• No results found

Information overload in Swedish emergency response command and control functions

N/A
N/A
Protected

Academic year: 2021

Share "Information overload in Swedish emergency response command and control functions"

Copied!
48
0
0

Loading.... (view fulltext now)

Full text

(1)

I

Information overload

in Swedish emergency response

command and control functions

(2)

II Kontaktadresser: Erik Prytz erik.prytz@liu.se Linköpings universitet SE - 581 83 Linköping

Center for Advanced Research in Emergency Response (CARER) Centrum för forskning inom respons- och räddningssystem (CARER) URL: http://www.liu.se/forskning/carer

E-Post: carer@liu.se

CARER rapport nr 35 ISBN: 978-91-7929-656-8 Granskad av: Susanna Lönnqvist

Rapporten än godkänd för publicering i CARER:s rapportserie

Publicerad av Linköping University Electronic Press URL: www.ep.liu.se

E-post: ep@ep.liu.se

Detta verk skyddas enligt lagen om upphovsrätt (URL 1960:729). Upphovsrätten ägs av författaren, 2021.

(3)

III

Förord

Denna rapport har skrivits inom ramen för det tvärvetenskapliga forskningsprogrammet ”Effektiv räddning på framtidens skadeplats” som fokuserar på komplexiteten i det framtida räddningssystemet, till exempel med avseende på resurser, teknik, metodutveckling och utvärdering. ”Effektiv räddning på framtidens skadeplats” är femårigt och finansieras gemensamt av Myndigheten för samhällsskydd och beredskap (MSB) och Linköpings universitet (LiU). Det bedrivs inom Centrum för forskning inom respons- och

räddningssystem (CARER) som också är att samarbete mellan LiU och MSB.

Sammanfattning

Informationsöverbelastning är fenomenet där mängden tillgänglig information en

beslutsfattare har tillgång till är så stor att den i sig leder till försämrade förutsättningar för beslutsfattande. Informationsöverbelastning är ett potentiellt framtida problem inom ledningssystem för respons och räddning, då nya informationskällor och mer effektiva datainsamlingsmöjligheter kan leda till en kraftigt ökad mängd information. En studie i två delar genomfördes för att undersöka informationsöverbelastning. Den första delen bestod av en litteraturöversikt över fenomenet både individuellt sett och från ett systemperspektiv. Den andra delen var en intervjustudie med 13 deltagare från svenska blåljusorganisationer som inkluderar polis, räddningstjänst och akutsjukvård. Deltagarna som intervjuades arbetar med ledning av räddning- och responsinsatser för mindre och större händelser på olika nivåer, från det operativa till det strategiska perspektivet. En tematisk analys på intervjumaterialet visade på fem teman rörande informationsöverbelastning i det svenska räddningssystemet:

1. Man behöver information för att förstå lägesbilden, och man behöver en lägesbild för att förstå informationen

2. Informationshantering är en medveten och manuell process

3. Information är nyckeln till framgång men all information är inte användbar 4. Informationsöverbelastning orsakad av mängden information är inte ett problem 5. Informationsöverbelastning är ett strukturellt problem

Sammantaget visar dessa teman att beslutsfattare i de svenska räddningssystemen arbetar manuellt, aktivt och medvetet med information och att de förlitar sig på sin tidigare erfarenhet för att kunna utföra sitt uppdrag effektivt. Vidare förlitar beslutsfattarna sig på själva

(4)

IV meningsfull och relevant för uppdraget och beslutsfattarens roll i ledningssystemet.

Informationsöverbelastning sågs inte som ett aktuellt problem, då ledningsstrukturen ska fungera som ett filter mot irrelevant eller dåligt strukturerad information. Dock såg de intervjuade att informationsöverbelastning kan bli ett problem om beslutsstöd eller nya informationsflöden införs i ledningssystemen, utan att dessa är utvecklade specifikt för att fungera i det existerande arbetsflödet i ledningssystemen.

(5)

V

Abstract

A current issue in the command and control (C2) community is information overload. With more information available and more efficient information-gathering capabilities, modern C2 systems are faced with the problem that the human operators will be confronted with an information environment that may adversely impact their ability to do their jobs. To gain insight into this problem, an interview study was conducted focusing on Swedish emergency response C2 operators to investigate their views on information overload and how the

phenomenon may be related to their work. Thirteen participants were recruited from three different Swedish agencies that engage in emergency response work: the police, rescue service, and medical emergency response. A thematic analysis was conducted to identify themes related to how Swedish emergency response C2 operators work with information and how they view information overload. A total of five themes were identified, showing that operators assess and integrate information manually to support their decision-making, mainly relying on their experience to do so, and that operators rely on the control structures of the C2 system to provide them with structured, rich, and relevant information. Operators do not view information overload as a present threat to their work environment, arguing that the C2 systems provide enough support to protect them from unnecessary information load.

However, if information overload were to become an issue, operators felt that this would be caused by poorly implemented decision aids, and new information flows that would erode the control structures that support the operators.

(6)

VI

Contents

Sammanfattning ... III Abstract ... V 1. Introduction ... 1 1.1. Purpose ... 2 2. Background ... 2

2.1. Concept of information overload ... 2

2.1.1. Information processing capacity ... 2

2.1.2. Information processing requirements ... 3

2.2. Information Overload in Command and Control environments ... 4

2.2.1. Event detection ... 8

2.2.2. Contextual control ... 9

2.2.3. Situation Awareness ... 11

2.3. Combating information overload in C2 ... 12

2.3.1. Information collection through data fusion ... 12

2.3.2. Information evaluation ... 13 3. Technical solutions ... 14 3.1. VACCINE ... 14 3.1.1. SMART ... 15 4. Method ... 16 4.1. Participants ... 16 4.2. Interview design ... 16 4.3. Interview procedure ... 17 4.4. Apparatus ... 17 4.5. Analysis ... 17

5. Results and analysis ... 18

(7)

VII 5.1.1. You need information to understand the operational picture, and you need the

operational picture to understand the information ... 18

5.1.2. The managing of information is a conscious and manual effort ... 20

5.1.3. Information is key, but not all information is useful ... 22

5.2. How do Swedish emergency response C2 workers view the problem of information overload? ... 25

5.2.1. IO caused by information load is not a problem ... 25

5.2.2. Information overload as a structural problem ... 26

6. Discussion ... 28 6.1. Results ... 28 6.2. Method ... 33 7. Conclusions ... 34 7.1. Future research ... 35 References ... 36

(8)

1

1.

