). Since the BFRS is not diagnostic (NASA, 2014), we have ana-lysed the crew feedback and video recordings of the test to understand what may have caused the work-load.

3 RESULTS

3.1.1 Procedure Review GA at Minimum Altitude (Scenario 1) Figure 1 shows the A320 GA procedure which was divided into four segments to easy the subsequent analysis. The task sequence corresponds to the base-line scenario which represents the GA at minimum altitude. The aim of this study is not to explain the procedure in detail. Instead, we focus on the func-tion allocations of the design especially in the first segment when initiating the GA. The task distribu-tion follows the Airbus philosophy (Airbus, 2006; Owens, 2013): The role of the PF is to fly the air-craft while controlling the flight path (Dehais et al., 2017) including energy management and supervision of automation as well as monitoring of tasks per-formed by the PM (Airbus, 2006; Flight Safety Foundation, 2009). Thus, the PF mainly performs psychomotor actions (e.g. thrust lever or sidestick inputs) to control the aircraft’s attitude and energy management. This task changes to visual monitoring the flight path after engaging the autopilot (AP). The teamwork is limited to commands (e.g. Go-Around Flaps, Gear up, Flaps One, Flaps Zero) by the PF. The PM performs a dual role since he or she moni-tors system related states, but also communicates with Air Traffic Control (ATC), monitors tasks per- formed by the PF and performs tasks on command

(Airbus, 2006; Flight Safety Foundation, 2009). Based on our observations we identified that the PM conducts these tasks often concurrently, especially when changing the configuration (Flaps and Landing Gear), Autopilot/Flight Director Modes (AP/FD) and when communicating with ATC. The latter case appears uncritical since the PM can focus on the Primary Flight Display (PFD) visually while listen-ing and communicating with ATC by using different modalities. However, for the configuration and AP/FD management, it may result in a detriment of monitoring if no flexible adjustment of the monitor-ing strategy in response to the task demands occurs (Anderson et al., 2017). An example is given in Fig-ure 2 indicating the PM defers the monitoring task when retracting the flaps at GA initiation. Our ob-servation is supported by two studies (BEA, 2013; Dehais et al., 2017) demonstrating that the role defi-nition of the PM results in broad spread of visual at-tention with briefer dwells to check that flight pa-rameter are within range. Looking into the crew coordination during GA, the flight crew moves be-tween periods of working alone, in parallel and to-gether. However, team-based tasks are performed based on triggering events (Airbus, 2006). The trig-ger opens the process gate (Cahill et al., 2014) and initiates an action block (Airbus, 2006). The per-forming operator closes the process gate by inform-ing the other agent about the aircraft system state (Cahill et al., 2014). Each action block consists of several human-machine interface (HMI) loops which together may form a shared mental model (SMM) loop (Kourdali & Sherry, 2016a). An example is given in Figure 2. By observing the interaction of the pilots with the flight deck interfaces during GA we identified 17 HMI-loops and seven SMM-loops that are mainly associated with configuration chang-es and mode awareness. Four SMM loops and nine

HMI-loops are tightly grouped in the first segment. The tight grouping is critical considering that most actions are performed during GA initiation. Figure 2. Human-Machine Sequence Diagram for the Flap Re-traction at Go-Around Initiation (Scenario 1) One flight crew member may predict needing to per-form specific team-based actions, but the timing is controlled by the other agent (Pritchett et al., 2014). Considering the diversity of multiple tasks per-formed by the PM, he or she might delay triggering the next action until feeling comfortable to devote the required attention. This condition may end in a delay of HMI and SMM-loops which results in fre-quent overlays of status reports by the PM and commands by the PF (Degani & Wiener, 1994). Callouts are the basement of the coordination (Pélegrin, 2007) but they can be a source of indirect pressure resulting in a potential increase in workload at a time when the flight crew is already occupied with multiple tasks (Loukupoulos et al., 2009). Thus, any overlay of numerous callouts may result in an overloading of the reporting channel (BEA, 2013) with a potential breakdown in teamwork (Degani & Wiener, 1994). Moreover, most trigger-ing events are initiated by the PF which exacerbates the condition for the PM when setting own priorities in task execution. This result supports the finding by the BEA (2013) that the task flow in particular dur-ing GA initiation requires an almost “ideal” crew coordination. We agree that the Airbus procedure can be executed smoothly if both flight crew mem-bers are proficient in coordinating their tasks. How-ever, the test results of this study have shown that

failures in crew coordination and overlays between status reports and commands occurred even at a sat-isfactory level of workload (BFRS ≤3) besides be-ing well familiar with the procedure. Other studies (Dehais et al., 2017; BEA, 2013) confirmed frequent observations of shortcomings in crew coordination during GA. Rejected Landing (Scenario 2) Figure 3. Human-Machine Sequence Diagram for the Flap Re-traction at GA Initiation (Scenario 2) A rejected landing is a GA after touchdown (Airbus, 2005). No specific procedure is given for this sce-nario. Instead, Airbus acknowledges that the rejected landing may require an adjustment of the task se-quence since the flap retraction should be delayed to provide sufficient lift until the aircraft achieves the target pitch attitude (Airbus, 2005). Thus, the reject-ed landing results in higher demands on controlling and monitoring the flight path. Figure 3 shows that the delay in flap retraction leads to a modification of the callout Go-Around-Flaps to Go-Around. The PF commands the flap retraction when the aircraft achieves the target pitch attitude (Airbus, 2005) which results in a delay of HMI and SMM-loops in comparison to the baseline scenario. However, the routine may lead to the automaticity of commanding Go-Around-Flaps resulting in an inadvertent flap re-traction (Barshi & Healy, 1993). Nevertheless, this event is not necessarily critical considering that de-pending on the performance and runway available,

