On the Coalescence of Supply Networks and Information ... › uploads › media... · networks, e.g. the electricity networks or road networks, and information systems. From a business

EMNet 2011 International Conference on Economics & Management of

Networks December 1 – 3, 2011, Limassol, Cyprus

On the Coalescence of Supply Networks

and Information Systems

Bernd Hellingrath, Jörg Becker,

Daniel Beverungen, Carsten Böhle,

Michael Räckers

ABSTRACT

This paper discusses the on-going coalescence of traditional supply networks and information systems which merge to hybrid, IT-enabled networks. These make possible new business models but their design and operation are influenced by emergent effects which are not yet fully understood. The aim of this paper is to describe properties of such networks and outline which theoretical foundations can be employed for further research and which questions arise.

Information Systems, Network Design, Smart Grids, Logistics, Research Approaches

2 On the Coalescence of Traditional Networks and Information Systems

1 Information Systems for Supply Networks

The purpose of our paper is to outline the ongoing coalescence of traditional supply

networks, e.g. the electricity networks or road networks, and information systems.

From a business perspective, this effects phenomena such as managing alliances and

networks or developing ‘smart’ business models that become feasible only with the

support of information systems (such as smart grids, power by the hour business

models, or advanced resource scheduling approaches in supply chains).

Although a precise and concise definition of hybrid networks, i.e. in this context

traditional supply networks that have been enriched with information systems, does

not yet exist, these kind of networks have been identified as a crucial asset for

services and operations that are essential for modern societies and economies. They

are comprised of assets both tangible (e.g. streets and sensors) and intangible (e.g.

information systems). Traditional networks such as electrical grids are augmented

with advanced information systems (i.e. they are informated) for enabling added value

such as real-time monitoring and control. A promising but not yet sufficiently

fathomed phenomenon in these networks is the apparent contrast of design and

emergence. While artifacts like information systems can be engineered purposefully

to solve a design problem, the high degree of connectedness in networks results in

emergent behavior which occurs and spreads in unexpected and nondeterministic

ways. As networks are increasingly interconnected, e.g. via the Internet,

understanding and directing emergent properties will be a crucial dynamic capability

in the near future.

The interdependencies of networks, information technology, social structures, and

human interaction in an IT-based network context is, therefore, a complex

phenomenon that has to be researched from different angles in order to identify and

influence evolution, aggregate behavior, and anticipation in networks. Among others,

this comprises the interplay of analytic research with synthetic research approaches,

one manifestation of which is the need to bridge the epistemological gap between

positivism and constructivism, and the research methods favored in both schools.

The upcoming challenges in hybrid networks require integrated research endeavors

that include a variety of perspectives and methodologies. The CAS perspective

predicts that when an engineered part is connected with the network, it will interact

with this network in unforeseeable ways. This implies that the high degree of

connectedness that was created to solve certain problems in fact might create new

problems as well as new opportunities. This renders the deterministic control and

management of networks impossible and raises entirely novel research questions,

such as: How can the businesses and economies make use of hybrid networks without

offsetting this value with unforeseen side effects? How much of the behavior in

hybrid networks can be controlled and how much is deliberately left to emergence?

How can strategies of the parts of a network be aligned with strategy on a network

level?

These questions become apparent in the scenario of industrial supply networks

which have become increasingly complex, global, and impossible to manage in their

entirety. A car manufacturer can have up to 1,000 1st tier suppliers and networks may

be as deep as 7-8 tiers. Supply networks heavily rely on IT systems such as ERP and

communication systems. While organizational and technical relations to 1st tier

Hellingrath, Becker, Beverungen, Böhle, Räckers 3

suppliers can be managed, this is usually not the case for suppliers further upstream.

From a positivist point of view, only additionally connecting these suppliers to the

manufacturer’s systems could bring an improvement. However, suppliers are reluctant

to do so as they are involved in multiple supply networks and fear being taken

advantage of or having to disclose confidential information. It is therefore interesting

to consider indirect analysis of behavioral patterns and respective measures from a

CAS point of view.

As another example, smart grids are power supply systems with the ability to

transfer information. They allow controlling electrical devices at customers’ homes

for making better use of electricity (e.g. a washing machine that starts operating as

soon as excess energy is available). Whereas the technological artifacts for smart

grids can be designed in an engineering approach, the usage patterns of power

suppliers and customers cannot be reliably estimated but emerge as the result of the

actions of millions of energy consumers, or may be subject to disruptions or attacks.

We outline the necessity for researching the coalescence of traditional networks

and information systems for their property that they can be partially controlled but

also feature emergent behavior. We argue that bringing together various research

disciplines for joined research on digitally-enhanced networks has to bridge the gap of

the traditional research paradigms applied in these disciplines. A reconciliation of

research methods can lead to a better understanding of the phenomena under

investigation. The objective of this paper is to contribute some thoughts and research

ideas concerning the design of hybrid networks.

The remainder of the paper is organized as follows. In Section 2, the historical

development as well as several theoretical lenses for investigating hybrid networks as

a phenomenon are discussed. In Section 3, different research methods are presented

which can be used for analyzing the process of coalescence. In Section 4, two

application scenarios are presented. In Section 5, an overview is given of the research

questions that are raised concerning the increasing use of IT systems within

traditional networks. Section 6 is an outlook section.

2 Theoretical Background

2.1 Historical Development of Supply Networks

Before describing hybrid networks, a discussion of the underlying traditional

supply networks, in their entirety often referred to as infrastructure, is necessary.

Integrating the infrastructural perspective and its body of literature is very beneficial

as it fits in nicely with constructivist ideas and helps extend the traditional

engineering approach. Therefore, a discussion of how the networks of concern did

develop and what properties can be derived from their status as infrastructures, will be

explored in the following.