Introduction

In their book, “Command and Control (C2) Re-envisioned”, Vassiliou, Alberts, and Russel Agre (2014) examined four mega-trends that constitute the driving factors for development in command and control (C2). One of these mega-trends was “Ubiquitous Data”, referring to the trend towards ever-growing data exposure that faces C2 systems. This poses new challenges for managing and processing raw data in C2, both for the human operator and for the system itself (Vassiliou et al., 2014). One such challenge is the risk of overloading operators with information. The human operator has long been established as a limited information processor, e.g., from Miller’s (1956) early research, which showed that humans can only process a certain number of bits of information at one time. This idea of the limited human information processor gave birth to the concept of information overload (IO), which refers to the state in which a human must contend with an information load that they are unable to process. The worry is that with the increasing amount of information being provided to the C2 operator, the likelihood that operators will suffer from IO increases. The concern about IO in C2 is mainly that it has adverse effects on decision-makers, potentially causing increases in decision-making time and reducing decision quality. In C2 environments, which are often engaged in high-risk activities such as firefighting, warfare, and rescue operations, these adverse effects may have catastrophic consequences.

C2 is practised by many different parties, ranging from the military to the private sector, and the methodologies and structures adopted within those systems may vary greatly depending on the purpose and goals of a C2 system. This study will focus on C2 as practised by vital societal functions tasked with emergency response in Sweden. These are functions that consist of agencies and institutions dedicated to responding to serious events and reducing the harm and risk they pose; e.g., police, the rescue services, and emergency medical care (MSB, 2013). These C2 systems are centred around the effort to coordinate and focus resources to achieve a desired effect in keeping with the goals of the C2 system. A Swedish emergency response C2 system is not always a formal organisation, but rather refers to all the different functions and their relations that are gathered under one management structure in order to achieve the overall system’s goal (MSB, 2013). These systems are often hierarchical in nature and may consist of one or several subsystems with hierarchies of their own (Hagel, 2013; MSB, 2013). These hierarchies are usually defined by levels of command, with each level managing decisions according to an operative time axis. The exact nature of this may vary from system to system, but generally there are strategic levels, tactical levels, and operative

(9)

2 levels (Cedergårdh & Wennström, 1998). In certain C2 systems, these levels may be referred to as outer and inner command, where outer command operates at an operative level from the scene of the incident whilst inner command operates at a strategic level detached from the minute-to-minute decision-making of the operation.

1.1. Purpose

The purpose of this study is to investigate IO in Swedish emergency response C2

environments. To accomplish this, relevant literature regarding C2 and IO has been reviewed to provide insight into the problem area. Then an interview study was devised to investigate how Swedish emergency response C2 operators work with information and view the issue of IO. A thematic analysis was used to analyse the interview data and identify themes that provide insight into the operators’ views on their information work and how this work relates to information and IO.

2.

Background

2.1. Concept of information overload

Due to the surplus of information available at all times in today’s society, IO is an issue that plagues most working environments where there is a reliance upon information to make decisions (Dorneich, Whitlow, Ververs, & Rogers, 2003; Eppler & Mengis, 2004; Hellar, 2009). Simply put, IO means having too much information to the point where it becomes detrimental to performance (Eppler & Mengis, 2004). Although there are many definitions of IO, the one most relevant to this study is that overload occurs when information processing requirements surpass information processing capacity (Eppler & Mengis, 2004; Tushman & Nadler, 1978). There are thus two important aspects constituting IO: the information

processing capacity of the individual operator, and the information processing requirements imposed on that operator.

2.1.1. Information processing capacity

When discussing information processing, a good place to start is with the concept of bits; that is, the number of bits of information that can be processed by a human. A bit of information refers to the amount of information that is needed to distinguish between two equally likely alternatives (Miller, 1953, 1956). To distinguish between four alternatives, two bits of information are needed. Each time the alternatives are increased by a factor of two, the number of bits of information needed is increased by one (Miller, 1956). The reason why bits

(10)

3 of information is an important concept in terms of IO is that the human capacity to process such bits is limited and the number that need to be processed may vary.

Channel capacity is a concept that refers to the previously mentioned limitations in the number of bits that can be processed and produce output. Miller (1956) argued that channel capacity has an upper limit of around 2.5 bits of information. This means that an individual can effectively respond to a maximum of six different stimuli at once. Beyond this point, the operator is argued to enter a state of confusion in which they are more prone to erroneous behaviour. A good example of this is provided by Giompapa, Farina, Gini, Gaziano, and Di Stefano (2006), who recorded that, when tracking an object, a radar operator could not integrate more than six information channels into their decision-making before feeling overloaded.

There are of course exceptions to this channel capacity, as Miller (1956) himself points out, and individuals exist who can surpass this capacity by quite a large margin. Although this paper is not concerned with the most extreme of these cases, what should be understood is how different human factors can affect a human’s processing capacity. Hendy, Farell, and East (2001) argue that an operator’s emotional state, e.g. motivation, stress, fatigue, is directly connected with how much of their capacity they can access. Other human factors at the

individual level, such as experience, subjective time assessment, and personal skill, are also argued to have an effect on channel capacity (Eppler & Mengis, 2004; Hendy et al., 2001; Jackson & Farzaneh, 2012; Wright, Chen, Barnes, & Boyce, 2015).

The important thing to appreciate regarding human information processing capacity is that it is limited and that trying to work beyond those limitations can have dire consequences.

2.1.2. Information processing requirements

Much as a physical task places a certain strain on an individual’s physical resources, so do cognitive tasks, like decision-making, put a strain on cognitive resources. The amount of strain on an individual’s cognitive resources is dictated partially by the inherent processing requirements of the information itself, and partially by the environment and context within which the information is to be processed.

Hendy, Farell, and East (2001) present a model that exemplifies the contextual aspects of information processing requirements. In their model, a human’s efficiency as an information processor is dictated by two factors: decision time and time pressure; where decision time is the product of task load and processing rate, while time pressure is the product of decision

(11)

4 time and time available. These subfactors, such as task load, are good examples of

information processing requirements. Task load is referred to by Hendy et al. (2001) as the amount of information (in bits) that needs to be processed to output a decision. Consequently, if the operator is not able to match the processing rate of bits of information to the task load then that operator will see an increase in decision time, which in turn influences the time available for future tasks and increases task load, thus, causing a downward spiral into performance loss (Hendy et al., 2001).

Requirements are not only set by the environment and the nature of a task. They are also set by the nature of the data that is used in completing the task. Schneider (1987) argues that the notion that IO is only caused by the amount of data is too simplistic. Amount of data to be processed is certainly linked to IO, but it is not the sole reason why information processing requirements increase. By this, Schneider (1987) means that information characteristics are equally responsible for causing increases in processing requirements. For example, data can be ambiguous, which increases the time it takes to interpret it, and it can also be novel in such a way that it lacks pre-existing models that can be used to aid in its interpretation (Schneider, 1987). Data characteristics such as these increase the requirements placed on the already-limited information processing mechanics of the operator, and thus contribute directly to overload (Jackson & Farzaneh, 2012; Schneider, 1987). As mentioned above in section 2.1.1, individual human factors are a key aspect of an individual’s capacity to process information. This is also true in relation to the characteristics of information. An experienced individual may rely on previous experience to map novel information onto a pre-existing decision template to help with interpretation, while a novice may struggle with complex information due to inexperience (Jackson & Farzaneh, 2012). Individual factors such as experience can thus be both a detriment and a boon in the case of regulating information processing requirements.

