Education

International Journal of Aviation, Aeronautics, and

Aerospace

Volume 5 | Issue 1 Article 6

2-19-2018

Flight Simulator Fidelity, Training Transfer, and the Role of Instructors in Optimizing Learning Paul L. Myers III Embry-Riddle Aeronautical University, myersIIp@my.erau.edu Arnold W. Starr Embry-Riddle Aeronautical University, starrjra@my.erau.edu Kadie Mullins Embry-Riddle Aeronautical University, haywardk@erau.edu

Follow this and additional works at: https://commons.erau.edu/ijaaa

Part of the Cognition and Perception Commons, Curriculum and Instruction Commons, and the Educational Psychology Commons

This Literature Review is brought to you for free and open access by the Journals at Scholarly Commons. It has been accepted for inclusion in International Journal of Aviation, Aeronautics, and Aerospace by an authorized administrator of Scholarly Commons. For more information, please contact commons@erau.edu, wolfe309@erau.edu.

Scholarly Commons Citation Myers, P. L., Starr, A. W., & Mullins, K. (2018). Flight Simulator Fidelity, Training Transfer, and the Role of Instructors in Optimizing Learning. International Journal of Aviation, Aeronautics, and Aerospace, 5(1). https://doi.org/10.15394/ijaaa.2018.1203

On November 12, 2001, American Airlines Flight 587, an Airbus A300,

departed from John F. Kennedy International Airport. Shortly after takeoff, the

aircraft encountered wake turbulence from a preceding departing aircraft. The

aircraft upset caused the copilot flying the aircraft to use excessive rudder input in

both directions, over-stressing the rudder and causing it to depart the aircraft

(NTSB, 2001; NTSB, 2004a; NTSB, 2004b). All 260 people on the aircraft and five

people on the ground were killed and the aircraft was destroyed.

In the mishap report, the National Transportation Safety Board (NTSB)

highlighted two contributing factors. Both relate to the simulator training and its

fidelity. First, incorrect rudder application was taught by simulator instructors.

Second, the rudder pedal responses in the simulator were significantly different

from the aircraft. The combination of the two may have caused the copilot to over-

control, leading to confusion and surprise (NTSB, 2001; NTSB, 2004a; NTSB,

2004b). The rudder input fidelity differences were caused by software

misrepresentation of an elastic cable stretch that was less stiff than the cable stretch

in the aircraft (NTSB, 2001; NTSB, 2004a; NTSB, 2004b). The differences

between the simulator and aircraft are shown in Figure 1.

Figure 1. Fidelity differences between simulator and aircraft rudder pedal inputs

(Courtesy of the NTSB, 2001).

Fidelity, the degree to which the simulator looks like the real aircraft and

the similarity to which it acts like the real aircraft, is closely linked to training

transfer. (Allen, Hays, & Buffardi, 2001; Noble, 2002). Training transfer refers to

the process by which knowledge, abilities, or skills acquired through training are

applied to the actual situation (Hochmitz, & Yuviler-Gavish, 2011). Negative

training transfer is the dampening effect of previous learning on the exercise of

skills or on new learning (Blaiwes, Pug, & Regan, 2001). The presence of poor

1

Myers et al.: Simulator Fidelity, Training Transfer, and the Role of Instructors

Published by Scholarly Commons, 2018

fidelity may lead to negative training transfer with unsafe, even catastrophic results

(Lee, 2009). Numerous mishaps, such as American Airlines Flight 587, can be

illustrative of this theory. This paper, therefore, examines the efficacy of both high

and low fidelity on training transfer and explores the flight simulator instructor’s

role in exploiting the simulator’s strengths as a training tool while minimizing

negative training transfer.

Simulator Training History

The history of flight simulation dates to 1929 when Edwin Link built his

first Link Trainer. The device had a basic set of instruments, a primitive motion

platform, and no visual display (Lee, 2009). When World War II began, the Link

Trainer was integrated into flight training and used extensively. At the time,

training accident rates were quite high and using simulators to reduce the aircraft

accident rate was believed to be a logical outcome (Valverde, 1973). The training

value of simulators substituting for aircraft was intuitive and based on common

sense (Lee, 2009). After the war, rapid simulator progress was achieved due to

many technological advancements during the war. Crucial to this evolution was the

development of analog computers. However, the academic study of flight

simulators did not start until around 1949 (Valverde, 1973). These studies continue

in earnest today.