The term “infrastructure” is commonplace, but giving a concise explanation of its

meaning is not easy. Etymologically, it combines the Latin prefix “infra”, i.e. below,

and “structure”. Van Laak (Van Laak 1999) argues that its first use dates back to 1875

when French railroad engineers referred to embankments, dikes, and bridges as

“infrastructure” as opposed to “superstructure”, e.g. rails, stations, or signals. French


dictionary Petit Robert seconds this and additionally states that its use for technical or

economical equipment started in the twentieth century (“(milieu XXe) Ensemble des

équipements économiques ou techniques.”). American English Merriam-Webster

dictionary gives 1927 as the year of its first known use. It was intensely used since

1951 when NATO started a program to standardize certain facilities such as airports

or pipelines which from then on were referred to as infrastructure. The first reference

to immaterial components can be traced back to Jacques Vignes, according to Van

Laak, in the context of development aid. He included social welfare, health care,

education and more, arguing that “any equipment and any dedicated investment” can

be understood as belonging to infrastructure, “although they are not directly

productive, but necessary for production or its development” (Vignes 1958). This

description bears some resemblance to Adam Smith who wrote that the state should

build and maintain such facilities which are useful for an economy but that can never

be profitable and consequently will not be operated by a private individual.

Finally, Jochimsen made ground-breaking work with his research on the theory of

infrastructures when he distinguished material, institutional, and personal

infrastructures (Jochimsen 1966). Material infrastructure “is understood to represent

capital goods in the form of transportation, education, and health facilities, equipment

of energy and water provision, facilities for sewage, garbage disposal, and air

purification, building and housing stock, facilities for administrative purposes and for

the conservation of natural resources”. Institutional infrastructures are “all customary

and established rules of the community as well as the facilities and procedures for

guaranteeing and implementing these rules by the state”. Last, personal infrastructure

“comprehends the number and the relevant properties of the working population (for

example, general and special education, qualification in different functions)”. These

categories cannot be seen separately, but they always have to be seen in combination.

Thinking of health care, material infrastructure is represented by hospitals, personal

infrastructure by doctors and nurses, and institutional infrastructure by health

insurance companies and legal regulations. One without the other is useless, e.g. it is

not advisable to build up material infrastructure while neglecting the complementing

categories.

In the field of Information Systems, “IT infrastructure” is the usual term for any

middleware, e.g. mail servers, that enable the use of the enterprise’s software

applications. Hanseth (Hanseth 2010) argues that IT solutions which are developed

today are fundamentally different from traditional information systems. Rather than

that, most solutions today are not entirely new implementations but are based on

existing structures. These are infrastructures which Hanseth defines as a “shared,

evolving, open, standardized, and heterogeneous installed base”. Thus, new strategies

and approaches are necessary, less of a system thinking and more of an infrastructure

perspective, because “the notion of systems make us believe that we through our IS

design methodologies are in complete control over the design process and accordingly

that we can design an IT solution exactly as we (and the users) want to”. Instead, “a

user cannot avoid being a designer”. So, whereas systems can be designed,

infrastructures have to be cultivated. Cultivation means that infrastructures cannot be

enforced but that they grow only with user acceptance. He cites the metaphor given

by Dahlbom and Janlet (Dahlbom and Janlet 1996): “the tomatoes themselves must

grow, just as the wound itself must heal”.


However, in this paper we are analyzing so-called hybrid networks as opposed to

“purely” digital infrastructures. As such we consider the combination of traditional

supply networks, e.g. roads or electricity, with information technology. They are a

sign of the increasing coalescence of technology and all other spheres of life and

business.

2.2 Positivism and Constructivism

A positivist point of view holds the epistemological assumption that truth is

objective and can be discovered by formulating and empirically testing hypothesis

based on predefined cause-and-effect relationships. From a constructivist perspective

researchers interact with the subjects under investigation and construct a subjective

understanding of the reality constrained by bounded rationality. Since traditionally the

solutions designed were operated in relative isolation, the positivist viewpoint was

favored in the design of information systems. Now that IT artifacts are embedded in

highly connected networks, previously unrecognized side effects occur that go beyond

the traditional cause-and-effect relationships. Instead, a novel understanding of how

the elements in hybrid networks interact and what aggregate behavior emerges is

required.

2.3 Complex Adaptive Systems

Hybrid networks can be viewed as Complex Adaptive Systems (CAS) that feature

three characteristics: Evolution, aggregate behavior, and anticipation. First, digitalized

infrastructures consist of autonomous parts which are able to learn and change

themselves. They can also connect to or disconnect from other parts which shapes the

evolution of the infrastructure. Second, the interplay of the parts leads to emergent

aggregate behavior that cannot be derived from adding up the actions of the parts, as

exemplified by the bullwhip effect in supply networks. Third, anticipation refers to

the ability of the parts of a network to anticipate the future based on current actions or

anticipated outcomes. This influences the behavior of the entire network in

unforeseen ways.

2.4 Adaptive Structuration Theory

On an IT level, Adaptive Structuration Theory (AST) has been proposed as a

theoretical base to conceptualize the interplay of advanced information technologies,

social structures, and human interaction. The idea of AST is that advanced

information technologies (such as hybrid networks) are designed based on the social

structures that prevail in the environment in which the network is intended to be

utilized. The act of design itself adds new structures that are inherent in the

technology. Consecutively, the IT constitutes an array of social structures that

provides occasions for the structuring of action. This means that a user can (and likely

will) do more and different things with the network than could be originally intended

by the designer. Therefore, hybrid networks encourage virtually unlimited


opportunities for networking which causes emergent aggregate behavior on an

alliance or network level.

