CONTENTS
3.4 Value of Information in Decision-Making
3.7 Nature and Processes of Sense-Making
3.9 Self-Organization and Self-Synchronization in the Value Chain
3.10 Complexity in Sense-Making and Command and Control
The history of network-centric concepts in the United States can be said to go back at least to the mid-1980s when the U.S. Defense Department was reorganized under the Goldwater–Nichols Act of 1986 that imputed the notions of “jointness” onto U.S. defense and military operations. Ten years later U.S. Admiral William Owens, in a paper for the Institute of National Strategic Studies at the National Defense University, wrote on the concept of “The Emerging U.S. System of Systems” (Owens 1995) as the foundation of the “Revolution in Military Affairs,” involving the extensive use of (and dependency on) information in a layered system framework connecting various military operational functions. A sequence of publications evolved that introduced the notions of net-centricity and eventually the military notion of network-centric warfare (NCW), in which the strong informational dependency persisted. In the networked case, which allows (or should allow) extensive sharing of information, the argument was that NCW enabled the following operational advantages:
• A robustly networked force improves information sharing.
• Information sharing enhances the quality of information and shared situational awareness.
• Shared situational awareness enables collaboration and self-synchronization and enhances sustainability and speed of command.
• These, in turn, dramatically increase mission effectiveness.
In these arguments, combat power is seen to be dependent on information. Related to these ideas, Evans and Wurster (2000) introduced the concepts of information richness and seek to explain how the Internet has changed the economics of information reach and the ability of information to create value. In this work, they defined information richness as an aggregate measure of the quality of information and information reach as an aggregate measure of the degree that information is shared. Alberts et al. (2001) add the parameter of “quality of interaction” to these factors as influencing the ability to create value, in this case combat value. So it can be argued, following these developments, that combat power and mission effectiveness depend on information quality, information “share-ability,” and the nature of interaction among people using information. In a network environment, every node has an opportunity to create information but also to modify it (say, improve its quality), send it forward to other nodes (expedite the sharing of that information), and the people at that node can interact with the information in a way that exploits it for task purposes. Thus, there is the potential for a “chain” of effects that impacts the overall combat value in such a system of systems; i.e., a “value chain” is a latent construct in any information network.
The term “value chain” is cited in the various open works on NCW or network-enabled capability (Alberts et al. 2001), but other sources suggest the term was coined by Michael Porter in 1985 (Porter 1985). The concept is an abstraction related to business processes that operate on a product as part of the product development, and the notion that each process should add value to the product. It seems to be a concept primarily useful for strategic planning that exposes the cost and value drivers at each stage of product development as a basis for analyzing and discerning the best tradeoff choices to make toward optimization of value and minimization of cost.
The term has been extended by the business community to apply to broadly based, multi-organizational processes under the phrase “value network,” which seems to be particularly applicable to service industries and processes involving nontangible components and products. It is generally presumed in the discussions about value networks that there is a dedicated and altruistic intent among the collaborators to fully cooperate through synchronized interactions toward the single purpose of product value optimization. Clearly, inter-agent communication is crucial to realizing the benefits of a value network (as argued in Alberts et al. [2001], where the “quality of interaction” parameter is introduced, as previously noted), and the overall system can be and usually is complex and exhibiting a variety of inter-agent dependencies, not unlike the complexities in a social network.
The value chain in the NCW case is descriptive of the interdependencies among, and value contributions of, the links from network-centric organizations and improved (value-adding) information processes—and information products—to more effective mission outcomes. As will be discussed later, there are two core assertions that underlie this concept: (1) that the collaborative framework that the net infrastructure provides will improve the quality of organic (individual-node) information, and (2) that the same net infrastructure will provide for improved shareability of information, in turn leading to more creative, agile, and timely situation assessments and decision making. As noted earlier for the business case, here too there is an assumption of an altruistic imperative and that the network nodes are cooperatively working toward a common goal. This is not unreasonable as an ideal goal but its realization is likely to depend on the specifics of given mission applications and the usual effects of the “fog of war,” and mission risks and urgencies in the defense or military context. Even among friendly forces, it is not always the case that the entire force is pulling in the same direction due to localized and random factors.
Also, no small part of the realization of the potential of NCW and the promise of the value chain process will be the willingness of the military to commit to the underlying open, cooperative, and proactive degree of information-sharing that these concepts depend on. As pointed out by Alberts and Hayes (2003), it was not too long ago that the phrase “Knowledge is Power” was employed to convey the notion that possession and control of information (i.e., making it scarce and not sharing it) was a means to achieve power and control. This paradigm thus argues for the control and caching of information, rather than sharing it and generally making it available. In part, these contrasting views relate to the economics of information availability in the general sense as well as the cost of sharing it. With the emergence of the web and the dramatic reductions in the availability-costs of extensive amounts of information and in the marked reductions of all types of networking costs comes the push for a new paradigm that factors sharing into the value-adding processes rather than purposefully resisting it. Of course, this will require a degree of revolution in the way “information-age forces” are structured and in the way they interoperate and in particular how they share information. Military organizations will need to go well beyond the current centralized planning-decentralized execution paradigm to the structures discussed in Alberts and Hayes (2003) to realize much more organizational agility and to empower those at the edge of organizations to decide about information sharing and action-taking.