2.2. Information Overload in Command and Control environments

McCann and Pigue (1998) define command and control (C2) as the process of establishing a common intent in order to achieve coordinated action. Command is defined as “the creative expression of human will necessary to accomplish the mission” (p.5), and control as “those structures and processes devised by Command to manage risk” (p.4). Although there is a conscious focus on command in these definitions, the two Cs of C2 complement each other (Mccann & Pigeau, 1998).

(12)

5 The above definition captures the essential purpose of C2, but it does lack the characteristics that define a C2 task environment. These characteristics separate C2 environments from other task environments. One of the hallmark characteristics of C2 environments is that they are dynamic, thus making planning an action a time-pressured activity that requires continuous monitoring of the environment. C2 thus emphasizes being able to produce actions quickly and effectively to get ahead of the situation at hand. This emphasis on quick decision-making is captured by John Boyd’s OODA (Observe-Orient-Decide-Act) loop, illustrated in Figure 1 below (Brehmer, 2005; Breton & Rousseau, 2005; Paradis, Breton, & Roy, 2014). The OODA loop describes the decision-making process taking place within C2, where action is built on the process of observing one’s environment, orienting oneself within the perceived environment, deciding on an appropriate response to the perceived environment, and then acting on that decision (Breton & Rousseau, 2005). Although a widely used template for how to achieve successful C2, the OODA loop has been criticized for not accurately capturing how C2 operates. As one of the critics of the OODA loop, Brehmer (2006) points out that the success of C2 is not only based on its ability to produce decisions quickly, but rather that decisions are made in a timely manner so they can have an appropriate effect on the situation. These abilities are argued to be the effects of C2 functions working within a system rather than a process that is to be traversed quickly (Brehmer, 2005).

Figure 1 John Boyd’s OODA loop

As a part of his critique of OODA, Brehmer (2005) developed the D-OODA (Dynamic OODA) loop as a substitute for OODA, arguing that it more closely represents the actual functionality of a C2 system (illustrated in Figure 2 below). The D-OODA loop introduces a

(13)

6 couple of changes to the original OODA loop. The biggest change is that it views C2 as a system, compared to OODA which is more akin to representing an individual process within C2 (Brehmer, 2006). There are three main functions acting within the system (illustrated below inside the box), namely information collection, sensemaking, and planning.

Figure 2 The D-OODA loop, adapted from Brehmer (2006)

The C2 system and its functions are placed inside a mission loop, which illustrates how the C2 system produces orders, which lead to action, which has an effect on the situation

(Brehmer, 2006). The box labelled friction refers to the different outside factors that may have an influence on the implementation of a plan, e.g. a sudden change of wind direction in a firefighting scenario may cause the current firefighting strategy to be less effective than originally planned. These effects are picked up by the information collection function’s sensors; in the above scenario, for example, anemometers reporting the change in windspeed, and reports from the field that the implemented actions are ineffective. This information is then integrated into the operating picture and, based on the understanding of the current situation and the overall mission statement (the sensemaking function), a new plan is formed and implemented (Brehmer, 2006). The definition of IO discussed in section 2.1, much like OODA, is focused on an individual level, but the same processes can be overloaded at a systematic level when viewed through the D-OODA loop.

(14)

7 Artman (2000) studied the impact of two C2 system structures for distributing information within information flow and how team situation awareness developed within these structures. Artman’s study provides a good example of how the systemic structure may affect the

decision-making and information management of individual operators. In his study, he used two C2 system structures for distributing information, one in which information is distributed sequentially and one in which it is distributed in parallel. The findings showed that the

different structures influenced information flow differently, and thus affected the strategies by which situation awareness was developed within the team and loaded the systems differently. The sequential distribution structure caused information to be processed and passed on

serially, which led to the controlled distribution of information since it could be attentively monitored as it traversed through the structure. The sequential distribution of information also created dependencies between each post in the command structure, meaning that as

information moved through the command hierarchy situation awareness was shared amongst the posts (Artman, 2000). For example, since post C does not receive information until it has been handled by post B, post C could monitor post B as B was handling the information, thus leading C to have greater insight into both the operational picture of B and the context in which B handled the information.

In the parallel distribution structure, however, information could be distributed along the command hierarchy more quickly, but the post dependencies and distribution of situation awareness gained from the sequential structure was lost. Since information could flow freely across the posts, staff at each level of the structure developed their own sense of the situation, thus creating a need to gather all the posts in the hierarchy together to discuss what the

situation looked like to ensure that the entire staff was working under the same operating picture (Artman, 2000). Although the parallel system was quicker at distributing information, there were two major drawbacks. The first drawback is mentioned above, with situation awareness being distributed unevenly. The second drawback, which is more directly relevant to the topic of IO, was that the parallel system suffered from IO due to an increase in

unfiltered information passing through the command hierarchy and was thus more prone to breaking down over time (Artman, 2000).

Perhaps a more intuitive way to explain the connection between a C2 system and IO is that information collection is one of the three main functions seen in the D-OODA loop. In the same way as an individual is exposed to being overloaded by increasing information processing requirements, a system can also be overloaded. This may be due to structural

(15)

8 factors, as discussed by Artman (2000) above, or due to the negative spiral discussed by Hendy et al. (2001) in section 2.1.2. Also, the efficiency of any sensemaking function in a C2 system is also dependent on the information delivered to them being sensible so as to make it easier to integrate that information into the operating picture.

The next section provides an overview of other areas closely related to C2 systems that can be inhibited by IO, thus causing a C2 system to break down.

2.2.1. Event detection

A key point of a functioning C2 system is to have the right information available at the right time, both to maintain a picture of the current tactical situation and to be able to make appropriate decisions. C2 systems thus often rely on a vast sensor networks to gather

information on the current situation to facilitate the information required to establish situation awareness. It is here, however, that IO becomes an issue. One of the challenges of working in a complex dynamic environment such as command and control is that the operational

environment is constantly changing. Depending on the mission, there may be several uncontrollable factors (frictions in the D-OODA loop, as seen in Figure 2) that are vital to identify and respond to if the overall mission is to be a success. For example, the change of wind direction in the firefighting scenario mentioned in section 2.2. A key part of achieving situation awareness is the incorporation of relevant elements into a comprehensive

understanding of their relevance to the current situation and operator goals (Endsley, 1995). However, for the relevant elements to be incorporated, they must first be detected.