3 Research Paradigms and Research Outline

3.1 Researching the Design of Supply Networks

Design and emergence are two principles for the evolution of hybrid networks. This is

inferred from the discussion of theories in section 2. The designer can consciously

design a hybrid network. However, due to the interplay with its socio-technical

environment, the network will behave in ways that cannot be anticipated fully by the

designer. Instead, unintended or even undesired side effects are likely to occur. A

prominent example is the introduction of email. Due to its general applicability for

manifold activities in an office, email is often used for purposes that it has not been

designed for in the first place, such as in the form of a knowledge repository or

archival system for information of any kind. This unintended side-effect emergent due

to the actions of people who were in the need for keeping their documents at a single

point of access that were lacking more elaborate systems, or who were unwilling or

unable to use these systems.

Investigating the complex interactions of design and emergence is very tricky to do

for various reasons. First, the analysis of the environment into which the artifact is

intended to function needs to be done by abstracting from the (seemingly) irrelevant

properties of the environment. Although this is necessary to keep the design process

focused, the properties that were abstracted from can be exactly the ones that will be

outside the scope of the artifact later on. Second, the design of IT artifacts itself can

be very complex and resource consuming. Therefore, even within the intended scope

of the IT artifact not all effects that result from the interactions of the artifact with its

environment can possibly be anticipated. Third, the artifact will likely need to interact

with other artifacts that do not remain stable along their lifecycle, but are subject to

extension, revision, or destruction. This will likely impact on the way in which the

artifact will act with its technological environment.

In order to sufficiently explore this multi-facetted relation between design and

emergence, a multi-disciplinary research endeavor is needed that integrates research

from two paradigmatic standpoints. On the one hand, design oriented research (Simon

1996, March and Smith 1995, Hevner et al. 2004, Peffers et al. 2008) is focused on

engineering man-made artifacts. On the other hand what Simon (Simon 1996) terms

as natural research is focused on building theory from observing and analyzing

phenomena without an intention to build artifacts. Notably, this view is not limited to

analyzing natural phenomena. Instead, the natural-type style of research is

conceptualized as research that is carried out with the traditional array of research

methods, based on collecting, interpreting data in order to identify causal

relationships.

We argue that integrating both perspectives in a comprehensive research

framework is required in order to intellectually grasp the complex interrelation of

design and emergence as found in hybrid networks. This argumentation is detailed in

the following sub-chapters.


3.2 Basic Premises of Design Science Research

Design Science Research (DSR) is a paradigm for research on how to develop

artifacts (artificial things as opposed to natural phenomena). In the IT domain, the

phenomenon under study is the design of IT artifacts, such as language constructs,

models, methods, and software instantiations (March and Smith 1995):

Constructs (vocabulary and symbols): Constructs form the vocabulary of a

domain (March and Smith 1995). They build the basis for defining problems and

specifying their solutions. Modeling languages are common collections of

constructs.

Models (abstractions and representations): Models are sets of statements

expressing relationships between constructs (March and Smith 1995). Building on

constructs, models are meant to represent certain situations of problems or

solutions. Reference models are special models and describe a class of relevant

real-world phenomena on an abstract level. They encapsulate knowledge and can

be reused for various purposes (Becker and Delfmann 2004), such as the design of

business processes, information systems other models.

Methods (algorithms and practices): Methods are sequences of steps used to

perform a task (March and Smith 1995). Typical examples are algorithms,

procedures or guidelines. Methods are often based on constructs and models to

represent the inputs and outputs of work steps.

Instantiations (implemented and prototype systems): Instantiations are realizations

of constructs, models or methods in information systems. „Models that work ‘on

paper’ will not necessarily work in real world contexts. Consequently,

instantiations provide the real proof” (March and Smith 1995). Instantiations can

be valuable to demonstrate or proof the utility of other artifacts by demonstrating

feasibility both of the design process and of the designed artifact (Hevner et al.

2004), as „Each new program that is built is an experiment. It poses a question to

nature, and its behavior offers clues to an answer.” (Newell and Simon 1976, p.

114). Thus, every information system instantiation can be treated as a proof by

construction (Nunamaker et al. 1991) in itself. For instance, a method for

modeling customer solutions can be shown to be implementable by presenting

software support for this method.

Each IT artifact needs to be designed carefully with respect to the characteristics of

the environment in which it is intended to function (Alexander 1970). For this

purpose, an analysis of the underlying outline and mechanisms that are present in the

environment is essential. However, such a field description can – by definition –

never be complete, since it is an abstraction of the complex reality that is made in

order to identify the crucial design parameters to be considered for the design of the

artifact (Alexander 1970). Notably, the purpose of designing an artifact, as envisioned

from a design science research standpoint, is utility, whereas the primary focus of

natural research is truth.

Various authors have proposed procedure models that can be used to systematize

the design process for IT artifacts. In this paper, we refer to the process proposed by

Peffers et al. (Peffers et al. 2008), since it builds on previous work in this area and

constitutes an integrated model based on these observations. The process model


features six stages, while four point of starting the design process are identified

(Peffers et al. 2008):

1. Identify problem and motivate: The research problem is identified and defined.

Also, the importance of solving the problem needs to be stated to justify the effort

for designing an appropriate solution.

2. Define objectives of a solution: The objectives of an appropriate solution for the

identified problem are developed. Objectives can be articulated qualitatively or

quantitatively. In the terminology of Hevner et al. (2004) this step corresponds to

identifying a business need that is perceived as important enough to start a design

process.

