Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
1 Tr ~ e e eam . ransoms Across a wide number of different systems in our society a common situation may be found: a team of operators, functioning for some period of time under relatively routine conditions, is abruptly confronted with abnor- maI and sometimes emergency circumstances, to which they must rapidly respond in an appropriate fashion. We call this the team transition situation. Let us consider a number of diverse examples, to realize the importance and . . . . . ubiquity of this situation. (1) The crew of a nuclear power control room becomes aware of a transient event somewhere in the reactor and within minutes (or even sec- onds) must try to establish the appropriate procedures to restore or maintain plant safety. This process may sometimes require a high level of cognitive activity as the nature of a fault is diagnosed (Rubinstein and Mason, 1979~. (2) Closely parallel to the nuclear power plant incident is the response of the crew of a commercial airline, which suddenly becomes aware of a life-threatening malfunction- the loss of an engine or, in the case of United Airlines flight 232, complete loss of hydraulic power near Sioux City, Iowa (Predmore, 1991~. Here again the team (flight deck personnel working with air traffic controllers and maintenance personnel) must rapidly perceive the problem and engage in problem-solving behavior to ensure the safety of the aircraft. (3) On a quiet night, personnel in a hospital emergency room are sud- denly confronted with the victims of a serious automobile accident. Rapid problem solving and perhaps prioritization of possible causes must be coupled 13
4 WORKLOAD TRANSITION with the precise exercise of procedural skills and coordination with the emergency room staff. (4) A firefighting unit is suddenly called to a multialarm blaze in the middle of the night. Multiple units arrive on the scene within minutes and find themselves confronted with a burning chemical plant with potential toxic and explosive material (Klein, 1989~. (5) Shipboard personnel respond to an explosion that not only results in a fire that they must combat, but also destroys the integrity of the hull and therefore the seaworthiness of the vessel (Wagenaar and Groeneweg, 1988). (6) A tornado, hurricane, or earthquake severely damages a commu- nity. Within minutes or hours a major disaster relief effort must be mobi- lized to care for wounded and homeless residents and attempt to restore necessary services. (7) An emergency medical service (EMS) team is suddenly called on to make a helicopter flight to pick up a critically ill patient. The flight will take place at night, in bad weather, to an unfamiliar destination. (8) Finally, an army tank crew stands ready and waiting at the edge of a battlefield, its crew of four having waited in a state of combat readiness for 36 hours. Hostile shots suddenly land close by, and the crew must immediately become an effective fighting unit. The situation confronted by this crew will provide the focus for much of the material in this report. BACKGROUND The Army plans to develop tanks with reduced crew sizes. To do this, tasks now performed by at least one crew member will be automated or redistributed among the remaining crew positions. During active deploy- ment, the reduction in crew size is likely to increase workload, thus increas- ing the potential for performance failures and errors unless compensatory measures are devised. Aside from a potentially high combat workload, the tank crew of the future may be faced with the additional problem of a long period of work underload and inactivity prior to active engagement in combat. Military doctrine calls for crews to remain with their tanks for periods as long as 72 hours before being called into combat or withdrawn. During such a period of passive waiting, the tasks of the crew will consist mainly of monitoring communication channels and visual displays, sleeping, and so forth. This period of low-level alertness may suddenly be followed by a high level of activity, when the order to engage the enemy is received. At this juncture, when stress and workload are high, decrements in the ability to focus atten- tion, to implement and prioritize activities, and to retain information in short-term memory may be differentially impaired.