Signal Detection Theory (SDT) refers to the process by which humans perceive and register stimuli in situations where the environment can be divided into either of two states: noise and signals (Wickens, Hollands, Banbury, & Parasuraman, 2013). The purpose of SDT is to explain how humans discriminate a signal from noise. If a signal is not detected, its content is missed. Depending on the environment, this can mean the loss of a vital piece of information. The detection of a signal is based on its power in comparison to the surrounding noise. If a signal is stronger than the noise, it is detected; if the signal is weaker than the noise, then it goes undetected (Wickens et al., 2013). There are also situations in which noise can be construed as the presence of a signal due to random variations in stimuli from the

environment or perhaps to heightened sensitivity of the operator. The process of detecting a signal is thus twofold; first, some stimulus is perceived regarding the potential presence of a signal, and a decision must then be made about whether the perceived stimulus is a signal or

(16)

9 not. This second part of the process gives rise to four different outcomes in SDT when a decision has been made about the presence of a signal: 1) The signal is identified correctly, meaning that there is a signal present and it is detected as such; 2) Noise is detected but it is acted upon as a signal causing a false alarm; 3) A signal is present but it is disregarded

causing a miss, and 4) No signal is present and it is correctly rejected as noise (Wickens et al., 2013). To be able to distinguish a signal from noise thus largely depends on the difference in strength between signal and noise; if the difference in strength is small, the possibility for error increases dramatically. It is thus important to avoid increases in unnecessary noise; one such noise factor could be that of IO. For example, the requirement in C2 to continuously monitor and process the information flow in order to identify and make sense of potential pieces of valuable information requires sustained attention. If that task is overloaded, not only does the task cause attentional resources to be depleted more quickly, but it also causes a higher mental workload, which in turn causes mental fatigue (Hendy et al., 2001). According to Wickens et al. (2013), both sustained attention and sustained high mental workload are connected to loss of signal sensitivity (the operator’s ability to pick up a signal), leading to more misses. For an operator in an overloaded state, it is plausible that signal sensitivity is decreased in a similar fashion, due to the operator’s attentional resources being drawn to additional tasks; for example, sorting the information flow rather than monitoring it (Kim, Yang, & Putri, 2016). In a C2 scenario, this would mean the potential loss of valuable information.

2.2.2. Contextual control

The potential loss of information due to IO is hugely detrimental to the functioning of a C2 system because, as mentioned at the beginning of this chapter, having the right information available at the right time is key to a functioning C2 system. To maintain control over a situation, it is imperative that emerging events are detected in time. If an event is detected early, the ability to plan for it and respond appropriately is increased, as a response may be planned proactively before the event has escalated. However, if an event is detected late, the C2 system is faced with a different task, where a response is reactive and action may be delayed until the emerging event has escalated into an entirely new situation (Caldwell & Garrett, 2011). One of the main benefits of having access to the entire operational picture is the ability to proactively anticipate future changes and thus obtain control, whereas reactive actions are more likely to cause a loss of control and thus impact negatively upon

(17)

10 Hollnagel (1998) proposed a similar notion with his contextual control model (COCOM), in which he argued that a system’s actions are determined based on the current context and that this context also decides the level of control with which one can plan and implement an action. COCOM as described by Hollnagel (1998) consists of four levels of control modes: Strategic – the optimal level of control – where time is abundant, and actions may be planned according to higher-level goals and their impact on the situation evaluated thoroughly;

Tactical – a controlled situation – where time is adequate and actions are planned according to guidelines and known procedures; Opportunistic – control is barely sustained – time is

limited, which causes planning to be limited and actions are largely executed based on habits and experience; Scrambled – control is lost or about to be lost – time is too inadequate for any type of planning to take place and actions are decided in a random fashion, the only goal in this control mode is to somehow regain an ounce of control. Although the strategic control mode is the most optimal, systems acting in a real environment often bounce between tactical and opportunistic (Hollnagel & Woods, 2005).

The time pressure present in each of the control modes is also directly connected to the information processing capacity. As control is lost, time becomes scarce and thus the capacity to process information is reduced in order to save time, but the information processing

requirements may also go up as the task becomes more complex due to losing control (Hollnagel & Woods, 2005). If a system is currently in the strategic mode, then that system will have a lot more time to process the information that is required. However, if a system moves into the opportunistic or even the scrambled state, then the risk of IO increases dramatically as time becomes limited. Thus, coping strategies must be implemented (either individually or systemically) to cope with the time pressure and the processing requirements (Hollnagel & Woods, 2005). One such coping strategy is omission, in which all additional information is ignored to manage the current task (e.g. retaining control in the case of the scrambled/opportunistic modes). Another coping strategy may be filtering, which relies on only accepting certain categories of information to reduce the amount; however, categories deemed not to be relevant are ignored and lost, as in the omission strategy. It is also possible to queue incoming information in the hope that it can be handled later; however, it could be said that deciding which information to queue will only add to the task and perhaps create a backlog that adds to the overload problem.

(18)

11 2.2.3. Situation Awareness

As stated by Breton and Rousseau (2005), situation awareness is a crucial aspect of being successful in any C2 activity. Endsley (1995) describes situation awareness as a three-level process. The environment is first perceived and relevant elements are identified, such as the position of friendly units, wind direction, and number of individuals hurt in an accident, etc. These distinct elements are then brought together to form comprehension of the current situation; for example, based on the above examples, the current situation of a firefighting operation. The third level of situation awareness involves the projection of future states based on the comprehension of the current situation; for example, predicting the future need to relocate firefighting units to prevent a new fire outbreak due to a change in wind direction. Endsley (2015) does not intend for this process to be linear; an operator may enter at any level, but situation awareness progresses according to these levels. If an operator has

progressed through all three levels, their situation awareness is better than that of an operator who tries to project future states but has no comprehension of the individual elements active in the situation (Endsley, 2015).

A common aspect of achieving situation awareness is the notion that the more information the operator receives, the better the situation awareness that may be achieved. However, it is the amount of information that can be transformed into knowledge that is the important measure, rather than the volume of information itself (Caldwell, 2008). Wright et al. (2015) found that greater amounts of information lead to longer decision times, which in itself is quite intuitive. But longer decision times can become detrimental to decision-making. Marusich et al.’s (2016) study, for example, showed that making more task-relevant information available to operators does not affect their task performance. There is only so much information that operators can integrate and transform into useful knowledge (Caldwell, 2008; Marusich et al., 2016). Similar findings have been presented by Giompapa et al. (2006), who noted that operators’ information processing capacity is overloaded when handling more than six information channels. These limitations may also be influenced by individual differences in the operator, of course. Caldwell (2008) refers to this individual limitation as the concept of bandwidth, which he defines as “the amount of information an individual may process per unit of time and transform into useful knowledge” (Caldwell, 2008, p. 430). A similar notion was proposed earlier by Hendy, Farell, and East (2001), who proposed a theory of humans as limited-capacity information processors, for whom information processing capacity devolves when time pressure is added and processing requirements go up.