Figure 2. Link Blue Box Trainer.

(https://www.link.com/media/gallery/Link_Blue_Box_Training_2.jpg. Image

Courtesy of L3 Link. Reprinted with permission.)

2

International Journal of Aviation, Aeronautics, and Aerospace, Vol. 5 [2018], Iss. 1, Art. 6

https://commons.erau.edu/ijaaa/vol5/iss1/6 DOI: https://doi.org/10.15394/ijaaa.2018.1203

Advantages

Simulators provide several advantages and are integral to modern aviation

training programs. These advantages include (a) providing a safe environment to

practice potentially dangerous procedures, such as an engine failure or rejected

takeoff, that should not or cannot be performed in the aircraft, (b) significantly

reducing training costs, (c) producing a positive impact on the environment by way

of conservation of resources and reduced carbon footprint, (d) providing a research

platform and laboratory, and (e) allowing rapid and multiple repetitions of events,

such as instrument approaches and landings (Williges, Roscoe, & Williges, 2001).

These many advantages have resulted in the requirement to use advanced

simulators in the FAA’s Advanced Qualification Program (Longridge, 1997).

Figure 3. Full Motion Level 3 / 4 Flight Simulators. BART International –

Simulators at SimCom Training Center. (Retrieved from

http://www.bartintl.com/content/simcom-pks-sim-bay3909rsjpg Reprinted with

permission.)

Disadvantages

Extensive simulator use for training does have some drawbacks, however.

Some disadvantages of simulators include (a) simulator sickness, (b) inducing

adaptation and compensatory skills, (c) poor motion cueing, (d) lack of user

motivation, (e) a complex system architecture, (f) over-regulation, and (g) high

costs associated with the most advanced simulators (Lee, 2009). Simulator sickness

3

Myers et al.: Simulator Fidelity, Training Transfer, and the Role of Instructors

Published by Scholarly Commons, 2018

is a form of motion sickness and can occur in both fixed and motion-based

simulators. Symptoms exhibited by pilots include sweating, fatigue, dizziness, and

vomiting (Lee, 2009). Pilots experiencing simulator sickness can be affected to the

point of needing to stop the training (Stein & Robinski, 2012). Motion cueing, the

algorithm used by simulators to align visual input with human motion sensing, is

employed on the more advanced machines and is expensive to install and maintain

(Williges et al., 2001). This process aims to replicate the feeling of being in a real

aircraft. However, poor motion cueing can cause diminished fidelity and increased

sickness.

Additionally, subjects know they are not in the aircraft potentially

impacting pilot motivation. Their perception of the danger and stress level may be

significantly reduced resulting in decisions that would not be made in the aircraft

(Lee, 2009). For example, a pilot may elect to continue an unstable approach and

attempt to land the aircraft in an unsafe situation. Whereas, if airborne, the decision

may have been to go-around. Another disadvantage is the multitude of technical

requirements of the simulator, creating a highly complex system architecture.

Making changes to the simulator system architecture is often impeded by the

plethora of government regulations in the certification process (Lee, 2009). Finally,

cost, especially for smaller carriers, is a significant determinant of the level of

fidelity that is incorporated into an operator’s training program and thus can be an

impediment to widespread use of simulators (Lee, 2009).

Figure 4. G1000 Part Task-Trainer. (Retrieved from http://www.flight1tech.com/

products/avionicssimulations/garming1000studentsimulatorsoftware.aspx. Flight 1

Tech Systems. Reprinted with permission.)

4

International Journal of Aviation, Aeronautics, and Aerospace, Vol. 5 [2018], Iss. 1, Art. 6

https://commons.erau.edu/ijaaa/vol5/iss1/6 DOI: https://doi.org/10.15394/ijaaa.2018.1203

Simulator Fidelity – Key Definitions

Fidelity is a fundamental concept in simulator design and is comprised of

three elements: physical fidelity, cognitive fidelity, and functional fidelity. These

elements are defined as follows.

Physical fidelity is the level to which the simulator replicates the

physical aircraft flight deck and feel (Allen et al, 1986). Physical

fidelity includes motion, visual, and sound replication. There are

limits to physical fidelity: Schroeder and Chung (2001) and Vaden

and Hall (2005) assert, for example, that current motion technology

cannot replicate the actual motion cues a pilot would receive in

coordinated flight to a 100% level.