Determination of whether the asserted benefits of the tenets of NCW and in particular those of the value chain can be realized begins with understanding the degree to which a force is in fact networked or connected. As is well known, connectivity at the information level is the result of a multilayered process; it begins with the physical connection layer (wires, fiber, transmitters/receivers) but goes well beyond that layer and in the military environment of course involves multilayer security issues and accessibility controls to information. We note the important requirement that to exploit and fuse shared information one must have to have been sent it from somewhere in the network, which in turn depends on what we call “information-sharing strategies (ISS),” those protocols or policies that define who sends what to whom, how often, and in what format. And as has been mentioned earlier, effective and efficient collaboration also presumes the unified focus and altruistic intentions of those nodes in a network that can contribute to the improved problem solution actually doing so, even under combat duress and confusion.
The NCW literature has various diagrammatic representations of the value chain; here we use a simple construct in Figure 3.1 depicting the process and its important components and functions, showing how value is built up in the course of “good” network operations. The figure shows that the first requirement to enable NCW is connectivity via some type of network infrastructure. Shared observational data, data fusion, and information management, done well, lead to significantly improved situational awareness, which when properly shared and integrated into a (possibly-new paradigm of) command and control (C2) and decision-making environment have the potential to yield measurable improvements in mission effectiveness. Closely related to the concept of the value chain is the “conceptual framework” of NCW, depicted here again using the diagram from the Network Centric Operations Conceptual Framework report prepared by Evidence Based Research, Inc. (2003) as Figure 3.2.
Most of Figure 3.2 is, first of all, all about information and its flow in the network but it is (toward the bottom) also about the use of the information in decision-making and action-taking. Important themes in this framework revolve around a few special words and the implied functions: quality—sharing—degree—synchronization. Also a new term appears: “sense-making.” Notice also that many of these terms and the associated functions happen to “a degree,” and ideally should be measurable through the development of relevant metrics; more is said on this in Chapter 17. Finally, not shown here but important to note in any case is that certain functions are in certain domains, across the physical, informational, cognitive, and social categories; some involve more than one domain. Everything begins here with improvements in the quality of information at some node; nothing of quality can happen across nodes if the individual nodes have nothing to offer.
FIGURE 3.1 Network-centric value chain concept diagram.
FIGURE 3.2 NCW conceptual framework. (From Evidence Based Research, Inc., Network Centric Operations Conceptual Framework Version 1.0. 2003, Report prepared for Office of Force Transformation, November 2003.)
3.4 VALUE OF INFORMATION IN DECISION-MAKING
One way to measure the quality of information at a node is by its contribution to both the local and team or network-level decision-making and action-taking that results from employing that information. Usually, the outcomes of actions taken in the context of estimated situational states may be assigned values or utilities, which represent the relative desirability of outcomes. This type of approach is typical for cases where rational decision-making and choice-making is appropriate. However there are many modern-day problems, e.g., asymmetric problems, that do not lend themselves to the rational choice, rational decision-making paradigm. If we denote a possible situation state as s, as an instance of S, and utility of action a, given s as U(a,s), we can describe the expected utility as
(3.1) |
where P(s) is the probability of any situation s. The situation, s, however is typically unobservable in a direct sense and can be treated here as being estimated by a fusion process on the basis of observable evidence e (of all possible observational data or other evidence, E). That is, the fusion process, assuming that it has a multi-hypothesis capability, produces the distribution of estimated situations P (s|e). If the maximum utility is that associated with taking the optimal action, then we have the maximum of the expectation as follows:
(3.2) |
If we want to gauge the value of any observable evidence or information e, assuming that what is being shared in the network is observational data or measurements, then we can marginalize over the possible values of e as
(3.3) |
The value of any observable informational element can then be computed as the difference in maximum utility when the information is included in the above vice excluding it. A similar calculation could be done if what are shared are situational estimates by using slight variations of these equations, using the marginal value of any situational estimate s. The viability and ability to implement calculations of this type will vary from case to case, but some type of quality measures are needed to drive the value-chain process; as footnoted previously, the Network Centric Operations Conceptual Framework report (Evidence Based Research, Inc. 2003) has a rather thorough characterization of a holistic approach to measuring the various “ilities” associated with the value chain process.