TEAM TRANSITIONS 15 The U.S. Army Human Engineering Laboratory requested that the Com- mittee on Human Factors conduct a study to provide advice and guidance on the effects of prolonged work underload on the subsequent performance of critical tasks and on approaches that could be employed to offset or compensate for decrements in performance that otherwise might occur. Given descriptions of the critical tasks required of each crew member at the onset of combat, the study had three objectives: (1) To review the concept of workload and the state of research knowl- edge of its effects on subsequent high workload task performance; (2) To evaluate the components of the critical tasks at combat onset to assess which components are most likely to be vulnerable to decrements from prior underload or the sudden onset of high workload; and (3) To identify and evaluate techniques that might be used by future tank planners and designers to compensate for, or offset, likely performance decrements. To obtain further information about the issues involved, staff received an orientation on current tank design and discussed the potential problem of underload followed by overload. Individuals with expertise in workload were then contacted to obtain their views about the effects of underload on sudden performance of important tasks. It was the consensus of those contacted that, while the concept of work underload-overload transition has been given little, if any, attention and even less research emphasis, it is nevertheless an important problem encountered in many work settings rang- ing from naval patrols in enemy waters (e.g., the Persian Gulf) to rescue squads and other public emergency service workplaces. However, for the purposes of this study the tank crew is the predominant focus. TEAM PERFORMANCE DURING TRANSITIONS Although there are many case studies of situations in which teams failed to perform effectively when faced with an abrupt transition from one level of workload to another, little research has been performed to examine team performance under such circumstances. The Three Mile Island incident (Rubinstein and Mason, 1979) and the grounding of the ship Exxon Valdez (National Transportation Safety Board, 1990) are examples of situations in which a team failed to respond appropriately. At other times, crews have performed these functions routinely and professionally, reflecting the fact that most such transitions go unnoticed. At still other times, the heroic and effective efforts of teams in times of crisis are worthy of note. Here the recent performance of the crew of United Flight 232 can be singled out in bringing the severely crippled airliner, without its steering control, to the ground (Predmore, 1991~.
6 WORKLOAD TRANSITION The point here is not to argue that performance in team transition situa- tions is either exceptionally good or exceptionally poor. Rather, it is to suggest that there is a confluence of elements defining a large class of events, which potentially impact the health and safety of a large number of people, about which little is known. Transition situations are characterized by: (1) a team of people who must coordinate and communicate; (2) an event that occurs at an unexpected time following a long duration of wait- ing; and (3) the impact of the event that may produce (a) substantial risk to the health and safety of the team and possibly others; (b) an increase, or qualitative change in workload; and (c) a concomitant increase in stress. This chapter briefly describes the duties of the tank crew and the char- acteristics of the tank environment. It provides an operational context for the in-depth analysis of human performance problems with teams in transi- tion that are treated in subsequent chapters. The chapter ends with an overview of some of the main features of the transition process. CREW PERFORMANCE REQUIREMENTS The duties of the four members of a tank crew (commander, gunner, driver, and loader), as well as the operators of many complex systems, may be grouped into seven categories: (1) planning, preparation, and organiza- tion; (2) monitoring; (3) inflation seeking; (4) decision making; (5) main- tenance; (6) control and operation; and (7) communications. The require- ment to perform each of these functions varies between mission segments and between crew members. Principles drawn from previous research can be used to develop an expected workload profile for each crew member and mission phase. However, unexpected events, emergencies, or the incapaci- tation of one crew member can alter the workload that the crew actually experiences. Planning and Preparation Before a mission or engagement, planning and preparation are the pri- mary sources of workload for a tank crew, particularly for the tank com- mander. The lead time for formulating plans may be hours or even days. Plans are developed at every level of command and transmitted down the chain of command through briefings, written orders, or radio transmissions. Planning is based on intelligence reports, which may or may not be current, and is driven by military doctrine and strategic goals. This type of planning is generally more self-motivated than event-driven, and it may or may not occur under external time pressure. If planning and preparation are suffi- cient, then performance during the subsequent engagement or mission is