(19)

12 IO in situation awareness is thus similar to IO with regard to signal detection. Too much information inhibits the operator’s ability to distinguish useful information from the rest and so hinders the operator from grasping the situation. It is important to distinguish between unnecessary and necessary information in the case of IO. Bennet et al. (2007) have shown that allowing operators to have access to task-relevant information regarding the complete

operation in a supply chain had a positive effect on task performance and level 3 situation awareness. Similarly, Artman (2000) suggested on the basis of his study that teams obtain greater situation awareness when allowed access to the complete operational picture within a command structure that processes information serially. Taking these findings in relation to findings mentioned in other sections, for example Marusich et al. (2016), suggest that in C2 there is a fine line between providing a complete operational picture – which supports situation awareness – and overloading the operator, which increases the time it takes to achieve good situation awareness.

2.3. Combating information overload in C2

To ensure the continued efficiency of C2 systems, the handling of information must be as great a part of the efficiency equation as information collection. Paradis et al. (2014)

emphasize the need for technological decision-making aids in C2 to be engineered according to both the technological need for that environment and the cognitive fit between that

technology and the task. In the context of IO, it is important for the technology to be fitted to the human so that it does not add to the overload problem (Karr-Wisniewski & Lu, 2010). The problem of IO has largely arisen out of the context of big data. With rapidly expanding options to gather information, it is important that the options for managing that information keep up. Thus, this section will review some of the work on the topics of gathering

information and managing it.

2.3.1. Information collection through data fusion

One of the dominant concepts in the area of information collection is data fusion. This is defined according to the JDL panel of C3 as: “A process dealing with the association, correlation, and combination of data and information from single and multiple sources to achieve refined position and identity estimates, and complete and timely assessments of situations and threats as well as their significance” (White, 1991, p. 5).

The purpose of data fusion is mainly to counter the user problems that arise from big data, and especially problems related to sensors storing data on different databases. The C2 operator

(20)

13 could thus be required to manually fuse together the data that is of use in order to gain the information necessary to make decisions. IO is one of the user problems mentioned above; another user problem arises from databases containing data that is inaccurate, unreliable, or redundant to differing extents due to their sensors having recorded the data differently or recorded the same data. Lastly, different databases may hold portions of data relating to the same information, such that the information cannot be fully understood unless all databases are searched and each piece found (Akita, 2002). Data fusion usually consists of five functional levels, numbered from 0–4, starting with the learning and production of data characteristics at level 0, then the grouping of those characteristics into objects at level 1, the integration of those objects into situation assessment at level 2, adding a threat assessment of the integrated objects in context at level 3, and lastly an iterative process of refinement regarding the information-gathering process at level 4 (Paradis et al., 2014).

Souza and Pinheiro (2013), as well as Carvalho, Souza, and Pinheiro (2015), have

investigated the method of using data fusion and correlation methods to identify single objects from heterogenous data sources. Such efforts would allow different sensors to observe several events, and through data fusion then merge those event data that are related into a single object, thus reducing the user problems of redundant and split information mentioned by Akita (2002), which could easily overload the operator (de Souza & Pinheiro, 2015; de Souza & Pinheiro, 2013). An interesting caveat can, however, be added to using data fusion to aggregate data. Speier and Price’s (1998) study indicated that using aggregated information is only effective when time is limited, and that providing decision-makers with more detailed information yields higher decision quality as long as there is time to process that information.

2.3.2. Information evaluation

The importance of providing the operator with appropriate information is something that has been mentioned several times in the previous sections. The process of evaluating the

information before presenting it to the operator is thus a large part of avoiding IO.

The ability to assess the relevance of information before presenting it to the operator is a key aspect of information evaluation. Breton, Bosse, Rousseau, and Tremblay (2012) proposed an information relevance evaluation framework called FAIR (Framework for Analysis of

Information Relevance) that uses pre-set information requirements as a way to filter relevant information from irrelevant. By using subject-matter experts and a breakdown of each critical information aspect in the context of situation awareness and COCOM, it is possible to identify

(21)

14 what information, in what quantity and form, is needed, when and where depending on the task and the time available for that task, thus allowing information to be filtered before reaching the operator (Breton et al., 2012). Although FAIR is not a technical aid, it is a good example of a theoretical framework that can be used when designing decision aids. Curts and Campbell (2001) propose using object-oriented technology that clusters massive data into coherent wholes in order to filter it. This is intended to reduce the IO caused by recurring and redundant data and allow the operator to traverse the OODA loop more quickly by being aided in the observe and orient part of the cycle (Curts & Campbell, 2001).

3.

Technical solutions

This section provides an example of a technical solution that aims to address IO problems in C2 environments, namely the SMART tool from the research group VACCINE.

3.1. VACCINE

The Visual Analytics for Command, Control, and Interoperability Environments Center (VACCINE) is a research group dedicated to the development of tools and methods for managing and analysing the vast amount of data available in mission environments (“About VACCINE”). VACCINE’s research focuses on providing tools that make users more effective through three main goals, as quoted from their website

(https://www.purdue.edu/discoverypark/vaccine/):

• Providing the right information, in the right format, within the right time to solve the problem

• Turning data deluge into a pool of relevant, actionable knowledge

• Enabling users to be more effective from planning, to detection, to response, and to recovery

VACCINE categorizes its research into five major areas (“About VACCINE”): • Investigative Analysis and Anomaly Detection

• Trend Identification and Predictive Analysis • Spatiotemporal Exploration and Visual Analytics • Risk-based Decision Making and Resource Allocation • Image/Video Analytics and Recognition

(22)

15 VACCINE is thus a prime example of a research group based around the issues covered in this paper. In the next sections, a couple of their tools and research projects will be covered.

3.1.1. SMART

In the field of Investigative Analysis and Anomaly Detection, VACCINE has developed a technological tool called Social Media Analysis and Reporting Toolkit (SMART). The purpose of SMART is to make the information available on social media feeds more accessible for real-time implementation into situation awareness (Department of Homeland Security, 2014). SMART has been implemented in a vast array of different fields and

practices, ranging from security agencies to the coastguard to police departments. It was used, for example, during the 2017 presidential inauguration to locate where protests were about to emerge (Zhang, 2017).

