Situation awareness

Situation awareness is sometimes confused with the term "situational understanding." In the context of military command and control applications, situational understanding refers to the "product of applying analysis and judgment to the unit's situation awareness to determine the relationships of the factors present and form logical conclusions concerning threats to the force or mission accomplishment, opportunities for mission accomplishment, and gaps in information".[27] Situational understanding is the same as Level 2 SA in the Endsley model—the comprehension of the meaning of the information as integrated with each other and in terms of the individual's goals. It is the "so what" of the data that is perceived.

In brief, situation awareness is viewed as "a state of knowledge," and situational assessment as "the processes" used to achieve that knowledge. Endsley argues that "it is important to distinguish the term situation awareness, as a state of knowledge, from the processes used to achieve that state.[1] These processes, which may vary widely among individuals and contexts, will be referred to as situational assessment or the process of achieving, acquiring, or maintaining SA." Note that SA is not only produced by the processes of situational assessment, it also drives those same processes in a recurrent fashion. For example, one's current awareness can determine what one pays attention to next and how one interprets the information perceived.[28]

Accurate mental models are one of the prerequisites for achieving SA.[29][30] A mental model can be described as a set of well-defined, highly organized yet dynamic knowledge structures developed over time from experience.[31][32] The volume of available data inherent in complex operational environments can overwhelm the capability of novice decision makers to attend, process, and integrate this information efficiently, resulting in information overload and negatively impacting their SA.[33] In contrast, experienced decision makers assess and interpret the current situation (Level 1 and 2 SA) and select an appropriate action based on conceptual patterns stored in their long-term memory as "mental models".[34] Cues in the environment activate these mental models, which in turn guide their decision making process.

Klein, Moon, and Hoffman distinguish between situation awareness and sensemaking as follows:

...situation awareness is about the knowledge state that's achieved—either knowledge of current data elements, or inferences drawn from these data, or predictions that can be made using these inferences. In contrast, sensemaking is about the process of achieving these kinds of outcomes, the strategies, and the barriers encountered.[35]

In brief, sensemaking is viewed more as "a motivated, continuous effort to understand connections (which can be among people, places, and events) in order to anticipate their trajectories and act effectively",[36] rather than the state of knowledge underlying situation awareness. Endsley points out that as an effortful process, sensemaking is actually considering a subset of the processes used to maintain situation awareness.[37] In the vast majority of the cases, SA is instantaneous and effortless, proceeding from pattern recognition of key factors in the environment—"The speed of operations in activities such as sports, driving, flying and air traffic control practically prohibits such conscious deliberation in the majority of cases, but rather reserves it for the exceptions." Endsley also points out that sensemaking is backward focused, forming reasons for past events, while situation awareness is typically forward looking, projecting what is likely to happen in order to inform effective decision processes.[37]

A recent holistic framework[38] has shown how SA emerges from causal mental models, where beliefs about causal relationships in the world are represented in the structure of the model and beliefs about the state of the world are reflected in the state of the model. Causal models support inferences about causes and consequences, as well as counterfactual thinking[39] and allow us to ask questions of why?, so-what?, and what-if?, that support SA comprehension and projection as well as naturalistic decision-making. When instantiated as a Bayesian network, these models also allow evaluation of the search for information and the integration of new information with existing knowledge. A computational causal model quantitatively predicted search behavior and changes in SA in fighter pilots.[38]

The SA states framework above extends an older (but simpler), theoretical framework of SA, provided by Dr. Mica Endsley (1995b), which has historically been widely used. Endsley's model describes SA states, and illustrates three stages or steps of SA formation: perception, comprehension, and projection.

Perception (Level 1 SA): The first step in achieving SA is to perceive the status, attributes, and dynamics of relevant elements in the environment. Thus, Level 1 SA, the most basic level of SA, involves the processes of monitoring, cue detection, and simple recognition, which lead to an awareness of multiple situational elements (objects, events, people, systems, environmental factors) and their current states (locations, conditions, modes, actions).

Comprehension (Level 2 SA): The next step in SA formation involves a synthesis of disjointed Level 1 SA elements through the processes of pattern recognition, interpretation, and evaluation. Level 2 SA requires integrating this information to understand how it will impact upon the individual's goals and objectives. This includes developing a comprehensive picture of the world, or of that portion of the world of concern to the individual.

