Using Evolutionary Theory in Explaining System of systems Development

Introduction

The paper describes one way to model evolution and development of System of systems. In the context of this paper, the System of systems is a socio-technical structure where Information and Communication Technology is intertwined with human and other structural components. Because of openness and intertwined components, System of systems is complex in nature. Thus, there is a need to understand better the relationships and interactions between components and their environment.

The hypothesis for a model is based on Darwinian (1859) Theory of evolution. It is expanded by Mokyr’s (1998) application of the evolution of technical systems. The Mokyr’s system is modularized using Andriani’s and Carigani’s (2012) approach to modular design and innovation. The modular system is then generalized as System of systems.

Mokyr defines knowledge as underlying structure needed to design systems. The model for knowledge is extended using features of Choo’s (1998) model for Knowing Organization. The extended model is projected to the environment of generic human community and military affairs to expose dependencies between entities and their environment. The forces exposed are explained by combining Mokyr’s and Choo’s theories and reviewing them with Bertalanffy’s (1968) general system theory.

Finally, the model is simplified to illustrate essential interactions and dependencies. The simplified model is easier to use in studies of Information and Communications Technology development. For this purpose, the extended and streamlined model is also explained using classical military development example of mechanized forces and maneuver tactics (Vego, 2007).

Darwinian evolutionary model as basis

The standard Darwinian model includes an Underlying Structure (genotype) that defines the Manifested Entity (phenotype). This connection between Structure and Entity is tight so at moment t Structure produces (maps) similar Entities. However, Darwinian model is a dynamic system. Within time, the Structure might change (mutation) and produce different Entities. When Entity (t) and Entity (t+1) are facing the similar stress of Environment, Natural selection may occur. One of the Entities may adapt better to the stress Environment is causing. The reproduction of better adapted Entity may be more vigorous and gradually overtake the less adapted Entity. See illustration in Figure 1.

Figure 1: a standard Darwinian model for Evolution

Natural selection shapes the character defined in Structure and expressed in Entity. Adaptation means that characters of an Entity evolve gradually to fit better into required function. Exaptation means that a character adapted to a particular function is adopted to a new use (Gloud; Vrba 1982). 

Extending the definition of Knowledge and System

Joel Mokyr (1998) uses the Darwinian model in explaining the evolution of technical systems in Human Societies. He names Knowledge as the Underlying Structure that defines the characters of technical System (Manifested Entity) pictured in Figure 2.

Figure 2: Mokyr’s model for evolution of technology

Mokyr proposes that designing or building a technical system needs useful knowledge. The useful knowledge at the moment (t) defines the characters of System (t). The system is used by a community of people for a function in a particular environment. New knowledge can be created by research or experimenting with existing systems. If the newly accumulated knowledge appears to be useful, then new system (t+1) is designed   and build. Natural selection occurs when the community is comparing new and old system and adopting either of them for further use.

An alternative evolution path is opened if users start to use old system for some new functions. While learning by doing, the user might not understand the principles of this new combination, but it seems to produce results. 

The Mokyr’s evolution model for technical systems has been further extended by Andriani and Carigani (2012) in their study for modular exaptation. It states that Systems are most usually evolving at substructure, module level. In combining System from different modules, it becomes possible to adopt modules elsewhere and use them as part of the new combination. The approach opens the third path for evolution model: new modules are designed, composed of previous modules, and existing modules elsewhere are included into the composition. The composition is further called System of systems (Wikipedia 2016) as it is defined by interrelated modules or sub-systems that produce or enable more functionality and performance together than separately. Figure 3 presents an extended evolutionary model for System of systems development.

Figure 3: Evolution model for System of systems

Cattani (2005) extends the evolutionary model further by proving that organizations can accelerate their knowledge creation by preadaptation. An organization can produce modules to fit better in existing the System of system and further usage within the community when they are running parallel design lines and experimenting with various combinations. Preadaptation provides a wider variety of possibilities to choose for production. Accumulated creative capital may appear to be useful further when new functional demands rise (Harford, 2011: 234).

Features of knowledge and information

In Mokyr’s evolutionary model, knowledge is accumulated by research and experimentation. There is also the possibility to learn by doing, but it transfers to knowledge only if the phenomena can be explained with some existing model, thus understood. When knowledge accumulation is studied from Choo’s (1998) knowing organization approach, knowledge creation provides a model to enrich tacit knowledge to explicit with Nonaka’s (2015) SECI process. Innovation is also often based on coopting modules or subsystems elsewhere and utilizing them in a different combination (Christensen, 2011). 