the aircraft might provide sufficient lift to get air-borne. Finally, a takeoff configuration warning due to the trim and flaps setting at GA initiation can oc-cur which may capture attention and disrupt the task sequence (BEA, 2013). Overall, it seems that the PM is potentially overloaded due to monitoring of mode changes, engine parameter and the flight path while verifying the warning and performing configuration changes. Revised Missed Approach (Scenario 3) Figure 4. Human-Machine Interface Sequence Diagram for the Interactions with the Flight Control Unit (Scenario 3) The third scenario introduces an element of surprise by assigning a new missed approach including a re-vised heading (20 degrees offset to the runway) and a restrictive level off altitude (reduction by 50 per-cent) at GA initiation. The scenario is similar to the study by Dehais et al. 2017 and the BEA (2013). The simultaneous climbing turn with an altitude lim-itations increases the time pressure and makes the energy and flight path management more difficult (Dehais et al., 2017). Also, the flight crew cannot re-late to the pre-programmed missed approach of the Flight Management System which requires the use of the basic autopilot modes. From a human factors perspective the third scenario demonstrates how the procedure sequence can impair the monitoring abili-ties and prospective memory (Toglia & McDaniel, 2016; Anderson et al., 2017) of the flight crew: The operational sequence is illustrated in Figure 4. To comply with the procedure and the golden Airbus

rule “Aviate, Navigate, Communicate” the PM should perform the configuration change before changing the heading and altitude at the Flight Con-trol Unit. Thereafter, the PM should read back the clearance. Thus, the PM must defer tasks and encode an intention to remember the heading and altitude changes as well as the readback of the clearance (Dodhia & Dismukes, 2003, 2009; Loukupoulos et al., 2003; Dismukes, 2012). However, the time to encode is limited, and other ongoing tasks like the flap and gear retraction divert the PM further away from remembering the required tasks (Dismukes, 2012). Dehais et al. (2017) demonstrated that the at-tention shift to perform the configuration change prevented the PM frequently from monitoring the flight path. As a result, the PM indicated a subopti-mal monitoring ability (Dehais et al., 2017) which is critical considering that the PF might have to fly against the flight director bars since the PM cannot enter the required lateral and vertical inputs at the Flight Control Unit immediately after receiving the clearance. Reducing the time for encoding by divid-ing the attention has been shown to impair the pro-spective memory performance (Einstein et al., 1997; McGann et al., 2002). As a result, prospective memory can be impaired if the flight crews adhere to the procedure sequence (Dismukes, 2012) resulting in a potential loss of the required heading and alti-tude information. In this event, further clarification within the crew and with ATC is required which in-terrupts the task sequence leading to an increase in taskload (Liu et al., 2017) and a possible delay of subsequent tasks (BEA, 2013). Dehais et al. (2017) and the BEA (2013) confirmed this hazard by ob-serving various lateral and vertical deviations from the flight path, mostly due to poor interaction with the Flight Control Unit and suboptimal monitoring. These deviations appear to lead to numerous acci-dents observed during GA (BEA, 2013). Summary The A320 GA procedure has proven its operational feasibility. Airline pilots are usually well familiar with the task sequence. However, concurrent task management when retracting the flaps as well as during changes of the AP/FD modes can result in a detriment of flight path monitoring notably by the PM. The design requires an almost “ideal” crew co-ordination while being in a startle effect situation. The tight cueing of HMI- and SMM-loops is vulner-able to potential breakdowns in teamwork. Adjust-ments of the task sequence during the rejected land-ing are not reflected in the design which requires regular airline training. Finally, adherence to the procedure can impair prospective memory indicating that the procedure is too inflexible to handle the complexity and dynamics of a revised missed ap-proach by ATC.

Figure 5. Airbus and OneTeamCockpit GA Procedure

3.2 Procedure Redesign

3.2.1 OneTeamCockpit Concept

The principle of giving each crew member the whole competencies except for flying the aircraft is essen-tial within the crew. The redundancy of skills pro-vides synergy for a safe and efficient GA. However, the specialisiation of both flight crew members on specific tasks does not exclude synergy but becomes necessary in line with higher operational demands and complexity. Therefore, procedures that describe the task sharing of multi-crew cockpits based on the physical set up of the flight deck are not sufficient because they do not adequately consider the dynam-ics and complexity of the operational environment (Kemény & Popp, 2016). Thus, the OneTeamCockpit Concept was introduced as a holistic approach to the development of human factors based procedures (Kemény & Popp, 2016). Because the goal is to achieve a OneTeam rather than a flight deck occupied by the two pilots, we aim at equal and active engagement to balance workload and to improve crew performance (Kemény & Popp, 2016). In this study, we achieved the OneTeam by new function allocations that allow for parallel workflows. We use the term Pilot Supporting (PS) instead of PM that is more appropriate to enhance collaboration and to encourage a OneTeam (Kemény & Popp, 2016).

3.2.2 Design Features Delayed Flap Retraction The first objective of the redesign was to improve the monitoring capabilities of both flight crew mem-bers at GA initiation. Maille (2016) has already shown that flight crews use different techniques when retracting the flaps and landing gear and occa-sionally adapt the sequence when using the Airbus GA procedure. This finding was used to improve the monitoring function at GA initiation by delaying the flap retraction after the Positive Climb callout. This prevents the PS from an attention shift away from the PFD/FMA to the flap lever. As a result, we can provide free space for a new task explicitly named as “monitoring the aircraft’s attitude” performed by the PS. Also, the FMA callout occurs earlier to provide mode awareness immediately at GA initiation. We argue, that these activities have priority for safe exe-cution independently of the scenario. It is assumed that the aircraft performance is adequate. However, the delay in flap retraction may result in a flaps overspeed. Those performance aspects will be con-sidered in the test. However, the two potential bene-fits of the redesign are: A mitigation of hazardous events associated with somatogravic illusions (Ludlow, 2016) at high g-loads as both flight crew members focus on flight path monitoring. Secondly, a hazard mitigation of an inadvertent flap retraction on the ground during the rejected landing since the configuration change is already delayed by design so that pilots do not have to consider or decide on the flap setting while being in a demanding situation.