Projection (Level 3 SA): The third and highest level of SA involves the ability to project the future actions of the elements in the environment. Level 3 SA is achieved through knowledge of the status and dynamics of the elements and comprehension of the situation (Levels 1 and 2 SA), and then extrapolating this information forward in time to determine how it will affect future states of the operational environment.

Endsley's model of SA also illustrates several variables that can influence the development and maintenance of SA, including individual, task, and environmental factors. For example, individuals vary in their ability to acquire SA; thus, simply providing the same system and training will not ensure similar SA across different individuals. Endsley's model shows how SA "provides the primary basis for subsequent decision making and performance in the operation of complex, dynamic systems" (Endsley, 1995a, p. 65). Although alone it cannot guarantee successful decision making, SA does support the necessary input processes (e.g., cue recognition, situation assessment, prediction) upon which good decisions are based (Artman, 2000).

Endsley's model of SA. This is a synthesis of versions she has given in several sources, notably Endsley (1995a) and Endsley et al (2000). Drawn by Dr. Peter Lankton, May 2007.

SA also involves both a temporal and a spatial component. Time is an important concept in SA, as SA is a dynamic construct, changing at a tempo dictated by the actions of individuals, task characteristics, and the surrounding environment. As new inputs enter the system, the individual incorporates them into this mental representation, making changes as necessary in plans and actions in order to achieve the desired goals. SA also involves spatial knowledge about the activities and events occurring in a specific location of interest to the individual. Thus, the concept of SA includes perception, comprehension, and projection of situational information, as well as temporal and spatial components.

In summary, the model consists of several key factors:

• Perception, comprehension, and projection as three levels of SA,
• The role of goals and goal directed processing in directing attention and interpreting the significance of perceived information,
• The role of information salience in "grabbing" attention in a data-driven fashion, and the importance of alternating goal-driven and data-driven processing,
• The role of expectations (fed by the current model of the situation and by long-term memory stores) in directing attention and interpreting information,
• The heavy demands on limited working memory restricting SA for novices and for those in novel situations, but the tremendous advantages of mental models and pattern matching to prototypical schema that largely circumvent these limits,
• The use of mental models for providing a means for integrating different bits of information and comprehending its meaning (relevant to goals) and for allowing people to make useful projections of likely future events and states,
• Pattern matching to schema—prototypical states of the mental model—that provides rapid retrieval of comprehension and projection relevant to the recognized situation and in many cases single-step retrieval of appropriate actions for the situation.

For a more complete description of the model, see Endsley (1995b) and Endsley (2004). See also Endsley (2000) for a review of other models of SA.

Any model of situation awareness depends on cognitive and shared cognitive processes, and yet '...models of SA refer to cognitive processes in general terms, but do not specify exactly what processes are involved and to what extent.' (Banbury & Tremblay, 2004, p. xiii). This criticism is an example of the difficulty that cognitive science has in addressing a concept such as SA, which through its definition and assumptions appears to stand robustly, however when the theorized processes are exposed at the cognitive level of analysis assumptions must be radically reviewed. Researchers have regularly raised these concerns - notably in Flach (1995) and more recently in Banbury & Tremblay (2004). To date the most widely cited model of SA is lacking in support from cognitive science, one notable observation that still stands is that:

'The test of situation awareness as a construct will be in its ability to be operationalized in terms of objective, clearly specified independent (stimulus manipulation) and dependent (response difference) variables ... Otherwise, SA will be yet another buzzword to cloak scientists' ignorance.' (Flach, J., 1995, p. 155)

Another criticism of the model comes from a study done in 2015 which looked at situational awareness in tasks where relevant knowledge about the task could be found through other mediums, other than directly asking the collaborator. It was found that in these types of tasks, verbal communication lengthens the time it takes to complete a task when compared to people completing a task individually.[40]

Team SA is defined as "the degree to which every team member possesses the SA required for his or her responsibilities" (Endsley, 1995b, p. 39; see also Endsley, 1989). The success or failure of a team depends on the success or failure of each of its team members. If any one of the team members has poor SA, it can lead to a critical error in performance that can undermine the success of the entire team. By this definition, each team member needs to have a high level of SA on those factors that are relevant for his or her job. It is not sufficient for one member of the team to be aware of critical information if the team member who needs that information is not aware.