It is important here to make a “fusion” remark in light of the implications of Figure 3.2. Any fusion node can only fuse two things: that information which is available to it organically—i.e., information over which it has control, such as locally managed sensor devices—and that information which comes to it somehow from the network. Notice the emphasis on “somehow”; it is only through the aforementioned ISS that some type of information flows to a node from the network. Such flows can be the result of a multiplicity of interwoven ISSs such as broadcasts from some nodes, responses to service requests from other nodes, and flows from nodes that the receiving node subscribes to, or yet other flow patterns driven by specified protocols. But it is emphasized that the nature of “non-organic” fusion that can happen at a node is only the result of the synthesis of any such directed or requested (and responsive) information flows, which in turn are the result of defined protocols and policies. A related remark is that fusion can be (should be, if well designed) a contributor to the quality of information and quality of sense-making and understanding, both at the individual or nodal level as well as at the shared level. It could be also argued that the “Level 4: Process Refinement” function of the fusion process could contribute to the nature of the information sharing and other inter-nodal interactions in a positive way, depending on the control authority aspects of how the network is managed.
Further, fusion process design is often spoken of as impacted by “push” requirements—those requirements driven by the input-side, and “pull” requirements, driven by the user-side. The network environment influences both of these requirements-sets in possibly many ways. It can be said that the information flow in the network can be characterized as both delayed and out-of-observation-time-order, and probably Poisson in arrival-rate distribution, all of which could potentially affect fusion algorithm and process operations. New user patterns involving self-organizing and self-synchronizing organizational dynamics will also likely affect how fused information products should optimally be constructed and delivered for use.
Following the flow of Figure 3.2, “sense-making” is a process and desired capability at both the individual node level and at the network level. It can be individualized to a person in which case the process would be largely cognitive with some degree of automated support at the individual level. For any netted level of sense-making capability whether within a sub-network at a node or across nodes, the sense-making process relies largely on patterns of collaboration and information exchange. As might be expected, the sense-making term seems to have a number of nominated definitions; a few are offered here to give a sampling:
• Sense-making as making sense of uncertainties in environments through interaction (Weick 1969).
• Sense-making encompasses the range of cognitive activities undertaken by individuals, teams, organizations, and indeed societies to develop awareness and understanding and to relate this understanding to a feasible action space (Alberts 2002).
• Sense-making is defined as the process of creating situation awareness in situations of uncertainty (Leedom 2001).
• Sense-making consists of a set of activities or processes in the cognitive and social domains that begins on the edge of the information domain with the perception of available information and ends prior to taking action(s) that are meant to create effects in any or all of the domains (Alberts and Hayes 2006).
One common theme through the definitions seems to be the notion of dealing with and clarifying an estimated world view while dealing with uncertainty, anomalies, and contradictions. Sieck et al. (2007) depict individualized sense-making as a six-step frame-building process (frames associated to mental representations in this approach), involving sub-processes that seek a frame, and elaborate, question, compare, reframe, and preserve the frame in an iterative process. Each step involves some type of adjudication or reconciliation process to deal with classes of complexity or uncertainty and ambiguity. In this process then, the drive to reduce uncertainty may not be immediately helpful since part of the sense-making process can be to understand the implications of uncertainty and ambiguity. The problem spaces addressed by sense-making processes involve an incomplete understanding of reality and are thus ontologically incomplete; they are also epistemologically incomplete in that available knowledge models are not adequate to describe the observed phenomena. Table 3.1 from Zack (1999) offers a characterization of types of ignorance that sense-making must deal with.
As pointed out in the sense-making literature (McCaskey 1982), the sense-making process is not constrained by the usual models and assumptions of rational decision-making, and a generalized maximization of a type of a utility-type function on the part of the decision-maker. Modern-day adversaries can be expected to act “irrationally” at least by certain standards, and certain arguments suggest that friendly decision-makers need to be equally “irrational” in their decision-making processes. Uncertainty reduction and optimization methods work well in support of the rational choice/rational decision-making model but may warrant reexamination as part of a sense-making process involving a collaborative situation assessment process that is constructing a subjective view of an unknowable, dynamic world and largely dealing with overt deception, equivocal information and the reconciliation of alternative views among the networked decision-making team. The use of bounded rationality models helps in this regard but such models are not the same as the typical descriptions of sense-making. In the sense-making case, it could be said that the networked group is constructing an interpretation of some complex reality sufficient to achieve a state of commitment to that interpretation and the decisions and actions that may result from it. This notion interacts with the concept of self-organizing teams in that the sense-making process is a logical precursor to a team setting its own goals and objectives for both action-taking and information-seeking. It could be said that a team can only be labeled as self-organizing if it dynamically sets its own goals and objectives. Commanders then need to limit themselves to presenting the team with an ambiguous challenge rather than defining terms of reference, etc.; whether traditional militaries can adapt to this process is to be seen. Moreover, most fusion processes operate on what could be called explicit information and to varying degrees may not exploit tacit knowledge and contextual information.
McCaskey (1982) offers the list shown in Table 3.2 of types of problems and questions that sense-making type processes are intended to address. It could be said that these are problems involving degrees of bewilderment for analysts or decision-makers. The term “wicked” has also been used to typify such problems involving contradicting information, discrepancies, etc., and the need for problem-solvers to significantly change their mindsets and shed historical preconceptions; see Rittel and Webber (1973).