TEAM TRANSITIONS 1 17 more likely to be acceptable. If events occur as expected and if few changes in the plans are required, then the workload of individual tank crews will be much lower than if planning is insufficient or inappropriate. During an engagement or once a mission has begun, planning and preparation occupy less of the crew's time. Previously developed plans have estab- lished the goals and tactics that are followed, at least until events occur that make these plans inoperative. At this point, individual tank crews may have to act independently, and control begins to shift down the chain of com- mand. The need to change the plan or develop a new plan is driven by rapidly evolving events external to the tank (see Chapter 9~. However, an individual tank commander's knowledge of the global situation may be extremely limited at this point; and less information is available up the chain of command about the position, status, and intentions of individual tanks or units. When planning and decision making must be performed on- line, time pressure may be acute, and the plans that are developed tend to focus on immediate events, rather than long-term strategies. The workload of planning in real time is high. Routine Monitoring Before a mission or while preparing for an engagement, tank crews are responsible for some types of monitoring (see Chapter 6~. The crew of each tank must stand watch, and members of a unit share the responsibility for mutual security. If the proposed reduction in tank crew complement occurs, standing watch will require more time for the remaining crew members, further contributing to fatigue problems already encountered in the field with crews of four (see Chapter 5~. Most of the monitoring activities asso- ciated with the tank and its weapons systems are preventive. In general, premission monitoring is of relatively long duration, self-motivated (rather than driven by changes in the situation), and precautionary. During a mission, continual monitoring is required to maintain situ- ational awareness (see Chapters 7 and 8~. The situation may change very rapidly and tank crews must rely on their own assessment of the local situation rather than on formal briefings. Thus, monitoring activities im- pose significantly higher workload during a mission than before. When the tank is moving, the driver is responsible for monitoring the state of the vehicle, the gunner is responsible for monitoring the weapons systems; the loader is responsible for monitoring ammunition supplies; and the tank commander must monitor the global situation inside and outside the tank. Each member of the tank crew must ensure that the others are aware of relevant informa- tion. The tank commander must transfer such information to other members of the unit and up the chain of command.
18 WORKLOAD TRANSITION Maintenance Before a mission, tank crews devote considerable time to maintaining their equipment (e.g., the tank, weapons systems, personal equipment), themselves (e.g., resting, eating), and resupply. These activities are generally routine, established by procedures and schedules. Depending on the time pressure under which these tasks are performed, workload may be very low or mod- erately high. Each crew member has specific duties that must be performed at regular intervals. The tank commander is responsible for ensuring that each crew member has completed his duties, as well as performing his own duties. He is also responsible for ensuring that each crew member has sufficient rest and food. If a defensive position is being established, the driver and loader may experience relatively high physical workload while positioning and concealing the tank. The primary difference between maintenance activities performed be- fore and during a mission is that the former are routine and proceduralized, whereas the latter are performed in response to problems or driven by events. During an engagement, only the most critical maintenance or repairs are performed in response to an urgent need. Rest and meals no longer occur at regular intervals or according to a schedule, but rather when the situation allows. Resupply becomes far less predictable as well; tank crews may have to make do with what they can carry with them for extended periods of time. Because maintenance performed during a mission is generally not routine, it is likely to impose higher workload. Furthermore, the support system available to facilitate maintenance before a mission may not be available once an engagement begins. Information Seeking Before a mission, intelligence briefings