Zhang (2017) outlines the functionality of SMART. Its main display integrates a

topographical map overview with several information elements that are stacked on top of the map to allow the spatial clustering of relevant information. SMART also supports the toggling of non-location-based information elements that the user may choose to display or hide at will. SMART utilizes two features for displaying information on the map, called Tag Map and Contentlens. Tag Map is a static representation of keywords displayed over the areas in which they are most prevalent. Contentlens is an interactive lens that can be moved and placed around the map to only display keywords based on a certain location, that is, where the user places the lens. SMART also displays related keywords around the initial keyword through a word-cloud. This gives the user an idea of what other keywords may be of use in potential searches. The user may also interactively decide which messages are going to be displayed on the map through interactive message categorization. Either by creating their own keywords or allowing SMART’s automatic classification algorithms to find keywords, the user is allowed full control over which messages are searched for and displayed by combining different keywords of interest. The user is also provided with real-time feedback about the usage of a certain keyword through a use-counter related to each keyword, which provides the user with an idea of how relevant that keyword is. The last main function of SMART is abnormal event detection and automatic alerts. By analysing keyword trends, SMART can assess anomalous keywords and present them to the user in an ordered fashion based on their anomaly rate. The user may then choose to investigate each keyword in more detail. Further analysis of a

keyword does, however, require the participant to interact directly with the keyword of interest to analyse it further. Thus, in cases where such active monitoring is not possible,

(23)

16 SMART provides automatic alerts for when a keyword hits a certain amount of usage within a specified period of time. All alerts can be customized to suit the user’s needs.

4.

Method

An interview study was conducted to assess how IO is viewed in modern C2 environments related to emergency response.

4.1. Participants

The interview study included 13 participants from three different emergency response

agencies in Sweden who engage in C2 work. Six participants were from the police, three were officials in emergency preparedness, three were from the rescue services, and one was a press secretary at a communications unit. The main criterion for inclusion in this study was that the participants were actively working in a C2 environment. The participants were thus recruited according to recommendations through contacts at each relevant agency. The agencies that were targeted for recruitment were the police, the rescue services, and the emergency medical command in the county of Östergötland, since these three are actively engaged in daily C2 work. Participants were of mixed rank and level of command, ranging from operator to strategic commander.

4.2. Interview design

The interviews followed a semi-structured approach in which questions regarding relevant themes had been established beforehand, but deviations from the questions were encouraged if relevant. These pre-established questions acted as a framework to ensure that key topics were covered, but the order in which the questions were asked was unspecified and could vary depending on the situation; for example, two questions might be answered at the same time. The pre-established questions were divided into three general themes: 1) how the interviewee works with and views information, 2) how familiar the interviewee is with IO and 3) how IO is interpreted in the context of the interviewee’s work. Lastly, there were questions about tools and aids in C2 work. Since the interviews had an exploratory purpose, all the questions were designed to be open-ended. The questions sought to gauge how the participants

subjectively viewed the topics.

The interviews were recorded via a voice recorder and the interviewer took written notes during each interview. The interviews took around 40–60 minutes each.

(24)

17 4.3. Interview procedure

Interviews were conducted both face to face and over the phone, but the procedure remained largely the same. Before the interview started, the participants were provided with a brief written introduction to the current study and they were also asked to sign a consent form. If the interview was conducted by telephone, the above two steps were carried out verbally. Once the participant had provided their consent and felt comfortable to begin, the interviewer started the voice recording and initiated the interview.

During the interview, the pre-established questions were not asked in any pre-set order. To create a sense of natural flow, it was left up to the interviewer, in relation to what the

participant was discussing, to decide in which order the questions should be asked. However, all the interviews started with a few questions regarding the participant and their job. This was mainly done to establish a connection with the participant, but it also allowed the interviewer to get a better sense of the participant and their work role.

4.4. Apparatus

An iPhone 6 application called “Voice Memos” was used to record face-to-face interviews and a mobile application called “Call Recorder” was used to record telephone interviews. A qualitative analysis tool called Nvivo 11 was used to analyse the data.

4.5. Analysis

Thematic analysis was chosen as the analysis method due to the study’s exploratory and descriptive nature. Thematic analysis is a method based on identifying thematic patterns, or themes, in interview data. These themes may identify phenomena or describe how the data explains a phenomenon (Howitt, 2013). Themes are derived from patterns in the data, and these patterns are derived through an iterative process of coding and grouping chunks of data. There is no precise methodology for how this iterative process should be conducted, so the exact details of how to proceed are left up to the researcher in question to decide (Howitt, 2013).

There are, however, different ways to approach thematic analysis regarding how the data set is approached and what the resulting themes may say about the data sets. As this study focuses on IO and information work in general within C2, the analysis concentrated on what the participants said about this. On occasion, off-topic elements were incorporated into the analysis if they were judged to be relatable to the main topics. Thematic analysis can be differentiated into different types in that it can be either theoretical or inductive. This means

(25)

18 that either the data is viewed from the perspective of a theoretical background adhering to a certain problem area (top-down) or it is data-driven in the sense that the data is not analysed from the perspective of existing theory (bottom-up) (Braun & Clarke, 2006). This study has used a methodology that is somewhere in between the top-down and bottom-up approaches. Since the investigation is exploratory in nature, viewing the data from the bottom up is needed, but since the analysis is intended to investigate a specific theoretical area and

participants were selected due to their perceived suitability to discuss this area, the data is also suitable to analyse in a top-down manner. The compromise was thus that a top-down

approach was used to decide which sections of the data were relevant and which were not in terms of theoretical relevance, but as far as investigating the data and how it relates to the theoretical area was concerned, the approach was bottom-up. The data was analysed based on two questions: how do emergency response C2 operators in Sweden work with information, and how do they view the problem of IO?

5.

Results and analysis

The results are presented in sections intended to answer the questions that were used when analysing the data. That is: how do you work with information in Swedish emergency response C2 today, and how do Swedish response C2 workers view the problem of IO? All quotes in italics are from the data set; all quotes have been translated into English from Swedish.

5.1. Working with information in Swedish emergency response C2

The first section is concerned with themes regarding how Swedish emergency response C2 workers currently work with information.

5.1.1. You need information to understand the operational picture, and you need the operational picture to understand the information

When the participants were asked about what aspect of their work was most important and how information is used in their work, the unanimous answer was establishing an operational picture. The process of establishing situational awareness and an operational picture is one of the main aspects of C2 work because, if you do not have a full understanding of the situation, it is impossible to navigate through it. “Any information that helps develop the operational picture is good information.” The thought processes of the participants were largely centred on the idea that you need to have an operational picture to make decisions, and you need to

(26)

19 have information to establish the operational picture. This is not a new finding, there is a reason why the information collection function is placed first in Brehmer’s (2006) D-OODA loop’s C2 system.

The more unexpected aspect of the need for an operational picture, however, is the picture’s apparent function in establishing information needs. It is not only that you must have

information to understand the operational landscape, but you must also understand the operational landscape in order to understand the information landscape.