Knowledge can be active when it is referred, and it remains accessible. Knowledge can also be dormant, kept in storage, and not utilized in designing new components. Knowledge has characteristics as follows (Dalkir 2011:2):

• Using knowledge does not consume it.
• Transferring knowledge does not diminish it.
• Knowledge is plentiful, but there are only a few opportunities to use it.
• Much of the explicit knowledge is lost in filing and much of the tacit knowledge out the door at the end of the day.

So there is a window of opportunity for each piece of knowledge if it is available at the time. As the human community is now doubling its amount of information every two years, there is always information which never reaches the level of knowledge. Besides learning, research, and knowledge creation, an organization can also acquire knowledge with new hires.

Mapping knowledge to design and building new or legacy systems depends on the following status (Mokyr 1998):

• As long some selecting agents do not share the consensus of knowledge, there remain old design and systems as niches. The military has a tendency to sustain traditions as horses in their ranks long after they were fully replaced by motor vehicles in operations (Guderian, 2001).
• If some part of knowledge is rejected by the majority of selecting agents, it does not vanish but survives in records. There is a change that after generation change among selectors, previously rejected knowledge find a new use. Frequent rotation of officers in several positions sometimes provide opportunities both to senior and junior officers to proceed with their earlier ideas (Harford, 2011:50).
• Most of the information are dormant. There is no know utilization for that information until new ways of analyzing information emerges. Business Intelligence and Big Data are trends in trying to make use of exponentially increasing the volume of data (Schmidt & Cohen, 2013:34).
• Human way to process information. There is a loss of information in human communications. All ways of communication need to be enforced to transfer knowledge. Human recognizes patterns and matches them with most recent or familiar ones to create understanding. Human thinking is tuned towards satisfying rather than optimizing (Snowden, 2012).

Mokyr modeled the awareness of existing knowledge, the unknown knowns, with access cost. The access cost is replaced with Choo’s (1998:61) model for information use as illustrated in Figure 4. For more complex approach one should see Dalkir’s (2011) Knowledge Management in Theory and Practice.

Figure 4: Mokyr’s Evolution Model extended with Choo’s model for Information Use

Choo’s (1998) process for information use includes three functions: Information Needs initiate Information Seeking, which enables Information Use. There is feedback from Use to Needs as process unveils possibilities and constraints. The process runs embedded in two sub-environments: 

• Information processing environment, where the need and seek processes are affected by cognitive needs and affective responses.
• Information uses environment, which is defined by the situation of use.

The cognitive needs in processing environment are affected by situational stops, where human faces different gaps, barriers or options when reaching after information. The human uses available bridging strategies to overcome these gaps. Human also has different habits in using acquired information. The affective responses include feelings that are defining behavior. Uncertainty, confusion, and anxiety are examples of prohibiting feelings. Feelings of confidence, optimism and clarity are driving information seeking. Relief, satisfaction or disappointment are the possible outcome from information usage.

Work and social settings define information usage environment. Varying the combination of people facing the same problem produces different outcomes as social structure affects their thinking. Organization, task domain and access to information also define the outcome. Also, expectations concerning resolution and the way it is communicated, have an impact on the solution.

Effects of Community and Environment of use

The model assumes that System of systems is open. It has external interactions with its environment and in this case particularly with its users, the community of human beings (Bertalanffy, 1969:141). Mokyr (2004) defines features of resistance in adopting new systems. Meanwhile, Harford (2011) defines features that might accelerate adaptation. 
In process of natural selection between legacy system and new system, Mokyr defines the following sources for resistance to adaptation:

• Value. The change will have effect in existing value chain and power structure. There will be stakeholders within the community that feel losing either value or power if the new system is being adapted. Thus, they will oppose the change.
• Ideology. Human community includes a degree of phobia and conformism that will resist all new. They may be defined at political, religious, cultural or ethnic levels.
• Epistemology. The existing system and knowledge hold a status of general acceptance. Human nature has a tendency to value possessions over future options. Thus, new information or components may be opposed as they do not fit into accepted understanding or usage.
• Systemic. Individual autonomic systems differentiate from others only per their features or price. In the other hand, if the new component has a decisive dependency on other parts of System of system, then systemic resistance needs to be considered. When the system of systems grows beyond control, as telex, land line telephone or TCP/IPv4 has, it slows down the evolution.
• Dependency on frequency. The rate of change and adoption depends on a number of users. In one hand, smaller and better-connected communities adapt new feature quicker than larger. The other hand, the large, well-connected population may adopt new features faster as social imitation/learning surface extends and touches more people. Some changes do follow a particular path of evolution that makes them more resistance to further changes. For example, a service established by legislation may enjoy protection from competition and natural selection.
• Comfort. Human being is seeking status quo with routines and rituals that provide comfort for everyday life. This comfort is further stabilized with habits that lessen the stress of decision making. In stressful situations, human often has a tendency to use familiar or previously proven solutions than unprecedented. Similar deepening comfort zone i.e. funnel of comfort is with communities as they seek stability (Duhigg, 2012).

There are also features in organizations ability to design new system that produces natural friction:

• Lack of complementarities. New knowledge may be available and useful to design a new module, but manufacturing technology is not mature to support building the physical form. The asymmetric situation is familiar to communities, whose strategy is defined by staying on the very edge of technical evolution.
• Speciation. Language is used in communicating existing and new knowledge. Some areas of expertise may develop jargon to differentiate themselves from the rest. This linguistic speciation may prevent dissemination of knowledge. Differentiation may also be founded at political, ethnic or ideological level. New understanding does not easily overcome barriers rooted in beliefs or upbringing. The base of trust is also powerful differentiator especially in the area of information security. It is hard to move the base of trust from physical formats towards virtual formats (Mattila, 2016).
• Organization structure. The organization, that is ought to produce new systems or their components, effects the outcome. Deeply functional and specialized organization can produce mainly components sized according to their functional unit. Cross-unit development does not occur unless there is a willing authority above all units involved. Matrix or composite organizations are more flexible and oppose less when creating new concepts and designs when a network of value changes (Christensen, 2011:33). Connectivity of the value network is also defining the adaptation of new modules. The more connected organization is, the more open it is to new knowledge and exaptation of useful modules.
• Entropy. The second law of thermodynamics can explain the loss of information as well as the gradual disorder that grows in open evolutionary systems. Information loses its content and integrity when communicated due uncertainty or entropy in a channel. The open socio-technical system also has some dynamic microstates with their agendas, both human and technical. The human community always has subcultures and unofficial organizations that affect the official organization and dominant culture. (Bertalanffy, 1969:143)

Figure 4 illustrates the effects of community and environment to embedded System of systems.

Figure 4: Environment and community effects in evolution of open System of systems

There are also positive forces in environment and community that may accelerate the evolution:

• Curiosity is a mean that fulfills the human need for knowledge or truth (Reiss, 200:31).
• Strategic transformation or changing environment is best approached with adaptation. Based on Donald Sull (Syrett, 2012:100) this is achieved by building three capabilities: 

1. A strategic anticipation to recognize patterns of change
2. Organizational agility to exploit opportunities or mitigate threats as they unfold
3. Resiliency to absorb uncertainty and surprises

• Adversary or Competitor may change the status quo of confrontation by action because of intolerable pressure or try to exploit occurred opportunity. Both states and enterprises are mutually dependent in a way that action prompts a reaction from other stakeholders (Porter, 1980:16). 
• Challenges of everyday life drive the process of evolution. When facing a problem, one tries out few variants available to solve it. Failures are weeded out and successes copied. This process of variation and selection goes on and provides constant drive (Harford, 2011:12). The sociotechnical system evolves with this process at the micro level (Tavistock, 1989). 
• Maturity. The S-curve theory of technical improvement explains that in the early versions of new systems, the pace of technical improvement is slow. With better understanding, control and diffusion, the rate of technological improvement will accelerate. In later stages of maturity, improvements will asymptotically reach a natural limit, hence the S profile (Christensen, 2011:44).

 

Optimistic model for evolution of System of systems

A system that composes of individual subsystems that are interrelated is called System of systems (SoS).  The SoS is open and interacting with its environment and community that is using it. The SoS has been designed to fulfill a function based on knowledge that the community possess and can use. As an open system, the SoS has a tendency to lose its coherence gradually with time. Friction and entropy are powers that change the structure and usage of SoS at the micro level. 