In a team, each member has a subgoal pertinent to his/her specific role that feeds into the overall team goal. Associated with each member's subgoal are a set of SA elements about which he/she is concerned. Team SA, therefore, can be represented as shown in Figure 2. As the members of a team are essentially interdependent in meeting the overall team goal, some overlap between each member's subgoal and their SA requirements will be present. It is this subset of information that constitutes much of team coordination. That coordination may occur as a verbal exchange, a duplication of displayed information, or by some other means.

Shared situation awareness can be defined as "the degree to which team members possess the same SA on shared SA requirements" (Endsley & Jones, 1997, p. 47; 2001, p. 48). As implied by this definition, there are information requirements that are relevant to multiple team members. A major part of teamwork involves the area where these SA requirements overlap—the shared SA requirements that exist as a function of the essential interdependency of the team members. In a poorly functioning team, two or more members may have different assessments on these shared SA requirements and thus behave in an uncoordinated or even counter-productive fashion. Yet in a smoothly functioning team, each team member shares a common understanding of what is happening on those SA elements that are common—shared SA. Thus, shared SA refers to the overlap between the SA requirements of the team members, as presented in Figure 3. As depicted by the clear areas of the figure, not all information needs to be shared. Clearly, each team member is aware of much that is not pertinent to the others on the team. Sharing every detail of each person's job would only create a great deal of "noise" to sort through to get needed information. It is only that information which is relevant to the SA requirements of each team member that is needed.

The situation awareness of the team as a whole, therefore, is dependent upon both (1) a high level of SA among individual team members for the aspects of the situation necessary for their job; and (2) a high level of shared SA between team members, providing an accurate common operating picture of those aspects of the situation common to the needs of each member (Endsley & Jones, 2001). Endsley and Jones (1997; 2001) describe a model of team situation awareness as a means of conceptualizing how teams develop high levels of shared SA across members. Each of these four factors—requirements, devices, mechanisms and processes—act to help build team and shared SA.

1. Team SA requirements – the degree to which the team members know which information needs to be shared, including their higher level assessments and projections (which are usually not otherwise available to fellow team members), and information on team members' task status and current capabilities.

2. Team SA devices – the devices available for sharing this information, which can include direct communication (both verbal and non-verbal), shared displays (e.g., visual or audio displays, or tactile devices), or a shared environment. As non-verbal communication, such as gestures and display of local artifacts, and a shared environment are usually not available in distributed teams, this places far more emphasis on verbal communication and communication technologies for creating shared information displays.

3. Team SA mechanisms – the degree to which team members possess mechanisms, such as shared mental models, which support their ability to interpret information in the same way and make accurate projections regarding each other's actions. The possession of shared mental models can greatly facilitate communication and coordination in team settings.

4. Team SA processes – the degree to which team members engage in effective processes for sharing SA information which may include a group norm of questioning assumptions, checking each other for conflicting information or perceptions, setting up coordination and prioritization of tasks, and establishing contingency planning among others.

Objective measures directly assess SA by comparing an individual's perceptions of the situation or environment to some "ground truth" reality. Specifically, objective measures collect data from the individual on his or her perceptions of the situation and compare them to what is actually happening to score the accuracy of their SA at a given moment in time. Thus, this type of assessment provides a direct measure of SA and does not require operators or observers to make judgments about situational knowledge on the basis of incomplete information. Objective measures can be gathered in one of three ways: real-time as the task is completed (e.g., "real-time probes" presented as open questions embedded as verbal communications during the task – Jones & Endsley, 2000), during an interruption in task performance (e.g., situation awareness global assessment technique (SAGAT) – Endsley, 1995a, or the WOMBAT situational awareness and stress tolerance test mostly used in aviation since the late 1980s and often called HUPEX in Europe), or post-test following completion of the task.