Cognitive fidelity refers to the ability of the simulator training

environment to replicate the cognitive skills required on the flight

deck (Lee, 2009). Specifically, factors that comprise cognitive

fidelity include psychological and perceptive factors such as

situational awareness, anxiety, stress, and decision making (Taber,

2014).

Functional fidelity is defined as to what degree the simulator acts

like the real equipment (Allen et al., 1986).

Another important definition related to fidelity is task, which is a goal or

problem to be solved (Lintern, 2001). It is important to distinguish between a

simulator and trainer as well. A simulator is a device that represents a specific

counterpart aircraft whereas a trainer represents a particular class of vehicles

(Williges et al., 2001).

Havighurst, Fields, and Fields (2010) define high fidelity as the required

equipment and materials necessary to adequately simulate the task the learner is

expected to perform. They define low fidelity as equipment and materials that are

less similar to what task the learner is expected to perform.

Self-efficacy is the belief in one’s capabilities to correctly perform a task.

Generally, high-perceived self-efficacy can lead to positive performance outcomes

(Holbrook & Cennamo, 2014).

5

Myers et al.: Simulator Fidelity, Training Transfer, and the Role of Instructors

Published by Scholarly Commons, 2018

High-fidelity Simulators

High-fidelity simulators have been shown to increase self-efficacy

(Holbrook & Cennamo, 2014). Taber (2014) reinforces this point stating that

reducing simulator capability to the minimum level impedes confidence in future

skill performance. This, in turn, impedes the development of necessary coping

strategies. Many simulator designers, operators, technicians, and behavioral

scientists believe that the simulator should be designed with the maximum fidelity

possible since that is postulated to provide the most training transfer. However,

doing so results in higher costs that may not be feasible for some organizations

(Lintern, Sheppard, Parker, Yates, & Nolan, 1989).

The final component of understanding the nature of fidelity includes Hays

and Singer’s (1989) four levels of fidelity. Level 1 is considered high fidelity and

includes two aspects: precise reproduction of the operational counterpart and

deliberate reduction in fidelity to reduce costs without compromising training

effectiveness. The other three levels have incremental reductions in the level of

fidelity (Hays & Singer, 1989). For this paper, fidelity refers to the extent to which

the training situation must be similar to the actual aircraft situation in order to

provide effective training.

The development of high fidelity simulators requires the ongoing

engagement of stakeholders throughout the process. Naweed, Ward, Gourlay, and

Dawson (2017) suggest cross-disciplinary teams with a transdisciplinary approach

are most beneficial. This allows for sharing of knowledge and innovation from one

field to another. Such practices, they argue, supports this collaboration where teams

are better able to anticipate challenges, formulate resolutions, and establish more

innovative responses to promote fidelity of simulators (Naweed et al., 2017).

Simulator Fidelity and Transfer of Training

There is considerable debate regarding the effect of simulator fidelity on

training transfer, particularly regarding the impact of motion on training transfer.

When performance in the aircraft is better than if there was no simulator training

provided, this is called positive training transfer. Conversely, negative training

transfer refers to those situations when performance in the aircraft is poorer than if

there was no pre-training at all (Listern, 2001). Several studies concluded that low

fidelity resulted in negative training transfer, while other studies concluded that the

degree of simulator fidelity had little or no effect on training transfer, making the

subject contentious among training experts (Listern, 2001).

6

International Journal of Aviation, Aeronautics, and Aerospace, Vol. 5 [2018], Iss. 1, Art. 6

https://commons.erau.edu/ijaaa/vol5/iss1/6 DOI: https://doi.org/10.15394/ijaaa.2018.1203

When considering training transfer, it can be further divided into three sub-

areas: self-transfer, near transfer, and far transfer. Self-transfer is the decrement or

improvement in training transfer resulting from the repeated practice of the same

event. Near transfer is the decrement or improvement in training transfer resulting

from practicing different but similar events. Far transfer is the decrement or

improvement in training transfer resulting from the repeated practice of dissimilar

events (Noble, 2002).

