Professional Documents
Culture Documents
Dickens 2002
Dickens 2002
discussions, stats, and author profiles for this publication at: http://www.researchgate.net/publication/247535919
CITATIONS READS
72 235
1 AUTHOR:
Christopher D Wickens
Alion Science and Technology
438 PUBLICATIONS 14,087 CITATIONS
SEE PROFILE
All in-text references underlined in blue are linked to publications on ResearchGate, Available from: Christopher D Wickens
letting you access and read them immediately. Retrieved on: 31 October 2015
128 VOLUME 11, NUMBER 4, AUGUST 2002
psychologists have represented and ond, situation awareness is relevant or information processing, chal-
studied these phenomena, and the to dynamic, evolving situations, lenges: First, the system has six
particular research challenges they and therefore is not the same as the variables that have to be monitored
present. more static knowledge of long-term and controlled as they change over
To understand the intertwining memory (e.g., knowledge of how time: three variables of orientation
of situation awareness and work- systems work or appropriate proce- around the axis of the aircraft—
load, consider a single pilot ap- dures to follow, mental models). pitch, roll, and yaw (heading)—and
proaching for a landing in rapidly Third, content or the product of sit- three position variables—altitude,
deteriorating weather. Already uation awareness is distinct from lateral deviation from a flight path,
overloaded just in keeping the the process of maintaining situation and position along a flight path.
plane aloft and tracking the air- awareness (Adams, Tenney, & Pew, Second, all of these six variables are
craft’s instruments along the flight 1995). The pilot who is situationally linked, or “cross-coupled,” in the
path, the pilot does not initially no- aware has rapid access to an accu- flight dynamics; for example, pitch
tice a failure of one of the naviga- rate mental representation of the determines future altitude, and roll
tional instruments. After noticing changing environment that is determines future heading, future
this failure, and while trying to di- broader and more enduring than lateral deviation, and future pitch.
agnose the nature of the instru- that which can be held within the The skilled pilot represents these
ment problem, the pilot loses very restricted capacity of working linkages and constraints in a mental
awareness of both the present off- memory, or “consciousness.” The model (Wickens, 1999). Third, these
course path and a large thunder- pilot with good situation awareness causal sequences produce time
storm that appears on the radar may not be consciously thinking lags, lasting several seconds, be-
display. Finally, after noticing the about the fact that there is an air- tween changes in the orientation
plane’s path and taking steps to craft close by to the side, but if variables, which can be directly
correct it, in this overloaded state, called upon suddenly to respond controlled, and the positional
the pilot has forgotten to lower the appropriately to this situation, the changes that result. The cognitive
landing gear. Thus, our pilot, be- pilot will do so rapidly and accu- challenges of anticipation when
cause of the high mental workload, rately because of the ability to rap- tracking systems with lags (Wickens
fails to maintain awareness of the idly access the information from & Hollands, 2000) impose an addi-
surrounding environment and the memory (Kintsch & Ericsson, 1995). tional mental workload beyond
state of the aircraft, and at various Thus, finally, good situation aware- that associated with the multivari-
points “sheds” certain critically im- ness supports the response to the able complexity of the system. (The
portant tasks. unexpected (Wickens, 1999). Hence, perceived lags between an air-traf-
aircraft design features that enhance fic controller’s instructions to a pi-
routine performance may inhibit lot and the change in the aircraft
situation awareness, creating a “tun- visible on the controller’s display
SITUATION AWARENESS neling of attention” focusing pri- are even greater than the lags for
marily on expected events and tasks. the pilot, and the cognitive chal-
Endsley (2000) defined situation Three aspects of situation lenges of anticipation are thus am-
awareness as “the (1) perception awareness that are most relevant plified for air-traffic control; Wick-
[noticing] of the elements in the en- for aviation are three-dimensional ens, Mavor, & McGee, 1997). Fourth,
vironment within a volume of time (3-D) spatial awareness, system the pilot must coordinate some-
and space, the (2) comprehension (mode) awareness, and task aware- times-conflicting goals of aviating
of their meaning, and the (3) projec- ness. I discuss each in turn, before (maintaining the proper orientation
tion of their status in the near fu- showing how all three are directly to preserve lift) and navigating
ture” (p. 5). This three-stage con- linked to the issues of workload (moving from point to point in 3-D
struct can be further understood in and task management. space). Fifth, such coordination is
terms of three important features. carried out in a 3-D space that itself
First, it involves cognition and contains numerous dynamic haz-
working memory (the temporary Spatial Awareness ards (weather, other traffic), whose
store of information that is rapidly 3-D trajectories cannot themselves
forgotten if not rehearsed), rather The concept of spatial awareness be predicted with certainty.