There are three main paths for System of systems to evolve:

1. Preadaptation is driven by the need to develop new SoS’. It includes research, experimenting or acquiring new knowledge with other means. Several optional solutions may be produced and explored to find the best fit. Gained knowledge and prototypes are used to design new SoS’ to fulfill the requirements of the new function.
2. Adaptation happens when the SoS is coopted gradually for different usage without necessarily understanding why it fit to the new function.
3. Exaptation occurs when component C from another system is coopted as part of SoS’ in making it more efficient or fitting to the purpose.

There are driving and resisting forces that effect in the evolution of function and SoS. This optimistic model is simplifying them into two opposing forces: Resistance and Drive. The optimism is the decline in the picture in Figure 5. The model assumes that there is a generic drive to improve and develop the performance of community, systems it is using, and knowledge it possesses.

Figure 5: An optimistic model to explain evolution of System of systems

Knowledge is imperative in preadaptation and exaptation. For knowledge creation there is a process of information need, seek, and usage. Environment and community are affecting this process from cognitive, affective and situational dimensions.

The model can be explained with different evolutions of socio-technical systems, but one of the military classics is the development of mechanized warfare before the II WW. Heinz Guderian has been recognized as one of the main inventors of mechanized warfare. He wrote 1936 a book called Achtung – Panzer that described battle tank as component enabling mobile warfare when supported with radio communications, motorcycle reconnaissance, and close air support. He proved the case by leading the attack of his armored forces through the Allied defense 1940 with speed that overtook the Allied leadership by surprise both at tactical and strategical level. Speed and firepower of maneuver tactics were named “Blitzkrieg”. Later 1951 in his book “Panzer Leader”, he described the challenges faced when trying to develop new mobile tactics within the Wehrmacht through 1930’s. The first battle tank was coopted (Exaptation) from earthmover as tracked vehicle that was able to carry armor, large gun or having high maneuverability.  Singular tanks were used during I WW with variable success. After the war, their technical development went further, but tactical use remained at the level of singular weapon platform (Adaptation). 

Guderian explained these facts, proceedings in technology, and defined new tactics for land warfare in 1936 (Preadaptation). The generals of cavalry did not favor his ideas (Resistance). Despite this, he continued developing (cognitive means and affective responses) panzer tactics in motorized logistics company using trucks and cars (Experimenting or Domain-driven Design). He also cooperated with Luftwaffe as they were developing dive bomber (Ju 87) 1935 and introduced close air support (Exaptation). Artillery pulled by horses were not quick enough to support mechanized spearheads.  The strategic need (Drive), for Germany to increase tactical speed, overdrive the resistance. When Germany launched the offensive 1940 against France and Allied defense, there were as many battle tanks on both sides. The Guderian way of using armored divisions and operation plan created by Manstein (Melvin, 2010) made the crucial difference and results prompted others (Soviet Union, USA) to copy (Adaptation) mobile tactics with variable outcomes.

Discussion

The main objective of this paper is to present a logical model to study the life cycle of open socio-technical systems and causation of their development. The model seeks to improve the understanding of evolution in the usage of Military Information and Communications Technology.

The model is based on Mokyr’s evolutionary model for technical development. It is further extended by Andriani and Carigani’s (2012) modular exaptation to create a better understanding of System of systems development. Intentional knowledge creation is explained using Cattani (2002) model of preadaptation. Choo’s (1998) approach for the Knowing Organization is used to understand better forces affecting information processes. Christensen (2011) introduces exaptation as a possible path for evolution. The composed model is evaluated against Bertalanffy’s general system theory as major forces effecting evolution are simplified. Finally, a practical, optimistic approach (exploitation phase in the flow of events in Gunderson & Holling’s (2002) Panarchy model) is chosen in providing the model with a positive incline towards development.

The model is defined only to sufficient level to gain a robust understanding of forces and interrelationships in the evolution of socio-technical complex systems. The model is used in other papers to study the evolution of military Command, Control, Communications, Computers and Information System of systems and related functions.

At the macro level, the model does not consider for example transformation model developed by Gunderson and Holling (2002). At the micro level, presented model does not describe a holistic open system with feedback and decision-making functions special for social behavior (Ackoff & Emery, 2008). It requires further Systems Thinking, Modeling and experimenting to be matured and detailed. The model creates a generic framework to study recent Information and Communications Technology related developments, especially in Military Communities.