The Debate

Simulator Fidelity Does Not Affect Training Transfer

Studies regarding simulator fidelity are inconclusive and, at times,

seemingly contradictory, with many asserting fidelity does not affect training

transfer while others affirm impacts. Burki-Cohen, Go, and Longridge (2001)

researched engine failure scenarios that resulted in either a rejected take-off or

continued take-off. They concluded from their study that motion for these tasks did

not affect evaluation and training simulator progress or transfer of training. The

authors, while exploring fidelity background information, did note that motion

improved pilot performance and control behavior when performing disturbance and

tracking tasks for low stability aircraft in the simulator (Burki-Cohen et al, 2001).

Norman, Dore, and Grierson (2012) in a study of medical students

performing clinical tasks in high-fidelity and low-fidelity simulators concluded that

there was no significant advantage (average differences 11% to 2%) of the high-

fidelity simulator use over the low-fidelity simulator. In a similar study, Fraser,

Peets, Walker, Tworek, Paget, Wright, and McLaughlin (2009) conducted a

medical students’ training study using a cardiorespiratory simulator (CRS). While

the simulator improved results when compared to students who did not use a

simulator, students displayed only a limited ability to transfer skills learned to other

real-world problems (Fraser et al, 2009).

Lintern et al. (1989) note in their study of ground attack bombing that

decreasing physical fidelity does not always lead to a decrease in training transfer.

Specifically, there was no difference in performance among crews using three

different visual simulation fields of view (Litern et al, 1989; Williges et al., 2001).

The researchers specifically mention, however, that visual simulation is needed for

ground-referenced maneuvers where high danger is present, such as high-speed/low

altitude military operations. They advocate that as learning changes, more

experience is gained (Lintern et al., 1989; Williges et al., 2001). Thus, fidelity

7

Myers et al.: Simulator Fidelity, Training Transfer, and the Role of Instructors

Published by Scholarly Commons, 2018

requirements change as learning progresses. Additionally, fidelity requirements

will also change based on learner ability and skill.

In a study of engine failure scenarios on takeoff, it was noted that no

significant training transfer differences between pilots who used motion and those

who did not. In the same study, however, the authors note that a lack of fidelity in

Air Traffic Control (ATC) communications offers an incomplete cognitive

environment and creates a false sense of simplicity to the pilots (Bürki-Cohen,

Boothe, Soja, DiSario, Go, & Longridge, 2000). This viewpoint is echoed by Lee

(2009) who stated that sound provides pilots needed feedback on both aircraft

systems and ATC communications to more fully simulate the aircraft operating

environment.

Vaden and Hall (2005) concluded that simulator performance and the

follow-on transfer performance did not show a direct relation. Additionally,

empirical evidence supporting the use of motion to improve training transfer is

lacking. Conclusions from the researchers cited in Vaden and Hall’s (2005) study

include the following:

• A comparison of T-37 pilot training students who used

motion and those who did not use motion yielded no

practical or statistical differences

• In a study of T-2C aircraft landings, motion was found to

provide no statistical benefit

• In a study of F-16 maneuver training in a fixed versus motion

simulator, there were no significant statistical differences in

performance, although it was noted that motion tended to

improve performance in some areas and degraded it in others

Neither field of view nor scene detail influenced training transfer from the

simulator to the aircraft (Caretta & Dunlap, 1998; Lintern & Garrison, 1992;

Lintern et al., 1997). Dahlström (2008) in a study of pilot training students up to

the first solo offered the following conclusions:

• High fidelity simulation has not necessarily resulted in

improved opportunities for learning coordinative and

cognitive skills • Despite high pilot acceptance, convincing visual effects,

and apparent validity of high-fidelity simulators, there is no

certainty as to whether training quality is improved

• With the introduction of new technology, operator work

demands are changed and new ways of performance and

possibilities for new forms of accidents can surface

8

International Journal of Aviation, Aeronautics, and Aerospace, Vol. 5 [2018], Iss. 1, Art. 6

https://commons.erau.edu/ijaaa/vol5/iss1/6 DOI: https://doi.org/10.15394/ijaaa.2018.1203

• To equate fidelity with better training reflects a limited

view on training

Simulator Fidelity Does Affect Training Transfer

Though many findings suggest no impact on training transfer by fidelity,

much research suggests fidelity does have an effect. Holbrook and Cennamo (2014)

in their study of high-fidelity and self-efficacy with law enforcement officers found

that high fidelity increased self-efficacy, emotional arousal, and led to positive

training transfer from the lessons learned in the simulator scenarios. The study

subjects remarked that no previous experience had prepared them as well as this

simulator period. Additionally, they commented that there was no experience better

than this because the simulator scenarios were so realistic (Holbrook & Cennamo,

2014).