provide the primary source of information. The plans and strategies that are evolved are relatively static (although they may change as new information is received). If adequate briefings are given, crews within a tank, unit, or company share common information about the situation and each other's intentions. During this phase, situational assessment is primarily predictive in nature. During a mission, the situation is no longer static, and gathering infor- mation about friendly as well as enemy forces is more difficult. Informa- tion transmitted by radio between tanks and intercom systems within a tank may be piecemeal and confusing. Each tank commander must update his mental model of the situation by piecing together incomplete information from a variety of sources and transmit relevant information to members of his crew, the unit, and up the chain of command. Obtaining information
TEAM TRANSITIONS 19 directly from the surrounding environment becomes an important element of the crew's duties during a mission. When the tank is moving, remaining geographically oriented is criti- cally important, imposes higher workload, and is more compensatory than predictive (see Chapter 7~. The tank commander and driver must work together to correlate the visual scene with information on maps to determine their current and projected position. The tank commander must work with the gunner to identify and classify targets as they are observed and transmit information about their location. All of these information-seeking, inter- preting, integrating, and sharing tasks are performed under time pressure and stress with only minimally sophisticated instruments. Thus, workload is much higher and performance may suffer. Uncertainty and complexity are greater, information quality is reduced, interruptions and interference are common, and the crew has little control over the rate or sequence of events. Decision Making At each level of command, the leader must select an appropriate strat- egy, given the mission goals and current situation, and adopt tactics that support the selected strategy. He must determine and convey the tank's position, situation, intentions, and ability to execute the original plan up the chain of command. If a tank's current position puts the crew in jeopardy or the crew will not be able to achieve its assignment, the tank commander must select a new position and the safest route to get there. If failures occur or if the tank is hit by enemy fire, the commander must determine the extent of damage and choose the most appropriate action, [based on status reports from the crew. Furthermore, the act of classifying and prioritizing targets and selecting the most appropriate weapons is an important decision that can affect the degree to which a tank succeeds or fails in accomplishing its goal and maintaining its security. Control and Operation Before a mission, tank control and operation of weapons systems plays a very small role in a tank crew's workload. Depending on whether a defensive or offensive position is being established, the crew may or may not have to move to a new position. However, this activity generally fol- lows a well-specified plan and may impose relatively low workload. During a mission, however, the tank may move rapidly over unfamiliar terrain. Vehicle control demands most of the driver's attention and consid- erable attention from the tank commander. The task is dynamic and unpre
20 WORKLOAD TRANSITION dictable and can impose relatively continuous workload. Workload is in- creased further if visibility is limited by terrain, weather, smoke, or restric- tions in forward visibility and field of view when the crew is closed up in the tank. The gunner is responsible for locating and aiming at targets with the primary weapons systems, although the tank commander shares this responsibility. The primary source of workload for the gunner in this phase is determining the range of a target and aiming accurately, while verbally coordinating his actions with the loader, the tank commander, and the driver. Communications Both before and during a mission, radio communications among tanks and intercom communications within a tank may impose significant workload. Before a mission, however, considerably more information is transmitted face to face through formal and informal briefings. This type of interaction is not possible during a mission, when the radio is the only link between tanks and when noise and physical separation between crew members makes direct verbal interactions difficult (see Chapter 8~. Particularly in the midst of a battle, communications between and within channels may overlap and transmission clarity is reduced. Although the format of transmissions is highly stylized, the content and organization of messages transmitted under stress is less than optimal, therefore increasing the workload of the recipi- er~t. During a mission, communications are likely to focus on the immedi- ate situation only. Particularly