Figure 6. Crew Coordination during Flap Retraction at F-Speed (Left Airbus, Right OneTeamCockpit) Parallel Workflows The next step was to design a procedure that is less vulnerable to failures in crew coordination due to overlays of status reports and commands. In the Air-bus SOP, the retraction of the flaps and landing gear is a shared task (see left side Figure 6). The PF initi-ates the SMM-loop, and the PM closes it by a verbal response. This design does not only require a good crew coordination but is also vulnerable to task in-terference, habit intrusion and misordering leading to a potential misselection of the flaps or landing gear. These actions slips are difficult to detect be-cause the execution by the PM is a highly practised, routine and largely a physical action (AAIB, 2016b) that itself is not under conscious control (AAIB, 2016a). However, it seems that these slips occur more frequently than we expected: Based on Flight Data Monitoring, Maille (2016) has identified 31 out of 439 cases in which the flaps where retracted out of the desired sequence during the GA. Recent inci-dents on the A320 fleet (Hradecky, 2017; AAIB, 2016a,b) during takeoff confirmed that besides the correct command Gear Up, the PM selected the flaps lever inadvertently. Therefore, the One-TeamCockpit aims at a higher level of gradual atten-tion by an active involvement of the PS (Kemény & Popp, 2016). The right side of Figure 6 shows that we fully allocated the configuration change to the PS without prior verbal confirmation or command by the PF. This is the key feature of introducing parallel

workflows. The PF can fully focus on flying the air-craft while the PM performs tasks independently. In this way, the PS becomes an active player on the flight deck that stays in the loop of flying the air-craft. The design modification not only affects the GA initiation but also changes the configuration management after passing the F- and S-Speed. The PS triggers and executes the action based on the pro-cedure sequence and the F/S- Speed indicated on the PFD. The PS informs the PF about the beginning of the action block by the callout Speed Checked. The PF can still interfere by announcing Hold Flaps or by informing the PS before reaching the scheduled flap retraction speed. Also, the PF may use the ap-proach briefing to inform the PS well in advance about the intended flap setting. Thus, the configura-tion change remains a shared duty, but while being in a dynamic and complex scenario, it is triggered and executed by the PS. However, the possibility to interfere is still necessary since the configuration change influences the flight path and energy man-agement of the aircraft which remains a primary task of the PF (Active Pilot Monitoring Working Group, 2014; Cahill et al., 2014). Adaptive Task Management A flexible design needs two modifications: First, the design of the task sequence should be usable across various scenarios. This was achieved by introducing the new function allocation to the PS related to the configuration change management.

Figure 7. Adaptive Task Management of the Pilot Supporting The callout Go-around-Flaps by the PF was modi-fied into Go-around. The command does not change in the three scenarios. However, the procedure de-sign process cannot consider all possible scenarios (Pélegrin, 2007) and even a flexible design will re-quire adjustments of the task sequence in some sce-narios. However, this study demonstrates that ad-justments can be controlled by the design of the procedure to support the flight crews in non-normal GAs. Thus, the second modification introduces an Adaptive Task Management (ATM) as shown in Figure 7. The ATM considers that it is intuitive to change the AP/FD modes immediately after receiv-ing a revised clearance by ATC. Therefore, two in-tegrated tasks were developed: ATM1 combines the flap and gear retraction while ATM2 includes the flight guidance phase. The ATM mainly affects the tasks of the PS. The normal sequence during the GA is ATM1 fol-lowed by ATM2 (left side Fig 7). However, in the event of an unexpected change in the missed ap-proach as demonstrated in scenario three, the PS can perform ATM2 before ATM1 (right side Fig. 7). It is assumed that this design provides a faster input at the Flight Control Unit so that it is not required to fly against the flight director bars for a prolonged time. However, once an ATM is initiated, the PS should complete the tasks without mixing it with the next ATM. An impending overspeed situation of the flaps is an exception to this rule. Thus, the PS can initiate ATM2 and performs the flap and gear retrac-

tion whenever required. The PF might inform the PS about the impending overspeed situation with the callout Speed.

3.3 Testing

3.3.1 Workload Assessment The objective of the test was to analyse the effect of the modified task sharing on workload distribution across the three scenarios. Workload ratings referred to three design modification: Two workload ratings were recorded at GA initiation: Flap and gear retrac-tion as well as changes of the AP/FD modes. The last workload score combines the configuration change management after passing the F-/S-Speed. The only valid interpretation of the workload scores is an analysis of the histograms and scatterplots (Wainwright, 1987; NASA, 2014). Therefore, Figure 8 shows the scatterplots of participant ratings for the PF and PM/PS applying the Airbus and One-TeamCockpit procedure in the three defined scenar-ios. The scatterplots indicate increasing workload ratings for both flight crew members which was ex-pected since the scenarios were representative from low (Scenario 1) to high (Scenario 2 & 3) workload. The maximum workload scores for the PF are higher for both procedures except for the third scenario. The effect of the OneTeamCockpit design on work-load is particularly apparent in the third scenario. Based on pilot feedback the lower workload ratings of the PS are attributed to the ATM which allows setting own priorities. The lower workload scores of