A study of platform-based simulator motion concluded that, for pilot

coordinated maneuvers, the motion platform must translate laterally when it rolls.

If not, the pilot feels an uncoordinated turn and the needle and ball indicate a slip

(Schroeder & Chung, 2001). The study determined that as the motion cues

degraded, both objective and subjective evaluation results worsened (Schroeder &

Chung, 2001).

Testing Boeing 747-400 Captains and First Officers in four maneuvers,

including both engine failure scenarios during takeoff and engine-out landing

maneuvers, Burki-Cohen et al. (2003) determined the advantage of motion fidelity

was small for the engine failures on takeoff. However, the results demonstrated the

early alerting function of motion (Burki-Cohen et al., 2003). In the other

maneuvers, only very slight differences were noted, such as the motion group had

slightly longer and softer landings than the non-motion group. No difference in

performance was noted during recurrent evaluations indicating there was no benefit

for recurrent training (Burki-Cohen et al., 2003).

Advocating that the closer the simulation is to real-world conditions, the

better the transfer of skills, it was found that a high level of cognitive and physical

fidelity was required for Helicopter Underwater Egress Training (programs)

(Baldwin & Ford, 1988; Hochmitz & Yuviler-Gavish, 2011; Taber, 2014). Doing

so provides individuals the ability to practice whole-task skill demonstration in

several critical areas and provides the best transfer of training.

Bürki-Cohen et al. (2000) in their study noted that in an FAA-sponsored

review of AC120-40B, Subject Matter Experts (SMEs) from industry, the FAA,

9

Myers et al.: Simulator Fidelity, Training Transfer, and the Role of Instructors

Published by Scholarly Commons, 2018

and academia generally perceived the absence of motion to likely have detrimental

effects on pilot control performance. This is especially true when performing

maneuvers where sudden motion onset cueing with limited visual references occur.

Motion is necessary for correct control inputs, especially for maneuvers that

involve high-G tolerance and spatial disorientation avoidance (Viden & Hall,

2005). Specifically, during training, trainees who receive no-motion training can

never achieve the same performance level as those with motion training. They

conclude:

• If pilot performance is dependent on motion in flight, then the simulator with motion will provide better transfer than

the simulator without motion

• A lack of motion caused trainees to be less successful in developing flight control strategies than those trainees who

had practiced the skill with motion

• Generally, student pilot and instructor feedback indicate that including motion provides greater simulator acceptance

and meets pilot performance expectations

• Pilots preferred motion to no motion when the task was to control an unstable aircraft

• In a study of helicopter coupled-hover departure procedures, motion was found to have a positive statistical

effect on pilot performance (Viden & Hall, 2005)

For high-altitude stall recovery and overbank recovery, motion improved

results in vertical motion simulators (Zaal, Schroeder, & Chung, 2015).

Additionally, simulation motion needs to be intense and abrupt enough to provide

the appropriate stimulus that the pilot can detect and input an appropriate control

response (Caro, 2001a). Pilot survey ratings of motion support this assertion (Zaal

et al., 2015). In approach and landing with sidestep and engine out on takeoff, little

difference was noted in training transfer (Zaal et al., 2015).

Those trainees who were trained with simulators that had high-

physical/high-functional fidelity or high-physical/medium-functional fidelity were

found to repeat assigned tasks less frequently (Allen, Hays, & Buffardi, 1986). In

contrast, the highest number of required repeated attempts occurred for those

trainees who were using medium-function/low-physical fidelity simulators. Thus,

the authors concluded that both functional and physical fidelity had a strong effect

on performance. Additionally, lower physical fidelity was associated with longer

solution times. Finally, they concluded that given the physical and functional

10

International Journal of Aviation, Aeronautics, and Aerospace, Vol. 5 [2018], Iss. 1, Art. 6

https://commons.erau.edu/ijaaa/vol5/iss1/6 DOI: https://doi.org/10.15394/ijaaa.2018.1203

fidelity relationships discovered, functional and physical fidelity should not be

dealt with in isolation (Allen et al., 1986).

Allen, Park, and Cook (2010), in their study on simulator driving scenarios,

found that those subjects training in a cab with a full-sized projected image had the

lowest crash rate and exhibited the least aggressive driving behavior. However, the

authors note several confounding variables were present which could have affected

the results.