for the tank commander, monitoring the radio and transmitting information impose a significant amount of workload during a mission. Within a tank, the shared intercom frequency not only facilitates coordination, but also creates a significant monitoring demand for all crew members. Both before and after transition, the efficiency of communications and the coordination of cooperative actions may be gov- erned very much by the leadership offered by the tank commander (see Chapter 104. WORKSTATION CHARACTERISTICS The physical environment of the tank crew establishes the context within which the above-mentioned aspects of performance are carried out. Crew workstation design has the potential for either enhancing or degrading crew performance. The MlA1 Main Battle Tank is a 60-plus ton mobile work- place for its crew members (Figure 1.11. Accurate weapons delivery is its primary purpose. Most of the design emphasis has thus been placed on propulsion, armor, and fire control systems. In an effort to produce a safe, or at least survivable, battlefield environment for crew members, the tank has been designed to be extremely mobile, present a small target to enemies,
TEAM TRANSITIONS 2 it.. ~ .~ ~ . if. . ~ .. . .,~..~,,~, ~.~,~..~.~.,~. ..~ ~ ~ ~ .. ~ ~ i. ~ ~ . ~ ~.~ ~ ~ ~ , ~ . . . i, ~ ~ ., ~.~.~ ....... ~ ~ ~ . ~ ~ .. .~ ~. . ~ .~ ~.~ ~ . .. .. ~.~ ~,~ .~ I. . ~ ~ ~..~ ~ . .~ i. ~ . , ~ ~ it.. ~ ~ ~ ~ ~ ~ ... .,~, ~ ~ i, ,., ,,,, ,,,., i, ,,, I,, ~ i. .~'~.~ ~.~'~'~'~ ''~"'~'~'~ '~''~'~'~.~''' ~. ,,,, ~,~,~-~: ~ ~ ~ ~ ~ , . ~ ~ ~ ~ ~ ~ . . ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ . ~ ~ ~ . ~ , : ~ ~ ~ . ~ ~ . . , ~ . ~ ~ . ~ ~ ~ ~ ~ ~ ~ . ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ . ~ ~ ~ ~ ~ ~ ~ ~ . ~ ~ ~ ~ ~ . ~ . ~ ~ : ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~'~'.~ ~'~.~.'~2.''~'.~ ~ : . . ~ . ~ ~ ~ I. , ~ ~ ~ ~ ~.~.~. ~ ~ ,~.. ~,.~ ~ .,. ~ . ~. ~.~ ~..~ ..- .... ~ id. .:.. ~..~ ~..~..~. ~.~ :..: .~:~ ~ ~ ::::: ::~:::: ::~:::::::::::::::::::: :::~: ,..,... FIGURE 1.1 Photo of an MlA1 Main Battle Tank. Courtesy of the U.S. Army Human Engineering Laboratory. provide a structurally sound shield against enemy ordnance, and provide a capsule that can be sealed against chemical/biological weapons (referred to as "buttoned ups. The crew workstations inside a tank (see, for example, Figure 1.2) are well defined, as are the normal work tasks of each crew member. There is one workstation for each of the four crew members. Each workstation consists of a seat and the displays, controls, and weapons that are required for the crew member to perform his job function. The driver is stationed in the front of the tank, beneath the main gun and turret. His seat is recumbent with a movable headrest. The steering yoke, throttle, transmission, and brake controls are positioned at the front of the driver's workstation. The driver is responsible for monitoring all engine parameters as well as the status of the hydraulic systems in the tank. He can look outside the vehicle by placing his head through a hatch in the tank deck. When the tank is buttoned up, the driver sees the outside world through three fixed vision blocks (narrow windows with ballistic glass) that are 7 1/4 inches long and 1 3/4 inches high built into the hatch cover. The other three crew members are stationed inside the turret of the tank. The loader's workstation is just to the left of the main gun breech.
WORKLOAD TRANSITION FIGURE 1.2 Top-down view of the tank commander's workstation in an MlA1 Main Battle Tank. Courtesy of the U.S. Army Human Engineering Laboratory. His seat is oriented slightly sideways to allow him to take ammunition from the storage locker on his right (at the rear of the turret) and transfer it to the main gun breech on his left. When the main gun is fired, the recoil causes the breech to move backward in the turret. The loader must move his legs and other appendages out of the way of this recoil. A hatch is located directly above his head. When the tank is open he can see the outside world through the hatch by standing on his seat. When the tank is buttoned up, the loader views the outside world through vision blocks. The gunner has the most complex workstation in the tank because he has been provided with the most advanced technology. He is positioned to the right of the main gun and slightly below the loader's level. His work- station consists of a seat and all the aiming and other fire control systems in the tank. In addition to the main sighting system, the gunner's workstation includes an auxiliary sighting system, a laser rangefinder, a thermal imaging system, and a ballistic fire control computer. The gunner has the main responsibility for acquiring targets visually, tracking them with the aiming system, adjusting any fire control solutions in the computer, and firing the main gun. The gunner is second-in-command of the tank crew; he takes over in the event that the tank commander is unable to function.