than action and response. Good sit- is inherent in the task of moving an To address these multiple chal-
uation awareness may support aircraft through a 3-D space filled lenges to spatial awareness, de-
good choice of action, but is not in- with hazards. The pilot in control signers are endeavoring to create
herently a part of that choice. Sec- confronts the following cognitive, instrument displays with 3-D
volume is represented on a 2-D Designers need to understand aware of the changes of their flight
viewing surface, 3-D ecological these three general issues and inte- modes that are relevant to the safety
views (e.g., viewpoint B in Fig. 1) grate the principles of visual atten- of flight.
create ambiguity in precisely lo- tion, spatial cognition, and manual
cating entities in the airspace control to craft the configuration of
(Wickens, in press). In contrast, aircraft displays that will support Task Awareness, Task
more analytical 2-D representa- all of the needed tasks, without over- Management, and Workload
tions avoid those ambiguities, loading attentional capabilities.
but impose greater attentional The pilot actually has four differ-
demands and cognitive load. For ent generic classes of tasks to per-
example, in the case of viewpoint System Awareness form. I have described the tasks of
C in Figure 1, the pilot must inte- aviating and navigating already.
grate the vertical and lateral dis- Just as the aircraft and its air- The two additional tasks are com-
plays in order to attain a mental space are complex and dynamic, so munication (on the flight deck and
“picture” of the 3-D airspace, also are the systems within the air- with air-traffic control) and systems
which is needed to fly a 3-D tra- craft. In modern aircraft, designers management (e.g., managing fuel,
jectory (Wickens & Prevett, 1995), have provided a host of automated cabin pressure, electricity). These
or mentally rotate the displays flight-control systems in an effort to four tasks are arranged somewhat
into a forward view, which is relieve pilots’ workload. Thus, in- on a priority hierarchy in the order
more characteristic of a 3-D view. creasing computer power has en- in which I have just referred to
• The prediction issue concerns the abled aircraft to perform many ac- them (aviating, navigating, com-
degree of prediction, or “look tions—status monitoring, situation munications, and systems manage-
ahead” capabilities, that should inference, and changing the mode ment, or “ANCS” for short), al-
be built into a display. For both of flight (e.g., climb, cruise, de- though the hierarchy has some
pilots and controllers, prediction scent)—that pilots need not usually flexibility. The busy pilot must al-
capabilities support Level 3 be aware of, unless the unexpected ways be aware of what tasks need
SA—the ability to deal with situation arises. The complexity of to be performed, and in what order
lagged systems. Reliable predic- many of these automated systems, (Adams et al., 1995). The number of
tive displays (e.g., seen in Fig. 1, coupled with poorly designed sym- such tasks is so extensive and, in
viewpoints A and C) are always bolic displays, makes system aware- some cases, their performance is so
helpful for controlling the air- ness difficult to maintain (Sarter & vital that a host of checklists pro-
craft. However, the reliability of Woods, 1995). Furthermore, the fact vide reminders of what to do and
prediction in an airspace full of that people remember actions that when to do it, thereby circumvent-
uncertainties regarding future they themselves have initiated bet- ing the frailties of human prospec-
wind, weather, and the pilot’s ter than those initiated by another tive memory (memory for things to
own control actions will be less agent (in this case, automation) be done in the future). Yet such
than perfect. An automated sys- poses an additional challenge to au- proceduralized checklists, reinforc-
tem can provide inferences tomation-system (or mode) aware- ing the knowledge acquired from
about the future (e.g., viewpoint ness. Finally, the high workload im- pilot training and experience, fall
A in Fig. 1 shows where the air- posed by other aspects of flying short in several respects (Wickens,
craft is predicted to be a few sec- may further degrade the pilots’ 1999). First, in some situations, per-