3. Design and development: IT artifacts are designed in a synthesis approach.

Consistent with Hevner et al. (2004), the design process can be informed and

shaped by preexisting IT artifacts and theories contained in the knowledge base.

4. Demonstration: The artifact is used to solve one or more instances of the problem

to demonstrate its effectiveness. This might include conducting activities such as

experiments, case studies, formal proofs, or simulations.

5. Evaluation: Compared to a demonstration, in an evaluation the artifact is subjected

to more comprehensive testing to indicate how well the artifact is able to solve the

identified problem. This can be done by comparing the results measured in the

demonstration with the identified design requirements, based on any appropriate

collection of data or based on a logical proof. Since the design of artifacts is

conceptualized as a search process (Hevner et al. 2004), the evaluation might

uncover the need to revise or extend the artifact in consecutive research.

6. Communication: The identified problem and also the properties, novelty, and

utility of the designed IT artifacts have to be effectively communicated. As this

might include scientific publications and communicating the results to

practitioners, the developed artifacts are intended to be transferred into the

knowledge base and also implemented and sustained in practical application.

Peffers et al. (2008) state that a design process can be started from multiple entry

points. A problem-centered approach is eligible to start by identifying a problem to be

solved and executing the entire process, which is recommended if the idea resulted

from an observation of the problem or from suggested further research in a previous

research project. In an objective-centered approach, the need for action is derived

from a research need or business need that can be addressed by designing an artifact.

In a design- and development-centered approach, the process is entered with

designing an artifact. This is in line with Gregor’s (2006, p.633) observation that the

construction of artifacts can „spring from inventiveness and imagination ahead of

good knowledge”. Peffers et al. (2008) outline, that this might include transferring an

artifact previously used to solve a similar problem into a new domain of application.

In a client-/context-initiated solution, the process is initiated by observing the

application of an artifact in its environment. This approach might be useful to re-

engineer the critical success factors embedded into the artifact for guiding the design

of new artifacts.


Identify Problem

& Motivate

Define problem

Show importance

Define Objectives

of a Solution

What would a

better artifact

accomplish?

Design &

Development

Artifact

Demonstration

Find suitable

context

Use artifact to

solve problem

Evaluation

Observe how

effective, efficient

Iterate back to

design

Communication

Scholarly

publications

Professional

publicationsIn

fere

nce

Th

eo

ry

Ho

w to

Kn

ow

led

ge

Me

trcis

, A

na

lysis

,

Kn

ow

led

ge

Dis

cip

line

Kn

ow

led

ge

Possible Research Entry Points

Problem-

Centered

Initiation

Objective-

Centered

Initiation

Design &

Development

Centered

Initiation

Client/ Context

Centered

Initiated

Process Interation

Nominal

Process

Sequence

Fig. 1. Design Science Research Process as proposed by Peffers et al. (2008)

3.3 Basic Premises of Theory Building

Theory building is the “classic” approach to developing knowledge about the world

with the “classic” set of research methods, such as experiments, surveys, and case

studies. It builds knowledge that is generalizable beyond individual cases. It informs

design processes as kernel theories.

Traditionally, several research methods are applied for theorizing. We provide a

brief overview here with a focus on studying emergence in order to outline the scope

of methods. The methods are introduced briefly and in their most basic form, while

they can be combined with each other in order to fit the needs of more complex

research endeavors.

Case study research: A case study is an empirical inquiry that investigates a

phenomenon within its real-life context. Case studies seem particularly useful if a

distinction between the phenomenon and its context cannot be made. They can be

involved with collecting multiple sources of evidence, such as qualitative and

quantitative data, at a single site or multiple sites (Yin 2003). Case study research

can be based on the assumption of ontological and epistemological realism (Yin

2003), and also on ontological idealism or constructivism (Walsham 1995). For

investigating emergence, this approach would be beneficial to investigate one case

of emergence related to a network, in order to develop rich insights based on

gathering and triangulating different types of quantitative and qualitative data. In

addition, a multiple case study could be conducted based on a previously set up

replication design, in order to make the findings better generalizable beyond the

investigated cases. Since we expect the emergence of IT-supported networks to be

a phenomenon that is intertwined into its socio-technical environment, case

studies represent a promising start for theorizing in this area.

Survey: Most surveys comprise a literature review to set up hypotheses that can be

empirically tested. Usually, data collected by means of sending out and receiving

questionnaires is analyzed with statistical data analysis techniques (such as

representative sampling, Kruskal and Mosteller (1979a; 1979b)). As a result,

based on the collected data the derived hypotheses are shown to be either

supported or not supported by the data. With respect to emergence, hypotheses

concerning the occurrence or consequences of emergence could be derived from


previous theory such as from a previously conducted case study. Consecutively,

quantitative data would have to be collected from many problem instances in

order to support or reject the developed hypotheses. This would help to further

generalize the findings.

Laboratory experiment: Laboratory experiments are studies that take place within

a designed and controlled environment (i.e. the lab) and „usually involve special

treatments of different groups to contrast the precise relationships among

variables” (Chen and Hirschheim 2004, p.206; Galliers 1991). As laboratory

experiments are analytical in nature, they can be used in the course of a design

science research project to evaluate IT artifacts. One approach is to specify two

groups of people, one of which applies the designed IT artifact to solve a problem,

whereas a second group of people is supplied with another artifact. As a result, the

performance of the designed artifact can be assessed by comparing the outcome

reached in each of the groups. However, a shortcoming of this approach is the

negligence of a real-world environment. It can be questioned if laboratory

experiments provide a means rich enough to study a complex development such as

emergence that is situated on a collective level of analysis and is so deeply

embedded into its organizational and technological environment.