“You need to assess the operational picture and the situation you’re in, and through that assessment you understand what information you need in order to understand the current situation, but also what information you need to understand how this situation will develop.” A part of this attitude also stems from the opinion that information should not just be

received, it must also be actively gathered in C2. It is therefore imperative in C2 operations that operators understand what information needs exist, and the only way to establish this, according to the participants, is to understand the current situation.

As mentioned in the quote above, understanding the information needs of a situation is not only essential for the C2 system’s current understanding of a situation, but it is also essential for a C2 system’s ability to proactively work with a dynamic situation. Participants placed some emphasis on being able to assess not only what information is needed now, but also what information will be needed in the future, as well as what information those working outside the C2 system will need from the system. The point was that, if you are to

successfully operate in a dynamic environment, it just as important to guard the future interests of the operation as the current moment. The reason for this is also why C2 often consists of levels of command working on different operational timescales. This is relevant here because it also ties into the relationship between operational picture and information needs. If a C2 system is to retain control of an operation from an information perspective, it is important not to be taken by surprise by sudden shifts in information needs, since this may cause overload in either workload or amount of information. However, in a dynamic environment, there are sure to be unexpected events causing new needs. The impact of suddenly fluctuating information needs may be remedied providing that a C2 system has control.

A final note on this theme comes from the perspective of cooperative C2 work, where several agencies with their own C2 networks cooperate during an operation. Here, the question

(27)

20 becomes how to create a shared operational picture between the cooperating C2 systems which, according to the participants, is essential for success. The reason why this is important seems to lie in the framework that the operational picture provides for understanding

information.

“If you don’t work under the same operational picture as the other agencies, you might not be able to take appropriate actions or, even worse, you might take actions that are

counterproductive to the goal of the operation.”

The key phrase in the above quote is “the goal of the operation”. A shared operational picture provides not only a framework in which everyone has the same understanding of the situation but also a framework for assessing how close you are to achieving the goal of the operation, and what actions can be taken to achieve that goal. This is a piece of intelligence of

paramount importance, which, according to certain participants, is hard to share properly in today’s emergency response C2. This is because it is often only the current or even the past operational picture that is shared between agencies. The direction that an operation is intended to take is often left out, according to some participants. The reason this topic is brought up under this theme is that it seems to be related to the operational picture’s role in guiding information work in C2. If there is no shared operational picture, shared information will not be managed correctly since everyone will be working under different assumptions based on their version of the operational picture.

Thus, understanding the operational picture is an essential piece of the puzzle in a C2

operator’s ability to gain a complete picture of a situation, in terms of both the operation itself and the relevant information.

5.1.2. The managing of information is a conscious and manual effort

This theme concerns how information is managed in Swedish emergency response C2

systems, and stems from questions probing how participants manage the gathered information in order to turn it into actionable intelligence.

As the theme’s title states, the assessment, filtering, and integration of information in today’s Swedish emergency response C2 systems is a largely manual process. Decision-makers at seemingly all levels of command are left to assess and decide upon which information to use and which to pass up at their own discretion. This is probably why the word “conscious assessment” was often brought up in the interviews, especially when discussing the topic of possibly integrating digital decision-aiding tools into the current C2 systems. Whilst no

(28)

21 participant directly opposed the use of digital tools, often claiming they were

pro-digitalization, they were unanimously cautious on the topic. This caution stemmed from a worry that digital tools would not provide enough insight into their reasoning so that the logic behind a decision could be followed by a human operator, thus disconnecting the human from a part of the decision-making process. For example, there are already decision aids and digital tools being used in Sweden’s emergency response C2 systems, but these were said to be bad at retaining this connection between the decision-maker and the decision-making process. The consequence of this, according to some participants, was that they simply would not use the systems in question, since they felt that they robbed them of oversight and control over the situation.

“I’m cautious since, well, the product of such a system is still built on raw data, and to understand the product you must understand the raw data, you assess the product through an understanding of possible errors sources in the data and so on. If I can’t do this, I don’t know if I can trust it.”

The word “conscious” is highly appropriate when explaining how Swedish emergency response C2 operators work with information; they manually process it, reason about it, and are thus conscious of, and very much in tune with, every step of the decision-making process. An apparent strength of this conscious connection with the decision-making process is that the C2 operators are allowed to make experience-based assessments, which can greatly influence the outcome of a decision. Since a decision-maker with experience can assess what they want the outcome of the decision to be much more specifically than a novice, they can make more delicately nuanced decisions. Although a novice could come up with the same decision in essence (e.g. disseminating a piece of information), the impact of that decision may vary greatly due to the two operators’ different experience of the work context. One example illustrating this point was mentioned by one participant who recalled a situation in which they had had to decide how to disseminate a certain piece of information. The operator had

received a report that there had been an accident at a hospital which would cause disruptions in the hospital’s capacity to treat certain patients. The task faced by the operator was thus to divert the ambulance traffic away from this hospital so that patients would not be needlessly shuffled around if it turned out that they could not be treated at the hospital in question. Standard procedure for such events is to inform all the involved parties, in this case all the ambulances in the region. The operator, however, being quite experienced in this role, knew that this information would not be important for all ambulances in the region, but only for

(29)

22 those active in and around the area of the affected hospital. Thus, the operator made the

decision to individually contact each ambulance that would be affected by this accident, rather than to distribute the information to all ambulances in the entire region.

In essence, this was only a question of who needed to know this piece of information but, based on experience, the operator understood that this particular situation did not require any large-scale action, reasoning that large-scale action (distributing local information to the whole region) could have unintended effects and cause unnecessary load to unaffected parties. By taking this into account, the operator could make a more tactical decision regarding how the information should be distributed. The operator thus consciously assessed the situation and, based on experience and understanding of the situation, could make a more nuanced decision than perhaps a novice would have made.

The example given in the above section provides a good segue into the next aspect of this theme. A direct consequence of this emphasis on conscious decision-making and largely working manually is that experience becomes a pivotal factor, as illustrated in the above example. Experience was constantly brought up by participants as being one of the main factors for being able to handle C2 work well. Almost always, when asked about how the participants went about manually assessing, integrating, and filtering information, they would answer that it is learned over time with experience. This answer was the same from tactical commanders, strategic commanders, all the way down to ground-level operators. It was thus hard to gauge the specifics of how this manual decision-making process was engaged in by operators, and exactly what strategies were used to be able to consciously manage information and make decisions. It also seemed entirely natural to the participants that this should be the case, since the complexity of their working environment and the decisions they were making could only be learnt through experience and exposure to operative work.

5.1.3. Information is key, but not all information is useful

As described in the previous chapters, e.g. in the D-OODA model in section 2.3, all C2 environments are information-centric. This is a notion that is also unanimously agreed upon by all the interviewed C2 workers in this study. As one interviewee put it: “Having no

information leads to bad decisions, which leads to a bad mission result.” However, it was also clear in the data that not all information that could be useful, is useful.