TEAM TRANSITIONS 23 The tank commander's workstation is located directly behind and above the gunner. It consists of a seat, all communication controls, the controls for actuating certain auxiliary armaments on the tank (e.g., grenade launch- ers, smoke generators, and chaff canisters), and a main gun firing mecha- nism to be used if the gunner's main firing switch is inoperative or the gunner is disabled. The tank commander is responsible for the overall operation of the tank and its crew. He is directly responsible for the tactical movement and placement of the tank, selection of targets, decisions about the type of ammunition to be used, the final decision to fire the main gun, and all communication with both his counterparts in other tanks in his platoon and the higher command structure. Several aspects of tank crew workstations fall short of acceptable hu- man factors design criteria. One characteristic common to all four worksta- tions is the lack of seatbelts or any other mechanism (such as pneumatic chair suspension) that might counteract the sudden and strong accelerations associated with vehicle movement. The driver's seat is the only one in the tank that is recumbent, allowing the driver to recline. The driver's seat provides much more body support than the other seats in the tank. This support, together with the availability of structures that allow the driver to brace himself for sideways accelerations, allows the driver to function un- der much higher accelerations than the other crew members. It is a com- monly accepted fact among tank crews that a driver can operate the tank in such a manner as to cause injury to the other crew members while causing none to himself. At least two of the tank crew workstations have obvious problems with the placement of controls and displays. In the driver's workstation, the steering and propulsion controls are well placed, allowing the driver to look outside the vehicle and steer the tank much like an automobile when the tank is not buttoned up. In many cases, he must rely on the commander's verbal steering commands and drive blind or drive with offset eyepoint (i.e., similar to using a periscope to see the outside world). However, many of the displays that the driver is supposed to monitor are placed orthogonal to his line of vision, so that he must lean forward and look sideways to see them. In the gunner's workstation, the main gunsight is centrally located in the gunner's field of vision and well within his reach. However, other components, most notably the fire control computer console, are placed in extremely awkward locations. Another common workstation design problem concerns ingress and egress It is quite difficult to get into and out of certain crew member workstations. The gunner's station is probably the most difficult to get into and out of, since it has no direct hatch access. The gunner must come through the tank commander's or the loader's hatch and then thread his way down into hi workstation. Such access problems are a direct result of competing design
24 WORKLOAD TRANSITION goals. As mentioned earlier, tank designers endeavor to make the tank as small as possible, as fast as possible, and as powerful in firepower, as possible. Crew member workstations are probably relegated to a fairly low priority in the overall design scheme. FRAMEWORK OF THE REPORT Given this description of the tank environment, features of which are elaborated in the report, we now turn to Figure 1.3, which provides a frame- work for the report. It describes the variables that characterize the transi- tion to action, the performance variables that are most affected by the tran- sition, and the variables that moderate the effect of the transition on performance and cognition. As mentioned before, this report focuses its attention most heavily on a single system the tank and its crew-that we see as proto- typical of many of the other systems in transition to be described in Chapter 2. It is our hope in doing so that we shall: (a) provide specific guidelines for human performance modeling and human factors engineering in the tank environment, (b) provide conclusions and offer a data base that will gener- alize to other teams in transition, and (c) stimulate research that may gener- alize to all such teams. Four key changes characterize the transition (Part A of Figure 1.31: (1) psychological stress level, a qualitative change from the anxiety and fear of waiting to the high workload, time pressure, and mortal danger of the battle; (2) environmental change from quiet to noisy and possibly clear to smoky; (3) cognitive mode changes from a good deal of proactive planning and routine following of procedures (pre) to a reactive mode involving rapid decision making and problem solving (post); and (4) target search and ac- quisition transitions from a low to a high probability of finding a target. Part B shows processes that are affected by the transition and will be the focus of subsequent chapters in this report: (1) target search, detection, and recognition; (2) spatial orientation, navigation, and situation awareness; (3) decision making; (4) strategic planning; (5) communications and team coor- dination; (6) following of procedures; and (7) perceptual-motor interaction. It should be noted that all of these activities characterize sustained battle- field activity, as well as the immediate post-transition period. However, it is likely that spatial orientation, navigation and situation awareness, and strategic planning will undergo a particularly intense increase during the immediate post-transition period. Finally, Part C is a partial list of variables that can affect or moderate the response to the transition: sleep loss, stress, human factors design, and training. Stress represents both a cause (Part A) and a moderating variable that can attenuate or enhance the effects of the transition on performance (Part C). Human factors design of the displays, controls, and working