Field experiment: Analogous to laboratory experiments, field experiments involve

special treatments to different groups, albeit they are conducted in real-world

settings. This might benefit the explanatory power of evaluations, as the IT artifact

can be evaluated with respect to its performance in the actual environment. With

respect to theorizing on emergence, field experiments might be difficult to

conduct, because the complexity of the phenomenon might render a definition of

homogeneous treatment groups and control groups inoperable. As a consequence,

other factors than those studied in the experiments might influence the results of

the study and might undermine their accuracy.

Grounded theory: The focus of grounded theory is developing theory from

systematically gathering and coding qualitative data (Glaser and Strauss 1967),

thereby identifying patterns in the data (Charmaz 2006). By analyzing these

patterns, researchers can build empirically valid theory (Urquhart and Fernández

2006). Accordingly, grounded theory is „an inductive, theory discovery

methodology that allows the researcher to develop a theoretical account in

empirical observations or data” (Martin and Turner 1986, p.141). Although

becoming more widespread in IS, grounded theory might still need to be adapted

to the specific need of considering the interface of people and technology that

constitutes IS research (Urquhart and Fernández 2006). With respect to

emergence, a grounded theory approach might shed light on the underlying

properties of the phenomenon without being subjected to bias from other theories.

3.4 Information Systems Research

DSR and natural research form comprehensive research cycles that complement each

other. In the light of design and emergence, this approach enables multi-disciplinary

research teams to work together in order to investigate the phenomenon in a

comprehensive way.


At the heart of this idea is to acknowledge the intimate embedding of the IT artifact

into its technological, organizational, and social environment (cf. Fig. 2). The design

of the IT artifact itself and the integration into its technological environment requires

to be carried out with respect to established design science techniques in order to

feature the desired functionality. In contrast, the analysis of the properties of the

organizational and social environments, as well as the nature of the interaction of the

IT artifact with these environments need to be theorized upon in natural research-like

projects. The social-technical nature of the entire system constitutes an important

property in this respect. While the IT artifact itself is a technological device, the

environment consists of other IT artifacts and technological devices, as well as of a

nested hierarchy of actors comprising such as people, groups of people, organizations,

networks of organizations, industries, ecosystems, and economies. On the most basic

level, these organizations are embedded into a system of cultural and social norms

that constitute the society in which the organizational units are situated.

Fig. 2. IT Artifact, as embedded into its socio-technical environment

Any attempt to investigate in interrelation of design and emergence in depth,

therefore, needs to combine research both paradigms in a complementary way. A

similar approach was formulated for the IS discipline by Hevner et al. (2004). They

stated that the development of the IT artifact must be informed from two directions

(cf. Fig. 3). On the one hand, design needs to be informed by theory that is contained

in the current body of knowledge. On the other hand, design must be informed by

exploring the environment into which the IT artifact is intended to function. Each step

of the design process itself needs to be justified or evaluated with information taken

from either side. After its development has been concluded, the IT artifact can be

embedded into the environment and tested. In addition, the knowledge encapsulated

Social Environment

Organizational Environment

Technological Environment

IT Artifact


by the IT artifact adds to the knowledge base as a new design theory (Gregor 2006).

Since there is not one best way to solve a problem with an artifact, design itself is

conceptualized as a search process (Hevner et al. 2004) that needs to be repeated until

a satisfactory degree of utility has been reached.

Environment IS Research Knowledge Base

People

Roles

Capabilities

Characteristics

Organizations

Strategies

Structure & Culture

Processes

Technology

Infrastructure

Applications

Communications

Architecture

Development

Capabilities

Develop/Build

Theories

Artifacts

Justify/Evaluate

Analytical

Case Study

Experimental

Field Study

Simulation

Assess Refine

Foundations

Theories

Frameworks

Instruments

Constructs

Models

Methods

Instantiations

Methodologies

Data Analysis

Techniques

Formalisms

Measures

Validation Criteria

Business

NeedsApplicable

Knowledge

Relevance Rigor

Application in the

Appropriate Environment

Additions to the

Knowledge Base

Fig. 3. Research Framework in Information Systems (Hevner et al. 2004)

4 Application scenarios

4.1 Examples for Supply Networks

There are many examples for traditional networks which have been coupled with

information technology. In fact, it would be hard to name networks without any IT

support. Roads have been equipped with sensors to allow for toll calculations and

might offer more functionality in the future, e.g. to avoid congestions. Similar

developments can be observed for other modes of transport. Another example are the

energy, water, and sewage systems. Together with IT systems, information can be

used to provide resources in the right amount at the right time so that they are used

efficiently. Two examples will be given in more detail in the following.

4.2 Smart Grids

According to (Pipattanasomporn et al. 2009) and the U.S. Department of Energy’s

Modern Grid Initiative (Modern Grid Initiative 2011) we define a smart grid as an

underlying network which “integrates advanced sensing technologies, control


methods and integrated communications into current electricity grid – both at

transmission and distribution levels.” (Pipattanasomporn et al. 2009, p. 1). The aim of

smart grids is to design more energy efficient and optimized energy provisioning

networks. In doing this, demand peaks within the energy grid shall be reduced and

load limits shall be avoided. In summary, the used electricity will not be less than at

the moment only through establishing a smart grid infrastructure. However, there will

be (mainly) cost advantages on both, the supplier’s and the customer’s side. While the

customers can buy their needed energy to off-peak times and, e.g., wash their clothes

automatically in the night, the suppliers, i.e. the common carriers and energy

producers, will be in the position to optimize the production of energy and reduce the

amount of electric circuits on the long term.