3.7 NATURE AND PROCESSES OF SENSE-MAKING
Sense-making is sometimes labeled as “constructive reality” and a process that is action-centered and retrospective. This is similar to what some in the fusion community have called “stimulative intelligence,” which involves taking actions to stimulate an adversary to an action that is either observable or that aids in clarifying a hypothesis. Such strategies will generally be more successful at the physical level, e.g., when trying to cause actions that manipulate physical objects, but both harder to define and execute and likely less successful at the informational and cognitive levels which are both fundamentally more difficult to manipulate and to observe. The sense-making processes are emergent and adaptive but are trying to be kept within a linear inferencing framework. It is also characterized by the problem-solvers’ reluctance to simplify interpretations and a reluctance to dispense with information that doesn’t fit nominated hypotheses; these teams are also characterized by having a commitment to resilience. With the process involving frequent adaptation, it can also be appreciated that most characterizations of sense-making describe the need for a knowledge management function that keeps track of the dynamics in nominated hypotheses and associated knowledge models to prevent thrashing and a failure to converge. Leedom (2004) shows the diagram of Figure 3.3 to convey the hybrid combination of linear and emergent processes working together in a mission/operational-tempo-based temporal context.
Form of Ignorance |
Definition |
Corrective Response |
Uncertainty |
Uncertainty is defined as not having sufficient information to describe a current state or to forecast future states, preferred outcomes, or the actions needed to achieve them. Uncertainty can be defined in degrees (e.g., in terms of probability); however, the context of uncertainty is well-defined and meaningful to decision-makers. |
Uncertainty can be reduced be acquiring additional information relevant to the problem context. Uncertainty can be tolerated by using assumptions to fill in missing information, or by developing agile responses that can accommodate critical areas of uncertainty. |
Complexity |
Complexity is defined as being faced with a situation made of an inter-related set of variables, solutions, and stakeholders – each individually understood, but together with exceed the processing capacity of the individual, the team or organization to synthesize. Complexity is defined relative to available experience and expertise: what is complex for one individual might be easily understood by another. |
Complexity can be accommodated by breaking problems down into manageable pieces (division of labor). However, this requires the addition of management overhead and the means to bring together the appropriate experts to synthesize the various pieces back into an integrated whole. |
Ambiguity |
Ambiguity is defined as the inability to make sense out of a situation, regardless of available information. Ambiguity arises when faced with novelty or situations that do not correspond to past experience. Here, what is lacking is not information but the experience and expertise to correctly frame and interpret the information. |
Ambiguity can be resolved by acquiring new sources of expertise and/or allowing iterative cycles of collaboration among experts and stakeholders to create new interpretations of the situation. Such collaboration requires well-established social networks for success. |
Equivocality |
Equivocality is defined as having multiple–equally plausible-interpretations of the same information. Here, interpretations may differ along one or more dimensions; descriptive criteria, problem boundary, relevance of specific underlying factors, multiple stakeholders who each have a vested interest in characterizing the current situation, forecasting its implications, and developing response actions. |
As with ambiguity, equivocality can be resolved through iterative cycles of interpretation, discussion, and negotiation among experts and stakeholders. This process can occur either democratically or in authoritative fashion, depending upon the relative influence of each stakeholder and the presence/absence of an overall decision authority. |
Source: Zack, M. H., Knowledge Directions, 1, 36, 1999.
TABLE 3.2
Sense-Making Problem Characteristics
Category |
Characteristics |
Nature of the problem |
The nature of the problem has shifted from the known (e.g., simple problem) to the unknown (e.g., wicked problem) Overall guidance and directions received from functional experts and stakeholders does not set forth a clear and consistent set of goals that address the present operational situation Time and other resource constraints necessitate trade-offs among competing goals and operational requirements |
Nature of the information |
The ability to effectively collect, interpret, and organize information becomes problematic because of the volume of available information or the reliability of this information There exist multiple, conflicting interpretations of the available information as different experts or stakeholders each apply their unique perspectives and expertise The operational situation appears to present decision-makers with a seemingly inconsistent pattern of features, relationships, or demands Functional experts and stakeholders employ symbols and metaphors to articulate their perspective, but these symbols and metaphors are not consistently understood by others |
Nature of the decision-makers and stakeholders |
Functional experts and stakeholders differ in terms of the underlying values, political goals, or emotional reactions Various relevant players lack a clear and consistent assignment of roles and responsibilities Decision-makers lack a clear and consistent set of success measures for judging operational progress and adjusting future decisions and actions Key decision-makers, functional experts, and stakeholders change as a function of the evolving operational situation |
Source: Adapted from McCaskey, M.B., The Executive Challenge: Managing Change and Ambiguity, Pitman Publishers, Marshfield, MA, 1982.
FIGURE 3.3 Sense-making dynamics. (From Leedom, D.K., The analytic representation of sensemaking and knowledge management within a military C2 organization, Air Force Research Laboratory Human Effectiveness Directorate Report AFRL-HE-WP-TR-2004-0083, 2004.)