Subjective measures directly assess SA by asking individuals to rate their own or the observed SA of individuals on an anchored scale (e.g., participant situation awareness questionnaire (PSAQ) – Strater, Endsley, Pleban, & Matthews, 2001; the situation awareness rating technique (SART) – Taylor, 1989). Subjective measures of SA are attractive in that they are relatively straightforward and easy to administer. However, several limitations should be noted. Individuals making subjective assessments of their own SA are often unaware of information they do not know (the "unknown unknowns"). Subjective measures also tend to be global in nature, and, as such, do not fully exploit the multivariate nature of SA to provide the detailed diagnostics available with objective measures. Nevertheless, self-ratings may be useful in that they can provide an assessment of operators' degree of confidence in their SA and their own performance. Measuring how SA is perceived by the operator may provide information as important as the operator's actual SA, since errors in perceived SA quality (over-confidence or under-confidence in SA) may have just as harmful an effect on an individual's or team's decision-making as errors in their actual SA (Endsley, 1998).

Subjective estimates of an individual's SA may also be made by experienced observers (e.g., peers, commanders, or trained external experts). These observer ratings may be somewhat superior to self-ratings of SA because more information about the true state of the environment is usually available to the observer than to the operator, who may be focused on performing the task (i.e., trained observers may have more complete knowledge of the situation). However, observers have only limited knowledge about the operator's concept of the situation and cannot have complete insight into the mental state of the individual being evaluated. Thus, observers are forced to rely more on operators' observable actions and verbalizations in order to infer their level of SA. In this case, such actions and verbalizations are best assessed using performance and behavioral measures of SA, as described next.

Performance measures "infer" SA from the end result (i.e., task performance outcomes), based on the assumption that better performance indicates better SA. Common performance metrics include quantity of output or productivity level, time to perform the task or respond to an event, and the accuracy of the response or, conversely, the number of errors committed. The main advantage of performance measures is that these can be collected objectively and without disrupting task performance. However, although evidence exists to suggest a positive relation between SA and performance, this connection is probabilistic and not always direct and unequivocal (Endsley, 1995b). In other words, good SA does not always lead to good performance and poor SA does not always lead to poor performance (Endsley, 1990). Thus, performance measures should be used in conjunction with others measures of SA that directly assess this construct.

Behavioral measures also "infer" SA from the actions that individuals choose to take, based on the assumption that good actions will follow from good SA and vice versa. Behavioral measures rely primarily on observer ratings, and are, thus, somewhat subjective in nature. To address this limitation, observers can be asked to evaluate the degree to which individuals are carrying out actions and exhibiting behaviors that would be expected to promote the achievement of higher levels of SA (see, for example, the situation awareness behaviorally anchored rating scale (SABARS) – Matthews, Pleban, Endsley, & Strater, 2000; Strater et al., 2001). This approach removes some of the subjectivity associated with making judgments about an individual's internal state of knowledge by allowing them to make judgments about SA indicators that are more readily observable.

Process indices examine how individuals process information in their environment, such as by analyzing communication patterns between team members or using eye tracking devices. Team communication (particularly verbal communication) supports the knowledge building and information processing that leads to SA construction (Endsley & Jones, 1997). Thus, since SA may be distributed via communication, computational linguistics and machine learning techniques can be combined with natural language analytical techniques (e.g., Latent Semantic Analysis) to create models that draw on the verbal expressions of the team to predict SA and task performance (Bolstad, Cuevas, Gonzalez, & Schneider, 2005; Bolstad, Foltz, Franzke, Cuevas, Rosenstein, & Costello, 2007). Although evidence exists to support the utility of communication analysis for predicting team SA (Foltz, Bolstad, Cuevas, Franzke, Rosenstein, & Costello, in press), time constraints and technological limitations (e.g., cost and availability of speech recording systems and speech-to-text translation software) may make this approach less practical and viable in time-pressured, fast-paced operations.