So what is necessary to evolve from the classical structure of energy delivery into a

smart grid structure which can be seen as a hybrid network as defined above which

means that the underlying ‘classical’ network moves to a new step, hybrid, IT-enabled

provision network? According to (Farhangi 2010) we propose three steps to flow into

a smart grid infrastructure (see Fig. 4).

Fig. 4. The Evolution of the Smart Grid (Farhangi 2010)

The basic state in the evolution of a smart grid is the use of electromechanical

meters and associated with this the introduction of automated meter reading (AMR)

systems in the electric circuits. These AMR allow for reading the consumption

records and further information of the customer’s side through the energy suppliers.

However, taking a return on investment point of view, a two-way automated

metering infrastructure (AMI) is necessary to really create a benefit because it will not

be possible for the energy providers to directly react on gathered information.

Through establishing AMI, a two-way communication will be established and not


only information fetching is possible. Desirable features like load management on the

customers’ side controlled by the energy suppliers and others allow for a higher

revenue on the one side and a beneficial buying and cost control on the other side.

The – for the moment – final logical step is to establish a complete smart grid as a

network of microgrids with distributed control. A permeating control that allows for

intelligent and efficient management of the smart grid has to include all components

and functions of the infrastructure.

Concluding on this procedure model, the movement from an undisputed classical

energy network as classical infrastructure into a smart grid is the establishment of an

IT-enabled infrastructure.

4.3 Logistics Networks

Logistics networks are based on several networks. The actual movement of goods

depends on traffic infrastructure and Supply Chain Management relies on an

institutional infrastructure of associated companies. Applying the infrastructure

perspective to logistics thus means to not only use the regular set of methods for

design, planning, and control that stem from an engineering point of view, but to also

consider theories which acknowledge that fact that complete control over the system

is not attainable.

Emergent effects can be observed in supply chains and have already been subject

to investigation. The best known phenomenon is commonly called the “bullwhip

effect”. It describes the interrelation of orders over multiple echelons. On each stage,

bundling, delay, etc. occurs and alters the order lot size for the next stage. Finally, a

constant demand from end customers may result in highly fluctuating orders at a

subordinate stage. The engineering approach to solving this problem fails as although

technically it would be easy to solve, actors have interests such as data privacy or

planning authority.

In their paper “Supply networks and complex adaptive systems: control versus

emergence” (Choi et al. 2001), Choi et al. formulate the thesis that supply networks

emerge rather than to be designed which implies that deterministic management can

only be efficient to a certain point. Consequently, they use complex adaptive systems

theory for their investigation. They give a concrete example for the interrelation of

design and emergence: “[…] when a buying firm develops one parts supplier as the

system supplier, this action in turn creates a whole new set of second-tier suppliers

[…]”. Therefore, their stated research goal is “[…] to develop a strategy on how much

of the SN to control and how much of it to let emerge”. After conceptualizing supply

networks as complex adaptive networks, the authors propose a tool combining

“control theoretic, agent-based, and discrete-event modeling approaches” to model the

network’s behavior and answer the question of determining the right degree of

freedom for emergence. They also predict that emergence patterns can be identified

by means of case studies and that these can be used for managing emergence.

Holweg and Pil (Holweg and Pil 2008) wrote on “Theoretical perspectives on the

coordination of supply chains”. They examine the changes in information flows at

three car manufacturers before and after the introduction of built-to-order production.

Methodically, they employ resource-based view (which will be omitted here),

complex adaptive systems, and adaptive structuration theory. Their main concern is in


how far technology is being used in the intended way. As in the automotive industry

the OEM is by far the most powerful actor, its environment adapts to its requirements

which is the desired setting for analyzing the interrelation of design and emergence.

CAS considers the entire system. The observation of changes and adjustments may

help to explain why technologies suffer deviations from their original purpose. The

aspects which are omitted from design are relevant, too, as mechanisms will emerge

in their environment to fill those gaps. Thus, e.g. buffers may result if the information

system does not provide sufficient data for planning. This is how emergent

phenomena regarding the missing design of a system can be explained. In literature,

there has already been mentioned that entire supply chains are based on emergent

effects. Last, the AST-perspective is considered. It serves as a framework for the

analysis of the interaction of intended behavior for information systems and the

effectively emerging structures. Taking the duality as an assumption, it provides a

basis to explain why it might be difficult to achieve certain effects in supply chains by

information systems. Nevertheless, the observations do not coincide completely with

the predictions. The authors attribute that to the quite unequal distribution of power

within supply chains. An emphasis is put on the fact that information systems are

used for several tasks which they had not been intended for. In a last step, the authors

elaborate that IT infrastructure by itself is unable to lead to any improvements. It is

rather to be seen in combination with the user so that new theories are needed.

In “Supply-chain networks: a complex adaptive systems perspective”, the authors

Surana, Kumara, Greaves, and Raghavan (Surana et al. 2005) state that supply chains

were driven to such a high level of complexity by IT that it almost reaches the one of

biological systems. However, there is a lack of coordination strategies that lead to an

adaptive, flexible and coherently collective attitude within supply chains. This is due

to a missing understanding of the principles of the formation and further development

of supply chains including their complex organizational structure and characteristics.