Weick (1995) depicts the sense-making process as having four functional components as shown in Table 3.3.
Positional arguing involves disparate functional experts coming together in a “community of interest” to develop a shared understanding of the problem space and to nominate actions that will aid in confirming current hypotheses or in aiding the inferencing process. Plausible expectations from the decided actions are formed by the key leaders of the team in the form of projected outcomes or events. Behavioral commitment, as indicated earlier, is action-based and is in a sense a way to help focus the sense-making process on particular components of the problem space for which a leader is committed to a course of action (reflects “commander’s intent”). Environmental manipulation is about those actions that are taken to help develop the “constructed reality” that forms the framework of interpretation of the group.
Understanding sense-making and the role for computer-based information fusion processing requires in part an understanding of the various types of information and knowledge involved with sense-making. In Leedom (2004), the knowledge sources described are codified information and knowledge, tacit knowledge, and social knowledge. Clearly, the knowledge coming from the output of an information fusion process falls into the codified knowledge domain. Information sources that are employed by a fusion process will mostly fall into the codified information domain. Among such sources, it can be argued that one particular important information source in this paradigm is that of contextual information. It has been said that “Sense-making is about contextual reality. It is built out of vague questions, muddy answers, and negotiated agreements that attempt to reduce confusion” (van Laere et al. 2007). Context is also a slippery word and has varying interpretations; it can be difficult to distinguish it from “situation” and tricky to discuss the interplay of situation and context. Contextual information, necessary to the determination of a context, can be seen to have two roles: (1) an “a priori” role where it is proactively designed into some fusion-based estimation algorithm—in this case the algorithm designed is able to prespecify what contextual information is relevant to the estimation process, and integrate it into the algorithm design (using terrain information in ground target tracking is one example), and (2) an “a posteriori” role, where contextual information is drawn upon to clarify or constrain an estimate that has been separately developed, i.e., contextual information is used after the fact of an externally asserted inference for the purpose of improved interpretation. In the latter case a type of “relevance filter” has to be designed to select, retrieve, and employ the pertinent contextual information for clarification purposes. The employment of contextual information, which can be relatively static but also dynamic (weather, e.g.), in the sense-making process adds a layer of complexity and also opens the process to various biasing effects.
TABLE 3.3
Sense-Making Process Characterization
Sense-Making Process |
What This Process Entails |
Why This Is an Essential Component of Sense-Making |
Positional arguing (belief-based) |
Various functional experts and/or stakeholders within the team or organization present their perspectives or positions in an attempt to shape the constructed problem framework As part of this collaborative process, each individual attempts to change or expand the knowledge state of others until there exists a commonly shared understanding of how each of the relevant problem elements and potential solution paths fit together in a cohesive whole Sometimes referred to as debative cooperation |
Whenever teams or organizations face wicked problems, the major challenge is constructing an appropriate problem framework within which to shape the resulting decisions Wicked problems—including their relevant threats and opportunities—will often be viewed differently by each expert or stakeholders, dependent upon their roles and tacit knowledge |
Plausible expectation (belief-based) |
Key leaders express their expectation of certain outcomes, events, or future states in order to focus the attention and thinking of their supporting team or organizational members |
The efficiency of sense-making within a team or organization depends upon its leaders focusing the attention and thinking of its members |
Expectations link belief to action in as much as constructed futures implicitly require certain actions or accomplishments that must be planned and executed by the team or organization |
Part of the responsibilities of a leader are to construct a vision for the team or organization out of many possible futures |
|
Expectations reflect constructed futures that evolve over time to conform with unfolding events and states |
Linking thoughts, teams, and accomplishment is a powerful motivational mechanism for shaping the decision behaviors of others |
|
Behavioral commitment (action-based) |
Key leaders demonstrate explicit, public, irrevocable commitment to specific plans and actions in order to further shape and focus the attention and thinking of their supporting team and organizational members |
Individuals, teams, and organizations try hardest to build meaning and understanding around those actions to which they are committed to |
Commitment is expressed in the form of approved plans and orders issued to subordinate elements |
Prior to leaders expressing commitment, all types of perceptions, experiences, and positions within the team or organization are loosely coupled to an evolving situation |
|
Commitment serves to provide a team or organization with purpose, order, and value |
Commitment transforms unorganized perceptions, experience, and positions into a more orderly and purposeful team |
|
Environmental manipulation (action-based) |
Teams and organizations selectively act within their operational environment to conform that environment to their constructed reality Manipulation reflects the role of the team or organization in actively shaping the future |
Sense-making is more than merely the passive interpretation of the operational environment as given; it involves the active constitution of a workable reality within which a team or organization operate |
Manipulation can take the form of pre-emptive actions taken to shape the problem space even before that problem space is completely understood |
Sense-making links beliefs and action together within an understandable framework; hence, the construction of a reality can involve both hypothesis building and action taking |
Source: Weick, K.E., Sensemaking in Organizations, Sage Publications, Thousand Oaks, CA, 1995.