Noble (2002) in his study, concluded that learner skill level must be

considered when determining fidelity. As the learner skill level improves, low-

quality fidelity devices become less effective when one considers the cost to build

them versus training efficiency. The learning stage of the student, the goals of the

training, and the level of fidelity are not mutually exclusive. Additionally, Noble

(2002) concluded from a study of KC-135 boom operators using a Boom-Operator

Part-Task Trainer (BOPTT), that both the environment and the task must be

considered when studying training transfer.

Hochmitz and Yuviler-Gavish (2011) conducted a simple study that divided

respondents into two groups consisting of a physical fidelity group and a cognitive

fidelity group. A three-dimensional virtual simulator was used. Performance

measures included training time, number of final errors, test time, number of

corrected errors, and time used to correct errors. The authors concluded that for

development of procedural skills in psychomotor tasks, a training approach using

both cognitive and physical training was required.

Lintern (2001) in his study concluded that skill transfer was based on some

type of similarity between the operational and training experience. His basic

premise is that the level of transfer is based on the extent to which the two

environments share common components.

Simulator Fidelity Training Transfer

After examining the various studies, the contradictory results of training

transfer were found to be caused primarily by lack of clearly defined study

methodologies, variances in study methodologies, and variances in the individual

tasks studied. Several authors note the study methodology problems (Caretta &

Dunlap, 1998; Vaden & Hall, 2005). Besides methodology problems, Caretta and

Dunlap (1998) also highlight lack of understanding of the mission and lack of true

simulator-to-aircraft transfer studies as contributors to conflicting results. Other

variables found to significantly influence results of studies include (a) criterion

11

Myers et al.: Simulator Fidelity, Training Transfer, and the Role of Instructors

Published by Scholarly Commons, 2018

measures where subjective measurements are used, (b) the subjects who have

different motor skills and cognitive capabilities, and (c) the instructor, who plays

an important role because of biases, attitudes, motivation, and abilities (Valverde,

1973).

It is understandable that results varied as often the specific tasks studied

also varied from study to study. For example, for upset recoveries, such as

American Flight 587, the NTSB, FAA, and other authors (Munshi, Lababidi, &

Alyousef, 2015; Vaden & Hall, 2005; Zaal et al., 2015) agree that high fidelity is

needed. For a task, such as being able to locate a switch in the cockpit, a part-task

trainer consisting of a cockpit diagram pasted on a piece of cardboard may suffice.

However, part-trainers are limited in their use (Caro, 2001b).

Ultimately, the amount of fidelity needed is specific to the training

objective, the individual task being trained, and the learning level of the student

(Caretta & Dunlap, 1998; Lee, 2009). Blaiwes et al. (2001) support this assertion

and further support the variance in study results. Different flight task types transfer

differently (Blaiwes et al., 2001). Further, Blaiwes et al. (2001) conclude (a)

particular motion types affect trainee training transfer and performance, (b) the

level of fidelity and type of trainer notably influences transfer, and (c) careful

specification of operational and trainer tasks is necessary to maximize training

transfer.

Finally, the FAA provides regulatory guidance through the Advanced

Qualification Program (AQP), Advanced Simulation Plan (ASP) and Advisory

Circular (AC)120-40b on simulator fidelity (Burki-Cohen, Go, & Longridge, 2001;

FAA, 2017). The AQP is designed to respond to changing training needs providing

pilots who not only have the requisite knowledge and hands-on skill but also are

proficient integrating cognitive and motor skills. The FAA is committed to

effectively preparing pilots to carry passengers. Thus, simulators must represent the

motor and cognitive challenges that would be experienced in an operational

environment. Additionally, the simulators must be sufficient to ensure full transfer

of performance and behaviors that would be experienced in the air (Burki-Cohen,

et al., 2001). To change regulatory guidance regarding fidelity, consistent study

results must present unequivocal evidence to prove that high-fidelity is not required

for the individual tasks in question (Bürki-Cohen, et al., 2000).