Therefore, supply chains are treated as CAS to allow the application of existing

concepts, tools and models to them. Referring to Choi et al., the authors take up the

question of what part is to be controlled within the network and how much emergence

is to be accepted. According to them, there is no framework to verify and generalize

this question. Nevertheless, behavior patterns can be identified and analyzed by

methods of complexity theory. The emergence of highly-structured collective attitude,

which arises from the interaction within the subsystems, is depicted as their most

important phenomena. Evolution in biological systems is contrasted with

configuration in artificially designed systems. Both systems have in common that they

develop protocols to achieve robustness via hierarchic organization. Complexity,

which is introduced to achieve robustness, can become a vulnerable element itself and

gives reasons for the “robustness/complexity/fragility spiral”. Regarding information

systems it is noted that the main obstacle in their use is caused by a lack of

understanding of the organizational, functional and evolutionary principles of the

supply chain. The proposed solution is considering them as CAS. After that, different

techniques for modeling and analysis are presented and transferred. In a last step, it is

stated that IT infrastructure partly made it possible to operate supply chains across the

whole network. However, the complexity of those nets is an obstacle and the existing

tools and techniques do not fulfill the requirements to achieve that goal.


In their paper “Structuration theory: its potential impact on logistics research”,

Lewis and Suchan (Lewis and Suchan 2003) claim that logisticians have to

understand better the effects on behavior and management caused by information

systems. The effects describe how individuals, groups, departments or corporations

interpret, implement, reject and use information systems. Therefore, structuration

theory is introduced as a method. First of all, it is mentioned why a process-based

theory is necessary in logistics. For this purpose, the contrast between positivism and

interpretivism is highlighted. They bring forward the argument that positivistic

approaches are limited regarding the understanding of organizations. That is why

methods are needed to understand the supply chain’s characteristics using the

subjective experiences of the members of a supply chain, their interpretation and their

subsequent actions. The complex relations within a supply chain and the increased

absorption, analysis and processing of information on physical goods have made that

even more necessary. Following the predominant positivistic approach this normally

takes place in the decomposition of the system and the composition of artifacts to

control a certain section. However, the solutions simplify the complexity of the

interactions and the behavior too much. This is why, referring to Choi et al., the

authors claim to regard supply chains as emergent CAS. Thus, the numerous

interdependencies between a corporation and its designed environment are made a

focal point. According to them, that is the only way to explain the emerging

phenomena in networks. Afterwards, the structuration theory is introduced which is

mainly based on the duality of the structure. Structures control the individual behavior

which in return influences the structures. The introduction of information systems can

cause very different reactions within different structures even if the range of functions

is the same. This is why those systems have to be seen in their context. An artifact has

basically the characteristics that were intended by the designer. The interpretation of

them highly depends on the interpretation of the user which may change over the

years. Nevertheless, at any moment, this interpretation influences the structure of the

organization which in return is interdependent with the behavior.

5 Research Directions

IT-enabled networks raise numerous interesting multi-disciplinary research questions.

Only if researchers from these disciplines work together, a comprehensive theory of

hybrid networks can be devised. As the informatization of our business and private

lives continues, this issue becomes increasingly important in order to keep track of

and steer the ongoing developments.

In terms of management, two major questions appear. First, what new business

models develop? And second, how can hybrid networks be set up, maintained, and

marketed? New business models are closely related to the possibilities offered by

information systems and their ability to collect and process huge amounts of data. For

example, service plans can be customized based on the consumer’s actual usage. In

addition, more business functions can be outsourced, e.g. cloud computing can serve

as an infrastructure for business processes. Most important, however, is the decision

of how much development to let emerge over time and how much to design. The

question is if emergence can be influenced. If too much is left to emerge, control over


the business might be lost, but if too much control is exerted, the status of an

infrastructure might never be attained. For example, Google seems to strictly enforce

the usage of real names on their social network Google+ whereas Facebook is

generally indifferent about this issue. It will be seen what the results of these

decisions are. Closely related is the marketing perspective and the question of how to

set standards. The possibility of losing control over a network arises when the state of

a monopoly is reached. Infrastructures are inclined to be monopolies, but antitrust

interests of the government and the society exist. Finally, companies might be forced

to respect these interests which are against their own. This happened to American

Airlines and their Sabre flight reservation system when the US government outlawed

the preference of American Airlines flights. Microsoft was under investigation for

bundling their operating system Windows, i.e. the de facto infrastructure for software

applications, with the Internet Explorer browser software. The same might happen in

the future to Google while they become the infrastructural backbone for all Internet

applications.

For law, the question when something becomes an infrastructure is an interesting

topic. An example is broadband Internet access. While it was virtually non-existent at

private homes only ten years ago, governments today discuss whether it should

become a right to have Internet access as the importance for daily live is growing and

not having Internet access might result in excluding somebody from society.

Information systems face the challenge of designing solutions for problems which

change with the introduction of the solution. The traditional way of analyzing an

isolated problem and designing a solution is obsolete and no longer successful. New

ways to devise artifacts have to be found and validated.

Finally, researchers have argued that we develop an “infrastructure literacy” based

on what we know from experience (Edwards 2003). This of course has most often a

western bias, so that a comparison and transfer of this research to the situation in

emerging countries might offer new insights.

6 Concluding Remarks

The boundaries between traditional supply networks and information systems are

becoming increasingly unrecognizable. They coalesce to form hybrid networks where

information flows are integrated and control can be exerted digitally. Due to the

amount of data which is instantly available and the sophisticated means to monitor

and regulate these networks, numerous new business models and ways to efficiently

organize operations in the networks become possible. However, networks are not

controlled by a single instance and they, especially their information systems parts,

cannot be designed like usual systems. New ways to analyze and steer emergent

properties are required to take full advantage of hybrid networks. Therefore, methods

from engineering have to be extended by more constructivist-influenced theories. This

paper gave some examples and outlined possible research directions.


7 References

Alexander, C. Notes on the Synthesis of Form, Harvard University Press, Cambridge, 1970.

Becker, J. and Delfmann, P. Reference Modeling. Efficient Information Systems Design

Through Reuse of Information Models, Physica, Berlin, Germany, 2007.