What seems to be needed to support the sense-making process is a type of non-monotonic logic; one appealing model is the abductive process, which pursues plausibility rather than accuracy (Lundberg 2000). Another way to view this is that we apply abduction when there is a lack of dependable causal models as typically driven by the traditionally deductive data fusion frameworks, i.e., when only “symptoms” are available and plausible causes have to be developed. However, there is likely no single inferencing process that can be argued as the foundation of sense-making; an inferencing toolkit is probably a better model.
There are various important messages for the information fusion community in reviewing the characterizations of sense-making:
• One is that sense-making and rational decision-making will in many cases need to coexist—they are each appropriate to different problem classes, and will very likely require different data fusion processes to support them.
• Another is that the fusion community needs to make a determination of whether it is possible for fusion processing as it is known today to fit into or be extended in some way to support the sense-making process.
• But the fusion community also needs to reflect on and develop a new model for fusion as supportive of sense-making per se, and what the new functional model of that process should be and what the technological challenges are toward implementing that model.
3.9 SELF-ORGANIZATION AND SELF-SYNCHRONIZATION IN THE VALUE CHAIN
The problem framework that gives rise to the need for a sense-making process can be said to form one of the drivers that fosters the need for self-organization of an operating unit: a sense of tension or difference, misunderstanding, or under-determination where meaning is in dispute. This tension necessarily or at least naturally leads to a need for communication and the new type of social dynamic that sense-making is. Hammond and Sanders (2002) argue that the dialogic creation of meaning (one could say sense-making) is a self-organizing process. They suggest that it is the tension between disorder created by randomness and order imposed by shared meaning that drives the need to communicate. However, while communicative activity aids in creating meaning and order in the face of equivocal information, the communication processes create disorder at the same time. What happens is that as the group begins converging on a problem solution, new directions begin to emerge in a kind of convergent-emergent tension. This engenders a bit of a twist on the sense-making process characterized as only convergent to a consensus; it is likely that in the confusing, equivocal environments that sense-making is designed for that divergent factors will enter into the process. Wheatley (1992) describes this as a productive localized “chaos” that enables the opportunity for participants to let go of previous assumptions and seek “out of the box” solutions.
As regards self-synchronization, the mostly widely quoted definition of self-synchronization related to NCW comes from Cebrowski (Cebrowski and Garstka 1998): “Self-synchronization is the ability of a well-informed force to organize and synchronize warfare activities from the bottom-up. The organizing principles are unity of effort, clearly articulated commander’s intent, and carefully crafted rules of engagement. Self-synchronization is enabled by a high level of [knowledge of] one’s own forces, enemy forces, and all appropriate elements of the operating environment. It overcomes the loss of combat power inherent in top-down command directed synchronization characteristics of more conventional doctrine and converts combat from a step function to a high-speed continuum.” A simpler definition of self-synchronization (Costanza 2003) is “the ability of a well-informed force to organize and coordinate complex warfare from the bottom up.”
It is usually considered that the “self” in “self-synchronization” implies the ability of an agent to arrange the timing aspects of its own activities without the influence or control of other agents, implying a sense of independence. In terms of analysis and decision-making style, to be independent an agent needs to be proactive in his actions otherwise he may be captive to the reactions driven by the adversary. Other factors necessary for enabling self-synchronization include maintaining an awareness of commander’s intent at all times, i.e., operating within that mind-set, and being able to dynamically prioritize activities. It is of course not usual that an agent acts strictly alone, so the notion of “self” in realistic cases relates to a kind of collective self-synchronization, and each agent in such collectives must be thinking synergistically, having a willingness to share resources and power. It also implies that such agents are synergistic communicators—empathetic listeners that understand the basic needs of a collaborator that enable achieving actions which are truly helpful to both agents, rather than compromises coming from negotiation-type communications. In the end, the self-synchronizing collective molds itself to the tasks and operations at hand; the molding forces are a kind of shaping context of people, problems, and resources. These factors are not unlike the “seven habits of highly successful people” that Covey (1990) sets as imperatives, e.g., being proactive, operating with an end in mind (e.g., commander’s intent), having priorities, thinking synergistically, and seeking first to understand.
3.10 COMPLEXITY IN SENSE-MAKING AND COMMAND AND CONTROL
Self-organization and self-synchronization are easy to talk about but very difficult to execute in the best way. Part of the rationale regarding the need for such agile behavior comes from the “Law of Requisite Variety” of Moffat (2003), where it follows from cybernetic arguments that to properly control a complex system (the dynamic asymmetric battlefield), the variety of the controller function (the number of accessible states which it can occupy) must match the variety of the combat system itself. In other words, the control system itself, here the C2 (human-based) organization, has to be complex. This Law of Requisite Variety implies that the control system must exhibit great agility in dealing with the dynamics and complexity of combat involving hybrid teams. But that agility must be controlled to some degree else it can result in chaotic behavior. According to Moffat (2003), “the representation of the C2 process must reflect two different mechanisms. The first is the lower level interaction of simple rules or algorithms, which generate the required system variety. The second is the need to damp these by a top-down C2 process focused on campaign objectives.” In a broad sense, the relationships between complex concepts and the behavior of an “information age force” are characterized as shown in Table 3.4.