TEAM TRANSITIONS A 25 Stress Anxiety ... Combat,Workload, Danger Environment Quiet Cognition Proactive Planning Target Events Low Noise Reactive Problem Solving High - B - W Search Situation Awareness Decision Making Strategic Planning Communications and Team Coordination Procedures Following Perceptual-Motor Interaction C _ in' - - - Sleep Loss Stress Human Factors Design Training FIGURE 1.3 Variables that (Al characterize the transition, (B) are most affected by the transition, and (C) moderate the effect of the transition on performance and cognition. environment may be assumed to influence transition performance in two respects: (a) good human factors design will improve the quality of perfor- mance as demands increase during the post-transition period and (b) good human factors principles may be applied specifically to address the prob- lems of updating situation awareness in the crisis period immediately fol- lowing the transition. This second application is quite analogous to the
26 WORKLOAD TRANSITION concerns expressed in the nuclear industry for displays that will support the rapid diagnosis of complex faults (Woods, 1988~. Finally, training may be the single most important variable to influence performance during transi- tion. Training may focus on the skills that are disrupted during transition or on mitigating the detrimental effects of stress. This framework sets a context for the chapters that follow. The first part of the report contains information from analogous systems that may provide insight on the variables that may affect performance during the transition and examines the workload domain--which is the overriding con- text that interacts with moderating variables to affect performance. Chapter 2 presents an analysis of the features of similarity and contrast between the tank environment and other analogous systems and references relevant case studies and accident reports in other domains. Chapter 3 reviews workload: its sources, measurement, consequences, relationship to performance, and so forth. The report then focuses on key variables that influence crews' abilities to adapt to a transition: stress (Chapter 4) and sleep disruption and fatigue (Chapter 5~. The report turns then to the specific domains of tank crew performance that may be influenced by the transition process: vigilance and target detection (Chapter 6), situation and geographic awareness (Chapter 7), decision making (Chapter 8), strategic task management (Chapter 9), and team structure, communications, and leadership (Chapter 10~. Finally, it discusses possible remediations of some of the moderating variables such as training (Chapter 11) and integrates and synthesizes the panel's most important recommendations (Chapter 12~. Two final points should be noted. First, our listing of specific tank crew behaviors is not exhaustive (i.e., Part B of Figure 1.3~. The panel selected a subset of tasks specific to particular members of the tank crew, primarily the commander and the gunner, as the focus of analysis. This selection was not random but was based on information obtained from a series of briefings provided by professional armor personnel at Aberdeen Proving Ground and in Washington, DC. These briefings revealed certain critical areas of vulnerability to transition-related effects, on which we have focused our analysis. Second, the chapters that follow, while maintaining a consistent structural format, vary to some degree in the extent to which they emphasize the broader issue of transitions across various systems or the specific tank system, emphasizing, but not exclusively, the transition as- pects of the tank. REFERENCES Klein, G.A. 1989 Recognition-primed decisions. Pp. 47-92 in W. Rouse, ea., Advances in Man- Machine Systems Research, Volume 5. Greenwich, Connecticut: JAI Press.
TEAM TRANSITIONS Rubinstein 1 979 27 National Transportation Safety Board 1990 Marine Accident Report~-ounding of the U.S. Tankship EXXON VALDEZ on Thigh Reef' Prince William Sound, near Valde, Alaska, March 24, 1989. Report No. NTSB/Mar-90/04. Washington, DC: National Transportation Safety Board. Predmore, S.C. 1991 Micro-coding of cockpit communications in accident analyses: Crew coordination in the United Airlines flight 232 accident. In R.S. Jensen, ea., Proceedings of the 6th international Symposium on Aviation Psychology. Columbus, Ohio: Ohio State University, Department of Aviation. , T., and A.F. Mason The accident that shouldn't have happened: An analysis of Three Mile Island. IEEE Spectrum 16:33 -57. Wagenaar, W.A., and J. Groeneweg 1988 Accidents at sea: Multiple causes and impossible consequences. Pp. 133-144 in E. Hollnagel, G. Mancini, and D.D. Woods, eds., Cognitive Engineering in Com- plex Dynamic Worlds. London: Academic Press. Woods, D.D. 1988 Commentary: Cognitive engineering in complex and dynamic worlds. Pp. 115- 129 in E. Hollnagel, G. Mancini, and D.D. Woods, eds., Cognitive Engineering in Complex Dynamic Worlds. London: Academic Press.