Psycho-physiological measures also serve as process indices of operator SA by providing an assessment of the relationship between human performance and a corrected change in the operator's physiology (e.g., French, Clark, Pomeroy, Seymour, & Clarke, 2007). In other words, cognitive activity is associated with changes in the operator's physiological states. For example, the operator's overall functional state (as assessed using psycho-physiological measures, such as electroencephalographic (EEG) data, eyeblinks, and cardiac activity) may provide an indication as to whether the operator is sleep fatigued at one end of the continuum, or mentally overloaded at the other end (Wilson, 2000). Other psycho-physiological measures, such as event related potentials (ERP), event related desynchronization (ERD), transient heart rate (HR), and electrodermal activity (EDA), may be useful for evaluating an operator's perception of critical environmental cues, that is, to determine if the operator has detected and perceived a task-relevant stimulus (Wilson, 2000). In addition, it is also possible to use psycho-physiological measures to monitor operators' environmental expectancies, that is, their physiological responses to upcoming events, as a measure of their current level of SA (Wilson, 2000).

The multivariate nature of SA significantly complicates its quantification and measurement, as it is conceivable that a metric may only tap into one aspect of the operator's SA. Further, studies have shown that different types of SA measures do not always correlate strongly with each other (cf. Durso, Truitt, Hackworth, Crutchfield, Nikolic, Moertl, Ohrt, & Manning, 1995; Endsley, Selcon, Hardiman, & Croft, 1998; Vidulich, 2000). Accordingly, rather than rely on a single approach or metric, valid and reliable measurement of SA should utilize a battery of distinct yet related measures that complement each other (e.g., Harwood, Barnett, & Wickens, 1988). Such a multi-faced approach to SA measurement capitalizes on the strengths of each measure while minimizing the limitations inherent in each.

In first aid medical training provided by the American Red Cross, the need to be aware of the situation within the area of influence as one approaches an individual requiring medical assistance is the first aspect for responders to consider[46] Examining the area and being aware of potential hazards, including the hazards which may have caused the injuries being treated, is an effort to ensure that responders do not themselves get injured and require treatment as well.

Situation awareness for first responders in medical situations also includes evaluating and understanding what happened[47] to avoid injury of responders and also to provide information to other rescue agencies which may need to know what the situation is via radio prior to their arrival on the scene.

In a medical context, situation awareness is applied to avoid further injury to already-injured individuals, to avoid injury to medical responders, and to inform other potential responders of hazardous conditions prior to their arrival.

A loss in situational awareness has led to many transportation disasters, including the 2015 Philadelphia train derailment.[48]

Within the search and rescue context, situational awareness is applied primarily to avoid injury to search crews however being aware of the environment, the lay of the land, and the many other factors of influence within one's surroundings assists in the location of injured or missing individuals[49] Public safety agencies are increasingly using situational awareness applications like ATAK on mobile devices to improve situational awareness.

In the United States Forest Service the use of chainsaws and crosscut saws requires training and certification.[50] A great deal of that training describes situational awareness as an approach toward environmental awareness but also self-awareness[51] which includes being aware of one's own emotional attitude, tiredness, and even caloric intake.

Situational awareness in the forest context also includes evaluating the environment and the potential safety hazards within a saw crew's area of influence. As a sawyer approaches a task, the ground, wind, cloud cover, hillsides, and many other factors are examined and are considered proactively as part of trained sawyers' ingrained training.

Dead or diseased trees within the reach of saw team crews are evaluated, the strength and direction of the wind is evaluated. The lay of tree sections to be bucked or the lean of a tree to be felled is evaluated within the context of being aware of where the tree will fall or move to when cut, where the other members of the saw team are located, how they are moving, whether hikers are within the area of influence, whether hikers are moving or are stationary.

Law enforcement training includes being situationally aware of what is going on around the police officer before, during, and after interactions with the general public[52] while also being fully aware of what is happening around the officer in areas not currently the focus of an officer's immediate task.

In cybersecurity, consider situational awareness, for threat operations, is being able to perceive threat activity and vulnerability in context so that the following can be actively defended: data, information, knowledge, and wisdom from compromise. Situational awareness is achieved by developing and using solutions that often consume data and information from many different sources. Technology and algorithms are then used to apply knowledge and wisdom in order to discern patterns of behavior that point to possible, probable, and real threats.