Thus, it is not surprising to see considerable literature discussing the NCW sense-making and C2 processes as modeled by a complex adaptive system (CAS). If a CAS model is appropriate, then there is a need to understand CASs well enough to predict their macro-level behavior, a result of nonlinear micro-level behaviors. A related goal is to design and construct a CAS-based C2 process having a desired, or perhaps bounded, emergent behavior with a theoretical understanding that the emergent behavior will be most fit for a particular C2 or mission objective. The CAS/C2 literature speaks of the C2 process as ideally operating “on the edge of chaos”; i.e., within the favorable, predictable macro-behavioral bounds of the inherent CAS C2 process, but not tipping into chaotic behavior.
TABLE 3.4
Relations between Complexity Factors and Force Factors
Complexity Concept |
Information Age Force |
Nonlinear interaction |
Combat forces composed of a large number of nonlinearly interacting parts |
Decentralized control |
There is no master “oracle” dictating the actions of each and every combatant |
Self-organization |
Local action, which often appears chaotic, induces long-range order |
Nonequilibrium order |
Military conflicts, by their nature, proceed far from equilibrium. Correlation of local effects is key |
Adaptation |
Combat forces must continually adapt and coevolve in a changing environment |
Collectivist dynamics |
There is continual feedback between the behavior of combatants and the command structure |
Source: Moffat, J., Complexity Theory and Network Centric Warfare, CCRP Press, Washington, DC, 2003.
Since information fusion processes are information-providing processes into such decision-making and C2 operations, it is then important for fusion process designers to understand that they are supplying information into this nonlinear decision-support environment. One way to study such interdependencies is via the multi-agent systems construct, and probably the most research in CAS for C2 has been along these lines. Some of the notable examples of using intelligent agents to study emergent behavior in warfare are the Irreducible Semi-Autonomous Adaptive Combat (ISAAC) works, and the Enhanced ISAAC Neural Simulation Toolkit (EINSTein), from the U.S. Marine Corps Combat Development Command (MCCDC) as part of their Project Albert research (Ilachinski, 1999). There are yet other efforts that have employed the agent paradigm for such research (Hummel et al., 2005, Yang et al., 2005, Lauren 2000). These test beds have been used for a wide variety of research studies that have aided in developing insights into the behaviors and performance of CASs. Other methods have been applied to explore the CAS-data fusion interdependency, but overall, the research and thus design knowledge is limited; this is considered a robust area for needed research.
Regarding other methods, Urken (2011) has studied “error-resilient data fusion” (ERDF) processes, in which the contributors to the formation of a composite situational estimate employ voting procedures. In the ERDF approach, the properties of the systems used to represent and aggregate votes produce a high probability of producing what Urken calls “error resilient collective outcomes” (ERCOs). When such a voting process produces a reliable ERCO, neither outstanding votes or data, nor unelapsed time, will change the collective inference, yielding a robust result or situational interpretation. So ERCO results provide a basis for ignoring uncollected critical data and enabling agents to take immediate action to adapt to changes in their environment. Alternate approaches to dealing with CAS aspects for both fusion and network design have been put forward in a limited body of work, such as the biologically inspired strategies described in Urken (2011) and Ferro and Pioggia (2009). However, by and large, the information fusion community has not developed an organized research strategy to explore the nature of fusion functions and processes in the context of CAS.
It is anticipated that not only the military but extensive business and civil systems will be operating in a network-centric context from the point of view of the underlying informational infrastructure. There are clearly advantages to employing networked systems but there is little doubt that there are also system design trade-off issues regarding the formation of the physical network and perhaps the even more important issue of how the network is used. In the value chain characterization, one can to some degree build in ways to improve information quality and sharing through mandated processes and protocols, but the intermodal interactions and human inputs and controls also play into the overall effectiveness equation. If the sense-making and CAS paradigms indeed apply toward modeling such interactions, the information fusion community will need to better study and understand how to design fusion processes to operate in these highly adaptive and nonlinear user environments. The implications of these new models of “sense-making,” consensus-formation, convergent–emergent interpretation dynamics, productive local chaos, etc., on the requirements for data fusion process design and development are likely to be rather revolutionary.
Alberts, D. S. 2002. Information age transformation. CCRP Program Monograph online, http://www.dodccrp.org/html4/books_downloads.html
Alberts, D. S., J. J. Garstka, R. E. Hayes, and D. T. Signori. 2001. Understanding Information Age Warfare. Washington, DC: CCRP Publications.
Alberts, D. S. and R. E. Hayes. 2003. Power to the edge. CCRP Program Monograph online, http://www.dodccrp.org/html4/books_downloads.html
Alberts, D. S. and R. E. Hayes. 2006. Understanding command and control. CCRP Program Monograph online, http://www.dodccrp.org/html4/books_downloads.html
Cebrowski, A. K. and J. J. Garstka. January 1998. Network centric earfare: Its origins and future. U.S. Naval Institute Proceedings, 124(1): 35.