the PF in the third scenario can be assigned to the faster inputs of the PS at the Flight Control Unit when performing ATM2 before ATM1. Based on the test data, the mean time until entering the correct heading by the PS could be reduced by 12.8 seconds when applying ATM2 before ATM1. Figure 8. BFRS Ratings by Participants for the PF (lower) and PM/PS (upper) across the two Procedures and three Scenarios Presumably, this allowed the PF to follow the flight director bars which decreased the workload during manual flight. However, there is no apparent change in workload of the PM/PS for both procedures when comparing the first and second scenario indicating that higher demands on monitoring the flight path and verifying the takeoff configuration warning have not affected the workload when retracting the flaps and landing gear in both procedures. The higher de-mands on controlling the aircraft after the touch-down are consistent with the higher workload scores of the PF in the second scenario for both procedures. Also, the PF shows higher workload ratings for both procedures in the first scenario. However, these scores are partially influenced by a confounding var-iable of system functions that were either not availa-ble in the research simulator or resulted in an unde-sired aircraft state which required a recovery by the PF. Table 2 shows that flight crews indicated a higher cumulative percentage of workload ratings between 1 and 3 when applying the OneTeamCockpit proce-dure across the three scenarios. The cumulative per-centage of workload ratings at or above 7 is lower when applying the OneTeamCockpit procedure. Al-so, the new procedure provides no observer ratings at or above 7 across the three scenarios. Participant and observer ratings are more consistent when refer-ring to the OneTeamCockpit procedure. The Airbus procedure shows a predominance of participant rat-ings between 1 and 3 while the observers almost equally rated between the two workload impact clas-

sifications satisfactory and unsatisfactory but tolera-ble. In closing, the OneTeamCockpit procedure seems to be beneficial to workload across the three scenarios although a between-subjects’ agreement on lower workload scores was only achieved in the third scenario. Table 2. Percentage of Flight Crew Workload Ratings for both Procedures across the three Scenarios ______________________________________________ Workload Airbus OneTeamCockpit Impact Participant/Observer Participant/Observer Classification ______________________________________________ N 45ú 30 45ú 30 Satisfactory: 1 – 3 64.5ú 40.0 73.3ú 80.0 Unsatisfactory but tolerable: 4 – 6 22.2ú 43.3 20.0ú 20.0 Not tolerable but possible to complete the task 7 – 9 13.3ú 16.7 6.7ú 0.0 Task not completed 10 0.0ú 0.0 0.0ú 0.0 ______________________________________________ Cumulative 35.5ú 60.0 26.7 ú 20.0 percentage > Bedford 3 (unsatisfactory) ______________________________________________ Nevertheless, the training effect in this study was significant due to the within-subject design and small sample size (Wainwright, 1987; Gawron, 2008). Finally, the redesign still indicates 26.7 per-cent of participant ratings beyond the satisfactory level (WL≤3).

3.3.2 Flight Crew Performance Looking into the crew performance, we have ob-served three potential conflicts of the One-TeamCockpit procedure: 1) A flap retraction before thrust reduction, 2) omitted or delayed status reports by the PS and 3) a mixture between tasks of ATM1 and ATM2. During a GA on the Airbus A320, there is no retrac-tion from Flaps 3 to Flaps 1 before reducing the thrust from TakeOff/Go-Around (TOGA) to Climb Thrust (CL) (Airbus, 2017). The trigger to initiate the thrust reduction by the PF is indicated on the FMA by a white flashing LVR CLB indication (Air-bus, 2017). However, in the test, we recorded a flap retraction to Flaps 1 before the thrust reduction alti-tude in both the Airbus and OneTeamCockpit proce-dure. However, it appears that this event happens more often when applying the redesign. Observa-

tions during the test indicated that depending on the aircraft performance, the F-Speed is often exceeded immediately after GA initiation. In the One-TeamCockpit procedure this speed is the trigger to initiate the flap retraction to Flaps 1. In this event, the OneTeamCockpit procedure seems to be more vulnerable to an inadvertent flap retraction since there is no command by the PF besides being able to call for Hold Flaps. However, in all observed events, the PF did not interfere when the PS retracted the flaps early. This observation provides evidence that parallel workflows can result in a detriment of cross-checking and visual tunneling of the PF on the PFD due to the new function allocation. The OneTeamCockpit procedure fully allocates the execution of the configuration change management to the PS but the PF remains in the loop by continu-ous status reports (e.g. Flaps 3, Gear Up, Speed-Checked-Flaps 1, Speed-Checked-Flaps 0). There-fore, omitting status reports affects the awareness of the PF related to the aircraft system states due to the one-way communication and parallel workflows (Degani & Wiener, 1994). This condition can impair the stability of the working environment (Pritchett et al., 2014). To enable the correct mental model of the aircraft system states, the PF must increase the monitoring work which can result in a higher work-load (Pritchett et al., 2014). Finally, the stability of the working environment is further influenced by a frequently observed mix between ATM1 and ATM2 when applying the OneTeamCockpit procedure. Figure 9 shows the operator action logic frequently observed in the test. Figure 9. Designer Action Logic (ATM2 followed by ATM1) compared to the Operator Action Logic (Mix of ATM1 and ATM2) during a revised Missed Approach.