onds in the future), but may monitoring of automation-con- forming two or more tasks at once,
prove wrong, and incorrect pre- trolled devices (Parasuraman & Ri- or rapidly switching between them,
dictions may destroy the hu- ley, 1997; Parasuraman, Sheridan, & is desired, if not required. Check-
man’s trust in the system, Wickens, 2000). How to exploit the lists provide no guidance on how to
thereby negating its benefits workload-reducing advantages of do this. Second, unexpected events
(Parasuraman & Riley, 1997). It automation while still keeping the can require actions that could never
is known that the reliability of pilot adequately “in the loop,” be fully laid out in written proce-
prediction decreases as the pre- aware of actions taken by auto- dures. Third, many cognitive tasks,
dicted interval increases. But de- mated systems, provides a tremen- such as maintaining situation
termining precisely how long dous challenge for aviation psychol- awareness, cannot easily be codified
this interval should be to sup- ogy (Adams et al., 1995). Recent in checklists and procedures, yet
port automation usage, without psychological research on change performance of such ongoing tasks
engendering mistrust, remains a detection should provide impor- is vital in supporting the response
major research challenge. tant insights into how to keep pilots to the unexpected.
One important domain of avia- ness is inherent in the fact that fectively addressed these issues in
tion research concerns the “rules” of maintaining situation awareness is isolation. However, understand-
task management and task switch- itself a task, requiring the allocation ing and then modeling the complex
ing. Research in this area has inves- of mental resources to the processes interactions among these phenom-
tigated both the tendencies that of selective attention (i.e., selec- ena remains a critical challenge
characterize pilots in general and tively attending to some events in posed by aviation to psychological
those that discriminate better from the environment and ignoring oth- researchers who are interested in
more poorly performing pilots. In ers) and of working memory or “scaling up” their theories to real-
the former category are studies that short-term memory. In this regard world problems. Overlaid upon
have examined task interruptions— it is important to realize that those this complexity are the fascinating
the extent to which the ANCS hier- attention-management strategies challenges of understanding how
archy is maintained when a task that support routine performance powerful yet imperfect computer
with a lower or higher priority ar- are likely to narrow the focus of at- automation, in predictor displays,
rives while another task is ongoing. tention on the parameters of flight warnings, inference making, and
Some evidence suggests that audi- control and potential hazards in the flight management, affects pilots’
tory tasks low on the ANCS hierar- forward path, but this narrow focus situation awareness, mental work-
chy, and particularly auditory com- does not necessarily support situa- load, and task management (Para-
munication tasks, tend to be both tion awareness. For example, a dis- suraman et al., 2000).
more interrupting and less inter- play (e.g., viewpoint A in Fig. 1)
ruptible than tasks with a higher that channels attention to the for-
priority (e.g., navigation). Studies ward flight path inhibits attention
Recommended Reading
comparing better and more poorly to surrounding hazards even when
performing pilots have indicated these are displayed elsewhere in Endsley, M.R., & Garland, D.J. (Eds.).
that better multitask performance the cockpit (Wickens, in press; (2000). Situation awareness analysis
results from rapid switching be- Wickens & Prevett, 1995). Thus, a and measurement. Mahwah, NJ: Erl-
baum.
tween tasks (Wickens, 1999). good pilot must allocate attention
Tsang, P., & Vidulich, M. (Eds.). (in
Task-management research has both to sources of information for press). Principles and practices of
a natural link to more traditional routine performance and to infor- aviation psychology. Mahwah, NJ:
studies of attention as a perceptual mation from the broader environ- Erlbaum.
phenomenon because highly sa- ment (in anticipation of unexpected Wiener, E.L., & Nagel, D.C. (Eds.).