Costanza, C. D. May, 2003. Self-Synchronization, the Future Joint Force and the United States Army’s Objective Force. Monograph, Fort Leavenworth, KS: School of Advanced Military Studies United States Army Command and General Staff College.
Covey, S. 1990. The Seven Habits of Highly Effective People: Powerful Lessons in Personal Change. New York: Fireside Publishers.
Evans, P. and T. S. Wurster. 2000. Blown to Bits: How the New Economics of Information Transforms Strategy. Boston, MA: Harvard Business Press.
Evidence Based Research, Inc. November 2003. Network centric operations conceptual framework version 1.0. 2003. Report prepared for Office of Force Transformation, http://www.oft.osd.mil/library/library_files/document_353_NCO%20CF%20Version%201.0%20(FINAL).doc
Ferro M. and G. Pioggia. 2009. A biologically based framework for distributed sensory fusion and data processing. In: Sensor and Data Fusion, N. Milisavljevic (Ed.). Vienna, Austria: InTech.
Hammond, S. C. and M. L. Sanders. 2002. Dialogue as social self-organization: An introduction. Emergence: Complexity and Organization 4(4): 7–24.
Hummel, J. R., J. H. Christiansen, C. M. Macal, and M. J. North. White paper, 2005. The development of complex adaptive systems based decision support systems. Decision and Information Sciences Division, Argonne National Laboratory, Argonne, IL.
Ilachinski, A. 1997. Irreducible semi-autonomous adaptive combat (ISAAC): An artificial-life approach to land combat. Military Operations Research 5(3): 29.
Ilachinski, A. February 1999. Towards a science of experimental complexity: An artificial-life approach to modeling warfare. Special issue of Kybernetes Journal.
van Laere, J., M. Nilsson, and T. Ziemke. 2007. Implications of a Weickian perspective on decision-making for information fusion research and practice. 10th International Conference on Information Fusion, Quebec City, Quebec, Canada.
Lauren, M. K. 2000. Modeling combat using fractals and the statistics of scaling systems. Military Operations Research 5(3): 47–58.
Leedom, D. K. 2001. Final Report: Sensemaking Symposium. (Technical Report prepared under contract for Office of Assistant Secretary of Defense for Command, Control, Communications & Intelligence.) Vienna, VA: Evidence Based Research. Inc., http://www.dodccrp.org/files/sensemaking_final_report.pdf
Leedom, D. K. 2004. The analytic representation of sensemaking and knowledge management within a military C2 organization. Air Force Research Laboratory Human Effectiveness Directorate Report AFRL-HE-WP-TR-2004-0083.
Lundberg, C. G. 2000. Made sense and remembered sense: Sensemaking through abduction. Journal of Economic Psychology 21(6): 691–709.
McCaskey, M. B. 1982. The Executive Challenge: Managing Change and Ambiguity. Marshfield, MA: Pitman Publishers.
Moffat, J. 2003. Complexity Theory and Network Centric Warfare. Washington, DC: CCRP Press.
Owens, W. A. May 1995. The emerging system of systems. U.S. Naval Institute Proceedings (121): 36–39.
Porter, M. E. 1985. Competitive Advantage: Creating and Sustaining Superior Performance. New York: The Free Press.
Rittel, H. and M. Webber. 1973. Dilemmas in a general theory of planning. In: Policy Sciences, Vol. 4, pp. 155–169. Amsterdam, the Netherlands: Elsevier Scientific Publishing Company, Inc. [Reprinted in N. Cross (Ed.). 1984. Developments in Design Methodology. Chichester, U.K.: John Wiley & Sons.]
Sieck, W. R. et al. May 2007. FOCUS: A model of sensemaking, Technical Report 1200, http://www.au.af.mil/au/awc/awcgate/army/tr1200.pdf
Urken, A. B. 2011. Voting theory, data fusion, and explanations of social behavior. Paper from the AAAI 2011 Spring Symposium, pp. 29–34. Stanford, CA.
Weick, K. E. 1969. The Social Psychology of Organizing. Reading, MA: Addison-Wesley.
Weick, K. E. 1995. Sensemaking in Organizations. Thousand Oaks, CA: Sage Publications.
Wheatley, M. 1992. Leadership and the New Science: Learning about Organization from an Orderly Universe. San Francisco, CA: Berrett-Koehler.
Yang, A., H. A. Abbass, R. Sarker, and M. Barlow. 2005. Network Centric Multi-Agent Systems: A Novel Architecture. TR-ALAR-200504004. The Artificial Life and Adaptive Robotics Laboratory, School of Information Technology and Electrical Engineering, University of New South Wales, Kensington, New South Wales, Australia.
Zack, M. H. 1999. Managing organizational ignorance. Knowledge Directions 1: 36–49.