Safety and Fidelity

Improved safety outcomes are among the key factors in desiring appropriate

12

International Journal of Aviation, Aeronautics, and Aerospace, Vol. 5 [2018], Iss. 1, Art. 6

https://commons.erau.edu/ijaaa/vol5/iss1/6 DOI: https://doi.org/10.15394/ijaaa.2018.1203

fidelity within training environments. To better understand the relationship between

safety and simulator fidelity, a search was conducted of the NTSB website for

safety recommendations that included the word simulator for the aviation

transportation mode. The search returned 37 results that consisted of 29

recommendations for procedure changes and 8 fidelity improvements. These

recommendations are the result of commercial or business aircraft mishaps and are

summarized chronologically below:

Eastern Airlines Flight 66. On June 24, 1975, Flight 66, a Boeing 727,

crashed while executing a precision approach to John F. Kennedy Airport in

Jamaica, New York, killing 113 people. The aircraft was flying through or at the

base of a mature thunderstorm. The NTSB recognized that from this mishap and

other mishaps, thunderstorms were a problem. Therefore, the NTSB recommended

that wind shear models be developed for simulators to train pilots on the effects of

mature thunderstorms (NTSB, 1976).

Pan Am Flight 759. On July 9, 1982, Flight 759 crashed while taking off

from New Orleans International Airport in Kenner, Louisiana, killing 145 people

on the aircraft and 8 people on the ground. Windshear conditions had been detected

by the airport just prior to takeoff and there were heavy rain showers on the

departure path. Because of this mishap and other mishaps involving wind shear, the

NTSB recommended that realistic microburst wind models be incorporated into

flight simulator training programs (NTSB, 1982).

Rejected Takeoffs. After a series of mishaps related to high-speed rejected

takeoffs, the NTSB issued two safety recommendations: A-90-043 and A-90-044.

As part of the A-90-043 recommendation, the NTSB required, to the maximum

extent possible, that cues and cockpit warnings that resulted in high speed rejected

takeoffs for other than engine failures be incorporated. An example would be a tire

failure during takeoff. Additionally, A-90-044 required that all simulators of

passenger carrying operators accurately produce stopping distance available for a

rejected takeoff (NTSB, 1990).

China Eastern Airlines, flight 583. On April 6, 1993, Flight 583 was

flying from Beijing to Los Angeles when the slats inadvertently deployed during

cruise flight. The Captain flew the aircraft through several violent pitch oscillations

and lost 5,000 feet. Two passengers were fatally injured and 149 passengers and 7

crewmembers were injured to some degree. As a result of the mishap, the NTSB

recommended that the Douglas Aircraft Company provide data needed to upgrade

the MD-11 simulators to accurately represent the longitudinal stability and control

characteristics of the aircraft (NTSB,1993).

13

Myers et al.: Simulator Fidelity, Training Transfer, and the Role of Instructors

Published by Scholarly Commons, 2018

U.S. Airways Flight 427. September 8, 1994, Flight 427, a Boeing 737,

crashed while maneuvering to land at the Pittsburgh International Airport in

Pittsburgh, Pennsylvania. The aircraft descended uncontrollably and impacted the

terrain killing all 132 people on board and destroying the aircraft. The cause of the

mishap was an un-commanded rudder reversal which caused the aircraft to depart

controlled flight. In the mishap report, the NTSB stated that the simulator

characteristics developed by Boeing and implemented by air carriers in simulators

did not adequately represent the crossover airspeed phenomenon and suggested that

the fidelity be addressed since data is readily available from flight tests (NTSB,

1999).

Tower Air Flight 41. December 20, 1995, Flight 41, a Boeing 747, veered

off the left side of the runway while taking off from John F. Kennedy Airport in

Jamaica, New York. There were no fatalities, but 24 passengers received minor

injuries, one flight attendant received serious injuries, and the aircraft sustained

substantial damage. The primary cause of the mishap was the failure to reject the

takeoff after loss of directional control. The NTSB stated their concern with the

inability of pilots to attain needed training for slippery runway procedures due to

poor simulator fidelity. They also found that while the Boeing simulators provided

a more accurate model, the air carrier simulators did not. Thus, the NTSB

concluded that improvements in slippery runway handling fidelity in 747 flight

simulators were both needed and feasible (NTSB, 1996).