It is reasonable to combine both action blocks by the PS since the flap retraction reduces the drag and the early heading input allows the PF to change the lat-eral flight path. Heading inputs before configuration changes were also observed when applying the Air-bus procedure indicating that pilots might rely on rule-based heuristics for following procedures (tech-niques) independent of the procedure design (Landry et al., 2006). However, it is contrary to the designer action logic and develops a condition where the PF experiences a combination of type one (the PF can-not predict the upcoming action of the PS) and type two unpredictability (the PF cannot anticipate the timing of actions of the PS) (Pritchett et al., 2014). Nonetheless, besides the undesired mixture of ATM 1 and 2, we did not identify any lateral or vertical deviations from the flight path by all crews in the third scenario when applying the OneTeamCockpit procedure. However, when applying the Airbus pro-cedure, we identified frequent omissions of pulling the heading selector. Therefore, the heading mode was selected but not engaged so that rollouts oc-curred far beyond the desired heading change of 20 degrees. Vertical deviations from the flight path were not observed.

3.3.3 Aircraft Performance Finally, at a weight at GA initiation of 54 tones there was no concern about a potential detriment in air-craft performance or flaps overspeed due to the de-layed flap retraction of the OneTeamCockpit proce-dure. None of the crews experienced a flap overspeed condition. Also, none of the pilots raised a concern regarding the aircraft performance even when staying below maximum landing weight of 64.5 tones (Airbus, 2017). However, the impact on the performance was evident when referring to an engine-out GA. Degraded aircraft performance is part of the contingency analysis and beyond the scope of this study. Nevertheless, two initial tests in the simulator applying the OneTeamCockpit proce-dure after an engine failure were conducted. The re-sults indicated a significant reduction in airspeed due to the increased aerodynamic drag when delaying the flap retraction at the recommended pitch of 12.5 degrees for an engine-out GA (Airbus, 2017). The minimum climb gradient during GA of 2.5 percent was exceeded. However, the airspeed was close to the lowest selectable speed (VLS) for autopilot and autothrust. Moreover, by delaying the flap retraction the aircraft becomes less maneuverable because the recommended flaps setting for maneuvering with engine-out is Flaps 3 (BEA, 2000). The decreased manoeuvrability may result in exaggerated control inputs or over-controlling which contributed to the Gulf Air Flight GF-072 accident (BEA, 2000). Nev-ertheless, this effect was not apparent during the test besides flying with Flaps Full for a longer time.

3.3.4 Pilot Feedback Participants valued the calm atmosphere on the flight deck due to the elimination of commands by the PF related to the configuration change manage-ment. The new function allocation defined clear re-sponsibilities while providing an enhanced active involvement of the PS in the workflow. The ATM improved the flexibility while the PF can primarily focus on flying the aircraft. The delayed flap retrac-tion was considered as beneficial to monitor the flight path by both crew members at GA initiation. Failures in crew coordination are less critical due parallel workflows. The loss of the SMM-loops through parallel workflows was considered as the critical aspect of the OneTeamCockpit procedure due to the higher probability of forgetting to perform tasks and a potential detriment of crosscheck-monitoring.

4 DISCUSSION

This study has demonstrated that the One-TeamCockpit Concept can be successfully imple-mented in the design of flight deck procedures. The redesign does not require any modifications by the flight crew in the three defined scenarios. Instead, the procedure introduces the ATM to enhance the flexibility of the design. The analysis indicated that the ATM could result in a lower workload of the flight crew. An overlay between status reports and commands is less likely compared to the Airbus pro-cedure which results in a calm atmosphere. This condition is critical considering that the operational environment is described by perturbations that likely interrupt the task sequence (Loukupoulos et al., 2009). The test did not provide evidence for a poten-tial flaps overspeed due to the delayed flap retrac-tions. Aircraft performance at the assigned weight was deemed satisfactory. However, the test data in-dicated a detriment in climb performance in the event of an engine-out GA. This aspect is safety-critical and requires further performance calcula-tions. The most critical aspect of the redesign is the loss of the SMM-loops based on parallel workflows which can result in an inadvertent flap retraction be-fore reaching the thrust reduction altitude. Moreo-ver, it appears that flight crews tend to omit cross-checking which may result in a higher probability of forgetting to perform tasks. This aspect is safety-critical and requires further verification. A checklist displayed on the Electronic Centralized Aircraft Monitoring (ECAM) display could be a potential de-sign solution to close the SMM-loop and to confirm that all configuration changes are performed. How-ever, the information density on the ECAM is al-ready high, and not all participants agreed that a checklist is necessary.

Overall, the Airbus A320 GA procedure is practica-ble and operational feasible. However, the One-TeamCockpit provides a promising concept that en-hances the flexibility of the flight crew during the GA.

5 CONCLUSIONS

This study has demonstrated that the One-TeamCockpit Concept can be integrated into the cur-rent design of flight deck procedures to enhance their resilience in both normal and non-normal GAs. The next step is to investigate the effect of the loss SMM-loops due to the parallel workflows in more detail. Finally, the ATM as part of the One-TeamCockpit Concept has been shown to be a prom-ising concept to enhance the flexibility in complex and dynamic flight phases.

6 LIMITATIONS

The results of the test are limited due to the small sample size and simulator limitations that have in-fluenced the workload ratings as a confounding vari-able. Moreover, performance estimations are based solely on pilot feedback and observation and require a numerical analysis for future certification.

7 ACKNOWLEDGEMENT

The authors would like to thank Per Ohme, Dominik Niedermeier and Jan Phillip Buch of the German Aerospace Center (DLR) for their superlative tech-nical support in conducing the simulator campaign. Also, we appreciated the support and feedback by all airline and test pilots who participated in this study.

8 REFERENCES

Active Pilot Monitoring Working Group 2014. A Pracitcal Guide for Improving Flight Path Monitoring. Flight Safety Foundation. Available at: https://flightsafety.org/files/ flightpath/EPMG.pdf (Accessed: 29 April 2018).