Charmaz, K. Constructing Grounded Theory: A practical guide through qualitative analysis,

Sage Publications, Thousand Oaks, CA, USA, 2006.

Chen, W. and Hirschheim, R. "A paradigmatic and methodological examination of information

systems research from 1991 to 2001," Information Systems Journal (14:3), 2004, pp. 197-

235.

Choi, T. Y., Dooley, K. J., & Rungtusanatham, M.: Supply networks and complex adaptive

systems: control versus emergence. Journal of Operations Management, 19(3), 351-366,

2001.

Dahlbom, B. and Janlert, L. E.: Computer Future. Manuscript. 1996.

Edwards, Paul: Infrastructure and Modernity: Force, Time and Social Organization in the

History of Sociotechnical Systems. In T. J. Misa, P. Brey and A. Feenberg (eds.) Modernity

and Technology, Cambridge, Ma: MIT Press. 185-225, 2003.

Farhangi, H.: The path of the smart grid. Power and Energy Magazine, IEEE, 2010.

Galliers, R.D. "Choosing appropriate information systems research approaches: a revised

taxonomy," in Information Systems Research: Contemporary Approaches and Emergent

Traditions, H.-E. Nissen, H.K. Klein, and R. Hirschheim (eds.), Elsevier Science Publishers,

North Holland, 1991, pp. 327-345.

Glaser, B.G. and Strauss, A.L. The discovery of grounded theory: Strategies for qualitative

research, Aldine Publishing Company, New York, NY, USA, 1967.

Gregor, S. "The Nature of Theory in Information Systems," MIS Quarterly (30:3), 2006, pp.

611-642.

Gregor, S.; Jones, D. “The Anatomy of a Design Theory” in Journal of the Association for

Information Systems, 8, 5, 2007, pp. 312-335.

Hanseth, O.: From systems and tools to networks and infrastructures – from design to

cultivation. Towards a design theory of information infrastructures. In H. Holmström, M.

Wiberg & A. Lund (Eds.), Industrial Informatics design, use and innovation:perspectives

and services. Hershey: IGI Global, 2010.

Hevner, A.R., March, S.T., Park, J., and Ram, S. "Design Science in Information Systems

Research," MIS Quarterly (28:1), 2004, pp. 75-105.

Holweg, M., & Pil, F.: Theoretical perspectives on the coordination of supply chains. Journal of

Operations Management, 26(3), 389-406, 2008.

Jochimsen, R.: Theorie der Infrastruktur. J. C. B. Mohr, Thüringen, 1966.

Kruskal, W. and Mosteller, F. "Representative Sampling 2, Scientific Literature, Excluding

Statistics," International Statistical Review (47:2), 1979a, pp. 111-127.

Kruskal, W. and Mosteller, F. "Representative Sampling 3: The Current Statistical Literature,"

International Statistical Review (47:3), 1979b, pp. 245-265.

Lewis, I., & Suchan, J.: Structuration theory: its potential impact on logistics research.

International Journal of Physical Distribution & Logistics Management, 33(4), 296-315,

2003.

March, S.T. and Smith, G.F. "Design and natural science research on information technology,"

Decision Support Systems (15:4), 1995, pp. 179-212.

Martin, P.Y. and Turner, B.A. "Grounded theory and organizational research," The Journal of

Applied Behavioral Science (22:2), 1986, pp. 141-157.

Modern grid initiative: http://www.netl.doe.gov/moderngrid/, Accessed 2011-11-10

Newell, A. and Simon, H.A. "Computer Science as empirical inquiry: Symbols and search,"

Communications of the ACM (19:3), 1976, pp. 113-126.


Nunamaker, J., Chen, M., and Purdin, T.D.M. "Systems development in information systems

research," Journal of Management Information Systems (7:3), 1991, pp. 89-106.

Peffers, K., Tuunanen, T., Rothenberger, M.A., and Chatterjee, A.S. "A Design Science

Research Methodology for Information Systems Research," Journal of Management

Information Systems (24:3), 2008, pp. 45-77.

Pipattanasomporn, M., Feroze, H., Rahman, S.: Multi-agent systems in a distributed smart grid:

Design and implementation. Power Systems Conference and Exposition, 2009. PSCE ’09.

IEEE/PES.

Simon, H.A. The Sciences of the Artificial, MIT Press, Cambridge, MA, USA, 1996.

Surana, A., Kumara, S., Greaves, M., & Raghavan, U. N.: Supply-chain networks: a complex

adaptive systems perspective. International Journal of Production Research, 43(20), 4235-

4265, 2005.

Urquhart, C. and Fernández, W. "Grounded theory method: The researcher as blank slate and

other myths," in Proceedings of the 27th International Conference on Information Systems,

Milwaukee, WI, USA, 2006.

Van Laak, D.: Der Begriff „Infrastruktur“ und was er vor seiner Erfindung besagte. Archiv für

Begriffsgeschichte. Nr. 41 (1999), pp. 280-299.

Vignes, J.: Die wesentlichen Probleme der Infrastruktur in Französisch-West- und Äquatorial-

Afrika. Investitionen, Infrastruktur und Kapitalbedarf. Erwartungen im Schwarzen Erdteil.

Schriftenreihe der Deutschen Afrika-Gesellschaft, 6, 20. Bonn, 1958.

Walsham, G. "Interpretive case studies in IS research: Nature and method," European Journal

of Information Systems (4:2), 1995, pp. 74-81.

Yin, R.K. Case Study Research. Design and Methods, Sage Publications, Thousand Oaks, CA,

USA, 2003.

Documents

On the Coalescence of Supply Networks and Information ... › uploads › media... · networks, e.g. the electricity networks or road networks, and information systems. From a business