Situational awareness for cybersecurity threat operations teams appears in the form of a condensed, enriched, often graphical, prioritized, and easily searchable view of systems that are inside or related to security areas of responsibility (such as corporate networks or those used for national security interests). Different studies have analyzed the perception of security and privacy in the context of eHealth,[53] network security,[54] or using collaborative approaches to improve the awareness of users.[55] There are also research efforts to automate the processing of communication network information in order to obtain or improve cyber-situational awareness.[56]

Scientists provided three standard levels of SAT in addition to a fourth level which included the agent's level of uncertainty in its decision in unmanned vehicles. The stated goal of this research was to determine how modifying levels of SAT affected user performance, situation awareness, and confidence in the agent. The scientists stated that their experimental results support that increased agent transparency improved the performance of the operator and human confidence on the agent without a significant effect on the workload. When the agent communicated levels of uncertainty in the task assigned, those involved in the experimentation displayed more trust in the agent.[59]

The ASM research was conducted by providing a simulation game in which the participant had to complete a training course with an ASM, a ground robot that communicates with infantry. The participants had to multitask, evaluating potential threats while monitoring the ASM's communications on the interface. According to that research, experimental results demonstrated that the greatest confidence calibration occurred when the agent communicated information of all three levels of SAT.[59] The group of scientists from the U.S. Army Research Laboratory developed transparency visualization concepts in which the agents can communicate their plans, motivations, and projected outcomes through icons. The agent has been reported to be able to relate its resource usage, reasoning, predicted resource loss, progress towards task completion, etc.[57] Unlike in the IMPACT research, the agent informing the user of its level of uncertainty in decision making, no increase in trust was observed.[59]

Crowdsourcing, made possible by the rise of social media and ubiquitous mobile access has a potential for considerably enhancing situation awareness of both, responsible authorities and citizens themselves for emergency and crisis situations by employing or using "citizens as sensors".[60][61][62][63][64][65][66][67] For instance, analysis of content posted on online social media like Facebook and Twitter using data mining, machine learning and natural language processing techniques may provide situational information.[68] A crowdsourcing approach to sensing, particularly in crisis situations, has been referred to as 'crowdsensing'.[69] Crowdmapping is a subtype of crowdsourcing[70][71] by which aggregation of crowd-generated inputs such as captured communications and social media feeds are combined with geographic data to create a digital map that is as up-to-date as possible[72][73][74][75] that can improve situational awareness during an incident and be used to support incident response.[76]

Since 2012 the National Information Sharing Consortium (NISC) has worked to provide "the right information to the right people at the right time" by use of common terminology among the emergency management community and first-responders with a mission of standardizing the structured geo-spatial data to be shared online over a variety of platforms. The result is to create a Common Operating Picture (COP) that generates accurate and timely information displayed visually, both at the strategic level for decision makers and at the tactical level for people on site. The NISC promotes the sharing of code, data widgets, and training in order to increase the quality of the situational awareness. Large scale exercises like the Capstone-14 week-long event coordinated by the Central United States Earthquake Consortium (CUSEC) and actual operational use of these data sharing methods have also advanced the work of NISC to expand the usability of GIS-based information sharing for enhanced situational awareness.

There are two training scenarios designed to increase the situational awareness skills of military professionals, and first responders in police and emergency services. The first, Kim's Game, has a more common place in the Marine Corps sniper school and police academies. The name is derived from the novel Kim which references the game to a spy school lesson. The game involves a tray with various items such as spoons, pencils, bullets, and any other items the soldiers would be familiar with. The participants are given one minute to view all of these items before they are covered up with a blanket. The participants would then individually list the items that they saw, the one with the most correct answers would win the game.[77] The same game is played in young scouting and girl guide groups as well to teach children quick memorisation skills.

The second method is a more practical military application of Kim's Game. It starts with a field area (jungle, bush or forest) of about five meters wide to 10 meters deep where various items, some camouflaged and some not, to be located in the area on the ground and in the trees at eyesight level. Again, these items would be ones that are familiar to the soldiers undergoing the exercise. The participants would be given 10 minutes to view the area from one place and take a mental note of the items they saw. Once their 10 minutes is up, the soldier would then be required to do a repetition of certain exercises such as burpees, designed to simulate the stress of a physically demanding environment. Once the participant completes the exercise, they would list the items they saw. The points would be tallied in the end to find the winner.[77]