Air Accident Investigation Branch, AAIB 2016a. Airbus A 319-111 G-EZTZ, EW/G2016/20/11. AAIB Bulletin 8: 27-29

Air Accident Investigation Branch, AAIB 2016b. Airbus A 320-214 G-EZTZ, EW/G2016/20/11. AAIB Bulletin 8: 30-31.

Airbus 2005. Flight Operations Briefing Notes - Landing Techniques - Bounce Recovery - Rejected Landing. Toulouse, France: Airbus Industries.

Airbus 2006. Flight Operations Briefing Notes - Standard Operating Procedures - Operating Philosophy. Toulouse, France: Airbus Industries.

Airbus 2017. A320 Flight Crew Training Manual. Toulouse: Airbus Industries.

Anderson, F., McDaniel, M. A. and Einstein, G. O. 2017. Remembering to Remember: An Examination of the Cognitive Processes Underlying Prospective Memory. In

Stein, O. (ed.) Reference Module in Neuroscience and Biobehavioural Psychology. Elsevier: 2–11.

ATG, Aeronautical Training Group 1999. Training Notes Airbus A320 - Book 2 - Airbus A320 Avionics Systems - Navigation and Autoflight. Luton, Bedfordshire, UK.

ATSB, Australian Transport Safety Bureau 2007. Go-around event Melbourne Airport, Victoria, 21 July 2007, VH-VQT, Airbus Industries A320-232, ATSB Transport Safety Report AO-2007-044 Final. Canberra City, Australia.

ATSB, Australian Transport Safety Bureau 2014. Flap overspeed and altitude exceedance during go-around Airbus A321, VH-VWY ,Carins Airport, Queensland, 3rd September 2012, ATSB Transport Safety Report AO-2012-116. Northbourne, Australia.

Barshi, I. & Healy, A. 1993. Checklist procedures and the cost of automaticity. Memory & Cognition 21(4): 496–505.

Barshi, I., Mauro, R., Degani, A. and Loukopoulou, L. 2016. Designing Flightdeck Procedures , NASA/TM-2016-219421. Moffett Field, California: NASA Ames Research Center.

BEA, Burea d'Enquêtes et d'Analyses 2000. Accident Investigation Report, Gulf Air Flight GF-072, 23.08.2000. Le Bourget, France.

BEA, Burea d'Enquêtes et d'Analyses 2007. Final report on the investigation into the accident involving the Armavia A320 near Sochi Airport on 3 May 2006. Le Bourget, France.

BEA, Burea d'Enquêtes et d'Analyses 2009. Rapport Incident grave survenu le 23 septembre 2009 sur l’aérodrome Paris Charles de Gaulle (95) à l’Airbus A 319-111 immatriculé F-GRHU. Le Bourget, France.

BEA, Burea d'Enquêtes et d'Analyses 2013. Study on Aeroplane State Awareness during Go-Around. Le Bourget, France.

de Brito, G. 1998. Study of the use of airbus flight deck procedures and perspectives for operational documentation. International Conference on Human Computer Interaction in Aeronautics (HCI-Aero 98), Ecole Polytechnique de Montréal, Canada: 195–201.

Cahill J., Callari, T.C., Fortmann, F., Suck, S., Javaux, D., Hasselberg, A. and Stoeve S. H. 2017. Adaptive Autmation and the Third Pilot: Managing Teamwork and Workload in an Airline Cockpit. In: Longo, L. and Chiara Leva, M. (eds). Human Mental Workload, Models and Applications, First International Symposium, H-Workload 2017, Dublin Ireland, June 28-30, 2017: 161 - 173.

Cahill, J., McDonald, N. and Losa, G. 2014. A Sociotechnical Model of the Flight Crew Task, Human Factors 56(8): 1337–1363. Available at: https://doi.org/10.1177/ 0018720814532684.

Corwin, W.H., Sandry-Garza, D.L., Biferno, M.H., Boucek, G.P., Logan, A.L., Jonsson J. E. and Metalis, S.A. 1989. Assessment of crew workload measurement methods, techniques and procedures, Vol. 1 - Process, emthods and results (WRDC-TR-89-7006). Wright Patterson Air Foce Base, OH.

Degani, A., Heymann, M. and Shafto, M. 1999. Formal Aspects of Procedures: The Problem of Sequential Correctness. Proceedings of the Human Factors Ergnomics Society 43rd Annual Meeting, TX, 43(20): 1113–1117. Available at: https://doi.org/10.1177/154193129904302012.

Degani, A. & Wiener, E. L. 1994. On the Design of Flight- Deck Procedures. Moffett Field, California: NASA Ames Research Center.

Degani, A. & Wiener, E. L. 1997. Procedures in complex systems: The airline cockpit. IEEE Transaction on Systems,

Man and Cybernetics 27(3): 302–312. Available at: doi:10.1109/3468.568739.

Degani, A. & Wiener, E. L. 1998. Design and operational aspects of flight-deck procedures. Proceedings of the International Air Transport Association (IATA) Annual Meeting, Montreal, Canada, 1998.

Dehais, F., Behrend J., Peysakhovich, V., Causse, M. and Wickens, C.D. 2017. Pilot Flying and Pilot Monitoring's Aircraft State Awareness During Go-Around Execution in Aviation: A Behavioral and Eye Tracking Study. The International Journal of Aerospace Psychology 27(2): 15-28. Available at: http://dx.doi.org/10.1080/10508414. 2017.1366269

Dismukes, R. K. 2012. Prospective Memory in Workplace and Everyday Situations. Current Directions in Psychological Science 21(4): 215–220. Available at: https://doi.org/10.1177/0963721412447621.