Airborne Express. On December 22, 1996, an Airborne Express Douglas

DC-8 crashed into mountainous terrain near Narrows, Virginia. The three

crewmembers and three maintenance personnel on the functional check flight were

killed and the aircraft was destroyed. The primary cause of the mishap was

improper applied control inputs during a stall recovery attempt. The NTSB

evaluation of the simulator fidelity found that the simulator did not produce the stall

characteristics of the DC-8 with adequate fidelity. Thus, the crew was provided

with a misleading expectation of the aircraft’s handling characteristics. The NTSB

recommended the FAA evaluate all simulator stall characteristics in air carrier

simulators and change them as necessary to represent to the maximum extent

possible the stall characteristics of each aircraft (NTSB, 1997).

Global Exec Aviation Bombardier Learjet Model 60. On September 19,

2008, a Bombardier Learjet overran the runway during a rejected takeoff at the

Columbia Metropolitan Airport in Columbia, South Carolina. Two passengers, the

Captain, and the First Officer were killed and two other passengers were seriously

injured. The right main landing gear tire had separated from the wheel causing

14

International Journal of Aviation, Aeronautics, and Aerospace, Vol. 5 [2018], Iss. 1, Art. 6

https://commons.erau.edu/ijaaa/vol5/iss1/6 DOI: https://doi.org/10.15394/ijaaa.2018.1203

vibration and shaking of the airframe and subsequent failure of the three main

landing gear tires during the attempted takeoff rejection. As part of the

recommendation to provide realistic tire failure training, the NTSB recommended

that the FAA define and codify minimum simulator fidelity for tire failure scenarios

(NTSB, 2010).

Based on their analysis, it is clear the NTSB considers simulator fidelity in

certain tasks a fundamental requirement to safely execute flight maneuvers. These

eight examples offer insight into mishaps where simulator fidelity, while not a

direct cause of the accident, was identified as a contributing factor. If improved,

simulator fidelity may well help prevent similar occurrences. Some industry

stakeholders and researchers would assert that since commercial aviation is

remarkably safe and instances of negative transfer from simulators are rare, the risk

of maintaining the status quo is acceptable. Conversely, it could be argued that

between just the eight mishap recommendations listed above, 410 lives were lost

and that is unacceptable. The moral dilemma is determining the correct balance

between cost-effectiveness and safety. These eight safety recommendations made

by the NTSB, however, reinforce that the agency views increased fidelity as a

fundamental requirement in enhancing safety. One such way to do so is to

understand the ways in which the instructor may impact training transfer.

The Instructor’s Role in Maximizing Training Transfer

The simulator training process is made of three parts that include (a) the

simulator, (b) the training syllabus and associated objectives, and (c) the instructor

(Lee, 2009). The simulator itself does not train, as it is simply a tool used in the

training process. Simulator design and flight training syllabi development receive

significant attention from various stakeholders. The instructor, however, is often

overlooked in this process despite the role of the instructor as a key element in the

success of training (Lee, 2009). It was found that a flight instructor influenced the

student’s progress more than syllabi variations or the simulator (Valverde, 1973).

In the most recent AC 120-54A, Advanced Qualification Program, considered the

agency’s most advanced and dynamic training system, the FAA asserts in the

section Instructors and Evaluators that “Instructors, evaluators, and supervisors are

the backbone of the Advanced Qualification Program” (FAA, 2017, p. 63).

The simulator instructor has many duties and plays many roles. The

instructor must be familiar with simulator capabilities and limitations, know the

lesson objectives, and instruct to attain the best performance from individual

students who vary in attitude, hands-on skill level, and cognitive ability (Lintern et

al., 1989). The instructor’s role is to obtain the highest performance possible from

15

Myers et al.: Simulator Fidelity, Training Transfer, and the Role of Instructors

Published by Scholarly Commons, 2018

the student for a task while optimizing the time in the simulator period to do so.

Simulator instructors must also be familiar with and role-play air traffic controllers,

ground/dispatch/maintenance personnel, and other crew positions throughout

numerous training scenarios.

It is imperative that the instructor identify the differences between the

simulator and the aircraft to prevent known simulator deficiencies from creating a

negative training transfer. This requires an extensive knowledge of and/or

experience in the actual aircraft to be able to discern the differences (Lintern et al.,

1989). The fidelity gaps must be filled in by instructing the students on the

deficiencies of the simulator. Caro (2001a) concluded from his pilot training study

that the instructors tended to concentrate on procedural task during simulator

training and not emphasize the training value of the simulators regarding dynamic

flight tasks.

Order now and get 10% discount on all orders above $50 now!!The professional are ready and willing handle your assignment.

ORDER NOW »»