(1989). Human factors in aviation.
lient perceptual events within a events), and training in task-man-
Orlando, FL: Academic Press.
task can capture attention and agement skills can help to maintain
thereby remind the pilot that the situation awareness. A good dis-
task needs to be performed. With- play designer must create displays
out such events, a task might be ne- that can effectively integrate the Note
glected. One direct linkage of task representation of the two classes of
management to situation aware- information. Finally, effective mod- 1. Address correspondence to
ness is inherent in the extent to els of mental workload should be Christopher Wickens, University of Illi-
which the pilot or air-traffic con- able to predict the circumstances in nois, Aviation Human Factors Divi-
sion, Willard Airport, 1 Airport Rd.,
troller notices that a task should be which the workload of routine per- Savoy, IL 61874.
attended to as a result of an auto- formance is raised to such a level
mated alerting signal (Level 1 SA), that resources are not available to
and retains an understanding that maintain situation awareness and
the task should be done now (in the latter task is shed. References
the face of competing task de-
mands, Level 2 SA), or in the future Adams, M.J., Tenney, Y.J., & Pew, R.W. (1995). Sit-
uation awareness and the cognitive manage-
(Level 3 SA). Remembering to take ment of complex systems. Human Factors, 37,
care of future requirements (such CONCLUSION 85–104.
Endsley, M.R. (2000). Theoretical underpinnings
as when an air-traffic controller re- of situation awareness. In M.R. Endsley &
members to take an arriving air- I have described an interlinking D.J. Garland (Eds.), Situation awareness analy-
sis and measurement (pp. 1–21). Mahwah, NJ:
craft out of a “holding stack” set of cognitive phenomena relat- Erlbaum.
around an airport) depends on ing to awareness, aircraft control, Gibson, J.J. (1979). The ecological approach to visual
prospective memory. attention, mental resources, and stra- perception. Boston: Houghton Mifflin.
Kintsch, W., & Ericsson, K.A. (1995). Long term
A second linkage between task tegic task management. Much ba- working memory. Psychological Review, 102,
management and situation aware- sic research in psychology has ef- 211–245.
Parasuraman, R., & Riley, V. (1997). Humans and Sarter, N., & Woods, D. (1995). How in the world Wickens, C.D., & Hollands, J. (2000). Engineering
automation: Use, misuse, disuse, abuse. Hu- did we ever get into that mode? Human Fac- psychology and human performance (3rd ed.). Up-
man Factors, 39, 230–253. tors, 37, 5–19. per Saddle River, NJ: Prentice Hall.
Parasuraman, R., Sheridan, T.B., & Wickens, C.D. Wickens, C.D. (1999). Cognitive factors in avia- Wickens, C.D., Mavor, A.S., & McGee, J.P. (Eds.).
(2000). A model for types and levels of human tion. In F. Durso (Ed.), Handbook of applied cog- (1997). Flight to the future: Human factors in air
interaction with automation. IEEE Transactions nition (pp. 247–282). New York: John Wiley & traffic control. Washington, DC: National Acad-
on Systems, Man, & Cybernetics, 30, 286–297. Sons. emy Press.
Previc, F.H., & Ercoline, W.R. (1999). The “outside- Wickens, C.D. (in press). Aviation displays. In P. Wickens, C.D., & Prevett, T.T. (1995). Exploring
in” attitude display concept revisited. The In- Tsang & M. Vidulich (Eds.), Principles and prac- the dimensions of egocentricity in aircraft nav-
ternational Journal of Aviation Psychology, 9, tices of aviation psychology. Mahwah, NJ: Erl- igation displays. Journal of Experimental Psy-
377–401. baum. chology: Applied, 1, 110–135.