Dodhia, R. M. & Dismukes, R. K. 2003. A task interrupted becomes a prospective task: Encoding and retrieval manipulations. Poster presented at the 44th Annual Meeting of the Psychonomic Society, Vancouver, British Columbia.

Dodhia, R. M. & Dismukes, R. K. 2009. Interruptions Create Prospective Memory Tasks. Applied Cognitive Psychology 23: 73–89. Available at: https://doi.org/10.1002/acp.1441.

Dornreich M.C., Passinger B., Hamblin C., Keinrath C., Vasek J., Whitlow S.D. and Beekhuyzen M. 2017. Evaluation of the Display of Cognitive State Feedback to Drive Adaptive Task Sharing. Fronties in Neuroscience 11(144):1-12. Available at: https://doi.org/10.3389/fnins.2017.00144.

Einstein, G. O., Smith, R. E., McDaniel, M. A. and Shaw, P. 1997. Aging and prospective memory: The influence of increased task demands at encoding and retrieval. Psychology and Aging 12(3): 479–488. Available at: http://psycnet.apa.org/doi/10.1037/0882-7974.12.3.479.

FAA, Federal Aviation Authority 2003. Standard Operating Procedures for Flight Deck Crewmembers. Advisory Circular No. 120-61A. Washington, DC: US Department of Transportation.

Fernandez, B., Lesceu, X. and Marconnet, D. 2017. Introduction to the Soft Go-Around Function. Safety First - The Airbus Safety Magazine (23): 14–21.

Flight Safety Foundation 2009. Manual Go-Around: FSF ALAR Briefing Note 6.2. Alexandria, VA.

Funk, K., Mauro, R. and Barshi, I. 2009. NextGen flight deck human factors issues. Proceedings of the 15th International Symposium on Aviation Psychology, Dayton, OH: 208–213.

Gawron, V. J. 2008. Human Performance, Workload, and Situational Awareness Measures, Boca Raton, FL: CRC Press.

Go-around (GA) Safety Forum 2013. Go-around Safety Forum 18 June 2013 Brussels: Findings and Conclusion. Brussels: Flight Safety Foundation, European Regions Airline Association, Eurocontrol.

Granger, S. & Jeanpierre, E. 2011. Automatic NAV engagement at Go Around. Safety First - The Airbus Safety Magazine (11): 19–21.

Harris, R.M., Hill, S.G., Lysaght, R.J., and Christ, R.E. 1992. Handbook of operating the OWL & NEST technology (ARI Research Note 92-49). Alexandria, VA: United States Army Research Institute for the Behavioral Social Science.

Hradecky, S. 2017. Incident: Easyjet A320 at Paris on Mar 2017, inadvertent flaps retraction on takeoff, The Aviation Herald. Available at: http://avherald.com/ h?article=4a741c67 (Accessed: 29 April 2018).

Kemény, C. J. & Popp, C. 2016. One Team, Journal for Civil Aviation Training (CAT) 27(2): 26–28.

Kingdom of Bahrain Civil Aviation Affairs (CAA) 2000. ACCIDENT INVESTIGATION REPORT Gulf Air Flight GF-072 Airbus A320-212, REG. A40-EK on 23 August 2000 at Bahrain.

Kourdali, H. K. & Sherry, L. 2016a. A Comparison of Two Takeoff Flap Retraction Standard Operating Procedures. Integrated Communications Navigation and Surveillance (ICNS), 2016, Herndon, VA, USA. Available at: doi:10.1109/ICNSURV.2016.7486344.

Kourdali, H. K. & Sherry, L. 2016b. A Systems Engineering Method and Simulation of Standard Operating Procedures. HCI-Aero ’16 Proceedings of the International Conference on Human-Computer Interaction in Aerospace, Paris, France — September 14 - 16, 2016. Available at: https://doi.org/10.1145/2950112.2964580.

Landry, S. J., Jacko, J. A. and Coulter, W. H. 2006. Impact of the use of techniques and situation awareness on pilots procedure compliance. Proceedings of the Human Factors and Ergonomics Society Annual Meeting 50(1): 40–44. Available at: https://doi.org/10.1177/154193120605000109.

Liu, H., Ma, X., Zhang Y. and Jin, Z. 2017: Practice for the Certification of Minimum Flight Crew Workload. HCI International 2017 - Posters' Extended Abstracts, 19th International Conference, Vancouver, BC, Canada, July 9-14: 339-346.

Loukupoulos, L. D., Dismukes, R. K. and Barshi, I. 2003. Concurrent task demands in the cockpit, challenges and vulnerabilities in routine flight operations. Proceedings of the 12th International Symposium on Aviation Psychology: 737–742.

Loukupoulos, L. D., Dismukes, R. K. and Barshi, I. 2009. The Multitasking Myth - Handling Complexity in Real-World Operations. Aldershot, UK: Ashgate.

Ludlow, S. 2016. Reducing the Threat of the Somatogravic Illusion. Flight Safety Foundation International Air Safety Summit (IASS), Dubai, November 2016.

Machanda, M. & Sikora, I. 2016. Mismanaged Go-around and the need to mitigate their occurrence. International Conference on Traffic and Transport Engineering - Belgrade, November 24-25, 2016: 44-48.

Maille, N. 2016. Modeling Airline Crew Activity to Improve Flight Safety Analysis. 38th IEEE Aerospace Conference, Big Sky, MT United States.

Marconnet, D. & Roland, M. 2014. Flying a Go-Around Managing Energy. Safety First - The Airbus Safety Magazine (17): 19–22.