In the case of all three agencies that were interviewed, information is often handled and assessed manually and thus it is not strange that it must retain certain characteristics to be

(30)

23 accessible for manual processing. Two primary characteristics were identified in the data: information structure and information utility. Each of these characteristics was identified by explicit comments regarding what the participant thought characterized “good” information and was also supported by implicit comments when answering other questions. It should also be noted that these two characteristics are partially dependent on each other, and not entirely distinct, but they will be presented separately in this section for the sake of clarity.

Structure

When asked if it is possible to have too much information, one participant said: “No you can never have too much information, but if the information is presented in the wrong way it may seem like there’s too much.” This quote captures the essence of why the structure

characteristic is important. This characteristic concerns the way in which information is presented to the C2 operator, either the format in which it is presented (e.g. written, vocal etc.) or how it is presented within a given format (e.g. a written report’s structure).

It was a recurring theme in the interviews that, when asked about social media as an

information source, participants were hesitant about saying that they thought it could hold any informational value. This is not because there is no important information to be gathered on social media, but because that information is not structured in a way that makes it possible to assess. Or rather, participants thought that the time and effort it would take to evaluate and process such information is simply not realistic. Since there is an element of speed required in C2, structured information becomes important because it allows for quick assessment. The more structured the information is, the easier it is to incorporate into the operational picture and turn into actionable intelligence. Unstructured information is thus often viewed as cumbersome and, although it might hold some informational value, it is simply not worth the effort of finding it. It should be made clear, however, that this reluctance is not due to

laziness, but because the analysis of such information would tie up resources that can be more efficiently used elsewhere.

The idea that unstructured information is too costly to process is supported in section 2.2.2 when discussing information-processing requirements and the information-processing model developed by Hendy et al. (2001). As task load increases and processing rate is not amended, decision time increases, which in turn increases time pressure on the following tasks. Thus, information that is structured carries lower processing requirements and lower decision times, which makes it more efficient to use in a C2 environment.

(31)

24 Utility

The term assessment was brought up frequently in almost all the interviews when participants were accounting for how they worked with information in their respective C2 roles; for example, this quote: “If you can’t assess the information then it’s useless, you simply won’t know what to do with it.” One of the reasons that information must be assessed before it is used is to decide its utility in the current situation. Several commanders mentioned that, although they may face a stream of information, they do not feel overloade, because they can simply focus on those pieces of information that bear the greatest utility for the current situation. For example: “I know what information I want and need so when they start talking about information I don’t need, I just zone out.” As the situation evolves, so do the

information needs, and operators must thus continuously assess whether or not information holds value in the current situation.

What makes utility particularly interesting as an information characteristic is that, compared to the structure characteristic, utility cannot be wrong or bad per se. For example, participants had no trouble arguing that information could be structured poorly; for example: “when information is too wordy and hides the points of interest” or “uses terminology that is not universal”. It is thus possible to view certain information types and presentation formats and objectively assess whether their structure is good or bad. Meanwhile, it is somewhat less clear what constitutes poor utility other than that information is not useful to the current situation. The interview data supports this; for example: “it’s not that the information is bad, it’s just not what I need at that moment,” or “I don’t want to, and I don’t have the time to handle information that isn’t relevant right now.” The characteristic of utility thus becomes more of a concern about information load rather than a direct problem with the information itself (as with structure). Although it is rarely admitted in the interviews that operators have

experienced IO, it is clear, for example in the two quotes above, that C2 operators experience moments when they are bombarded with information they cannot use. Operators do seem confident, however, in their ability to filter information efficiently, and few expressed concerns about feeling overloaded with unnecessary information.

One of the primary aspects of information that operators use to assess utility is whether it fits into the operational timescale of that particular operator. One internal commander said: “For example, if you follow the timescale, you know you need certain information at the beginning, and some other information in the future, and so on. So, you know what info to prioritize and what you need to put aside.” If operators work more with minute-to-minute decisions, e.g.

(32)

25 operational command and external command, a similar rule applies but regarding the

freshness of information in relation to the operational timescale. Information is filtered according to whether it is fresh, e.g. “I must assess how information stands in relation to the events on the timescale, is the new information I received before or after?” explained one external commander. The second aspect that operators assess is how the information adds to and expands the operational picture. As in the situations mentioned above, if information provides new actionable intelligence, it is considered to hold high utility and is thus retained. However, information that is retrospective, explaining what has happened rather than what is happening, seems to hold less value in the eyes of the operators.

5.2. How do Swedish emergency response C2 workers view the problem of information overload?

This section addresses themes regarding how the phenomenon of IO is viewed by participants.

5.2.1. IO caused by information load is not a problem

IO according to the definition used in this study is centred on the information load being too heavy for a human operator to handle. That is, when information load is increased, the information processing requirements are increased to the point where they surpass the information processing capacities of an operator, and thus cause IO. This is also a definition that is echoed in most of the IO literature. It was therefore interesting to find that the operators had a hard time relating to this definition. The idea of having too much information in terms of amount and load seemed utterly foreign to most participants, and when presented with the idea of such a scenario, most found it hard to believe that they could ever be placed in such a situation. Even when presented with futuristic scenarios in which C2 work involved vast sensor networks and had access to numerous information sources, the participants could hardly imagine that they would ever be troubled by the amount of information per se. Also, when presented with the idea that, in the future there may be an increased number of information sources incorporated into C2, the participants did not worry about the implications this would have for information load but rather what it implied for human resources. It is possible that the definition used here for IO is hard to relate to due to C2 operators often having to work from an information deficit. It is often such a deficit that a C2 system is constantly battling during the active operation and that is continuously growing as a situation evolves. Some participants did, however, acknowledge the possibility that you could be overwhelmed by information, but attributed this mainly to a lack of experience on the

References

Related documents

The increasing availability of data and attention to services has increased the understanding of the contribution of services to innovation and productivity in

The government formally announced on April 28 that it will seek a 15 percent across-the- board reduction in summer power consumption, a step back from its initial plan to seek a

Det finns många initiativ och aktiviteter för att främja och stärka internationellt samarbete bland forskare och studenter, de flesta på initiativ av och med budget från departementet

Den här utvecklingen, att både Kina och Indien satsar för att öka antalet kliniska pröv- ningar kan potentiellt sett bidra till att minska antalet kliniska prövningar i Sverige.. Men

Av 2012 års danska handlingsplan för Indien framgår att det finns en ambition att även ingå ett samförståndsavtal avseende högre utbildning vilket skulle främja utbildnings-,

Semi-structured interviews were held with information security professionals in order to further understand the connection between the CIA-triad concepts and trust

critical infrastructure protection; public strategic management; multi-level planning; complex planning environment; decision process; emergency power supply.. © 2018

Floridi’s main point seems to be that the infosphere is growing and evolving into a new, digital state, where everything is informational not merely because of our individual