Mauro, R., Degani, A., Loukopoulos, L. and Barshi, I. 2012. The Operational Context of Procedures and Checklists in Commercial Aviation. Proceedings of the Human Factors and Ergonomics Society 56th Annual Meeting - 2012. Available at: doi:10.1177/1071181312561158.

McGann, D., Ellis, J. A. and Milne, A. 2002. Conceptual and perceptual processes in prospective remembering: Differential influence of attentional resources. Memory and Cognition 30(7): 1021–1032. Available at: https://doi.org/10.3758/ BF03194320.

NASA, National Aeronautics and Space Administration 2014. Human Integration Design Processes (HIDP) - NASA/TP-2014-218556. Houston, Texas: NASA.

Niedermeier D. & Buch, J.P. 2016. Das Man4Gen Projekt: Unterstützung von Airline- Crews in unerwarteten Situationen. Deutscher Luft- und Raumfahrtkongress 2016:

13-15 Norden, C. 2010. Go-around handling. Safety First - The

Airbus Safety Magazine (10): 12–13. Owens, D. 2013. The Golden Rules for Pilots Moving from

PNF to PM. Safety First - The Airbus Safety Magazine (15): 5–7.

Pélegrin, C. 2007. Compliance to operational procedures: Why do well trained and experienced pilots not always follow procedures?. Safety First - The Airbus Safety Magazine (5): 20–23.

Pritchett, A. R., Kim, S. Y. and Feigh, K. M. 2014. Measuring Human-Automation Function Allocation. Journal of Cognitive Engineering and Decision Making 8(1): 52–77. Available at: https://doi.org/10.1177/1555343413490166.

Reynal, M., Colineaux, Y., Vernay, A. and Dehais, F. 2016. Pilot flying vs. Pilot monitoring during approach phase: An eye-tracking study. Paper presented at HCI-Aero 2016, International Confoerence on Human Computer Interaction in Aerospace, Paris.

Richter, H.-L. 2014. Outbound flight to the future. DLR Magazine (140–141): 37–39.

Roscoe, A.H. 1984. Assessing Pilot Workload in Flight. Flight Test Techniques AGARD No 373, Proceedings of the Flight Mechanics Panel Symposium Held in Lisbon, Portugal on 3-5 April 1984. Advisory Group for Aerospace Research and Development, Neuilly-sur-Seine: 12-1 - 12-13.

Roscoe, A. H. & Ellis, G. A. 1990. A Subjective Rating Scale for Assessing Pilot Workload in Flight: A Decade of Practical Use, Technical Report TR 90019. Farnborough, UK: Ministry of Defence.

de Terssac, G. & Chabaud, C. 1990. Référentiel opératif commun et fiabilité’. In Leplat, J. and de Terssac, G. (eds) Les facteurs humains de la fiabilité dans les systèmes complexes. Tolouse, France: Octarès: 111–139.

Toglia, J. & McDaniel, M. A. 2016. Prospective Memory. In Strobach, T. and Karbach, J. (eds) Cognitive training: An overview of features and applications. Heidelberg, New York: Springer: 81–91.

Wainright, W.A. 1987. Flight test evaluation of crew workload, In Roscoe A.H. (ed.) The practical assessment of pilot workload, AGARDograph No. 282. Neuilly sur Seine, France: 60-68 437–443.

AIRBUS GO-AROUND PROCEDURE PF PM

THRUST LEVERS………………………………………TOGA ROTATION………………………………………..PERFORM GO AROUND-FLAPS……………….……..ANNOUNCE FMA………………………………………………ANNOUNCE LANDING GEAR UP……………………………….ORDER NAV OR HDG MODE ………………….AS REQUIRED AP………………………………………………AS REQUIRED

FLAPS LEVER………..REDUCE BY ONE STEP POSTIVE CLIMB……………………ANNOUNCE LANDING GEAR LEVER………..…SELECT UP NAV OR HDG MODE…..SET AS REQUIRED

AT GO-AROUND THRUST REDUCTION ALTITUDE THRUST LEVERS…………..……………………………….CL

MONITOR TARGET SPEED INCREASE TO GREEN DOT AT F-SPEED FLAPS 1………………………………………………..ORDER AT-SPEED FLAPS 0………………………………………………..ORDER

FLAPS LEVER………………………………FLAPS 1 FLAPS LEVER………………………………FLAPS 0 GROUND SPOILERS……………………DISARM NOSE LIGHT SWITCH………………………..OFF RUNWAR TURNSWITCH…………………..OFF OTHER EXTERIOR LIGHTS….AS REQUIRED

ONETEAMCOCKPIT GO-AROUND PROCEDURE

PF PM THRUST LEVERS………………………………………TOGA ROTATION………………………………………..PERFORM GO AROUND………………………….……..ANNOUNCE FMA………………………………………………ANNOUNCE NAV OR HDG MODE ………………….AS REQUIRED AP………………………………………………AS REQUIRED

AIRCRAFT ATTITUDE………………MONITOR POSTIVE CLIMB……………………ANNOUNCE ATM1 FLAPS LEVER………..REDUCE BY ONE STEP LANDING GEAR LEVER………..…SELECT UP ATM 2 NAV OR HDG MODE…..SET AS REQUIRED

AT GO-AROUND THRUST REDUCTION ALTITUDE THRUST LEVERS…………..……………………………….CL

MONITOR TARGET SPEED INCREASE TO GREEN DOT

AT F-SPEED FLAPS LEVER………………………………FLAPS 1 AT S-SPEED FLAPS LEVER………………………………FLAPS 0 GROUND SPOILERS……………………DISARM NOSE LIGHT SWITCH………………………..OFF RUNWAR TURNSWITCH…………………..OFF OTHER EXTERIOR LIGHTS….AS REQUIRED