IMPACT OF PROCESS INNOVATION ON ENTERPRISE
NETWORKS FOR COMPETENCES EXCHANGE
E³, a Multi Agent based Model
Marco Remondino, Marco Pironti and Paola Pisano
e-Business L@B, University of Turin, Torino, Italy
Keywords: Process Innovation, Network Topology, Business Process, Agent Based Simulation.
Abstract: A business process is a set of logically related tasks performed to achieve a defined business and related to
improving organizational processes. A process innovation can happen at various levels: incremental way,
redesign of existing processes, totally new processes. The knowledge behind a process innovation can be
shared, acquired, changed and increased by the enterprises inside a network. An enterprise can decide to
exploit the innovative process it owns, thus potentially gaining competitive advantage, but risking, in turn,
that other players could reach the same technological level. Or it could decide to share it, in exchange for
other competencies or money. These activities could be the basis for a network formation and/or impact the
topology of an existing network, by changing the number and topology of ties and links. In the present work
an agent based model is introduced (E³), that aims to explore how a process innovation can facilitate
network formation of existing enterprises, affect the network topology (e.g.: an enterprise owning an
innovative process could become a focal point), induce new players to enter the market and spread onto the
network by being shared or internally acquired by new players.
1 INTRODUCTION
Unlike product innovation, which is targeted
towards product engineering, development and
commercialization activities, process innovation
relates to improving organizational processes. Our
understandings of business process innovation come
from the growing researches on organizational
learning and knowledge management. The transfer
and sharing of process innovation is not easy to
attain, but information sharing/knowledge transfer
(both within and across the boundary of the
organization) is seen as an essential element for
innovation. The network promote not only the
transfer of knowledge (and the possible transfer of
process) but also the creation of new knowledge as
well, through synergies or competition. Within an
organization, cross-unit knowledge transfer can
produce “creative abrasion” (Leonard-Barton, 1995),
generate “improvisational sparks” (Brown and
Duguid, 1991) and create new information patterns
by rearranging information already in use and
incorporating information previously neglected
(Isabella, 1990; Macdonald, 1995). Enterprises also
actively look for external knowledge, for example
by expanding their networks to learn new practices
and technologies (Kogut, 1988). The process
innovation could impact on the network not only by
improving the knowledge of the involved
enterprises, but also by changing the number of
actors (exit and entry), and changing the numbers
and patterns of link information (Koka, 2006). The
network can expand, churn, strengthen or shrink. At
the level of a single enterprise, if it is the only one
(or among the few) possessing an innovative
process, it could become the focal point in a
network, attracting others, wishing to link with it.
Each network change is brought about by specific
combination of changes in tie creation, tie deletion,
and by changes in an actor’s portfolio size (number
of link) and portfolio range (numbers of partners)
(Koka et al. 2006). While Koka et al. (2006) present
four types of network changes, they find that only an
expanding network and a churning network are a
reflection of a structural change, because new
alliances are formed with new partners. An
expanding network is brought about by an increase
of new alliances without a deletion of old ones
(meaning a large average of portfolio), together with
an increasing portfolio range (more difference in
167
Remondino M., Pironti M. and Pisano P. (2009).
IMPACT OF PROCESS INNOVATION ON ENTERPRISE NETWORKS FOR COMPETENCES EXCHANGE - E
3
, a Multi Agent based Model.
In Proceedings of the International Conference on e-Business, pages 167-174
DOI: 10.5220/0002233001670174
Copyright
c
SciTePress
partners). A churning network reflects the formation
of new alliances and the deletion of existing
alliances. While the average portfolio remains stable
in term of the number of partners, there is a rotation
of partners.
In order to empirically study how process
innovation can affect an enterprise network, an agent
based model is used. Agent based simulation is an
effective paradigm for studying complex systems. It
allows the creation of virtual societies, in which each
agent can interact with others basing on certain
rules. In this way, a social system can be observed as
if it were a laboratory study, by repeating the
experiments all the needed times, and changing just
some parameters, by leaving all the others still
(coeteris paribus analysis), something that would be
impossible in the real system. The agents are basic
entities, endowed with the capacity of performing
certain actions, and with certain variables defining
their state. In the model presented here, the agents
are reactive, meaning that they simply react to the
stimuli coming from the environment and from other
agents, without cognitively elaborating their own
strategies. An agent based model consists of a
multitude of software agents (both homogeneous or
heterogeneous), each type being endowed with
particular local properties and rules, put together
within an environment, formally described as a set
of parameters and rules. When the model is formally
built and implemented, emergent results can be
observed, thus inferring cause-effect relations by
simulating different core scenarios.
In the present work, social network theory briefly
is analyzed and a definition of process innovation is
given. Then, the comprehensive agent based model
used is formally introduced, and it is discussed how
it can be employed to study how a process
innovation affects an enterprise network. Last, some
empirical results coming from the model are given
and the future work in this direction is discussed.
2 SOCIAL NETWORKS
A social network is a social structure made of nodes
(which are generally individuals or organizations)
that are tied by one or more specific types of
interdependency, such as values, visions, ideas,
financial exchange, friendship. Social network
analysis views social relationships in terms of nodes
and ties. Nodes are the individual actors within the
networks, and ties are the relationships between the
actors. These concepts are often displayed in a social
network diagram, where nodes are the points and
ties are the lines.
The idea of drawing a picture (called a
sociogram”) of who is connected to whom for a
specific set of people is credited to Dr. J.L. Moreno
(1934), an early social psychologist who envisioned
mapping the entire population of New York City.
Cultural anthropologists independently invented the
notion of social networks to provide a new way to
think about social structure and the concepts of role
and position (Nadel, 1957; Mitchell 1969), an
approach that culminated in rigorous algebraic
treatments of kinship systems (White, 1963; Boyd,
1969). At the same time, in mathematics, the nascent
field of graph theory (Harary, 1969) began to grow
rapidly, providing the underpinnings for the
analytical techniques of modern social network
analysis. The strategic network perspective avers
that the embeddedness of enterprises in networks of
external relationships with other organizations holds
significant implications for enterprise performance
(Gulati, Nohria, and Zaheer, 2000).
Specifically, since resources and capabilities
such as access to diverse knowledge (Burt, 1992),
pooled resources and cooperation (Uzzi, 1996), are
often acquired through networks of inter-firm ties,
and since access to such resources and capabilities
influences enterprise performance (Mowery, Oxley,
and Silverman, 1996), it is important from a strategy
perspective to examine the effect of network
structure on enterprise performance (Gulati et al.,
2000). Relationships between enterprises and their
partners affect enterprises’ alliance-building,
behaviour and performance (Ahuja, 2000; Almeida,
Dokko, & Rosenkopf, 2003; Powell, Koput, Smith-
Doerr, & Owen- Smith, 1999; Stuart, 2000). There is
evidence that enterprises’ network positions have an
impact on their survival (Baum, Calabrese, &
Silverman, 2000), innovativeness (Ahuja, 2000),
market share (Shipilov, 2005), and financial returns
(Rowley, Behrens, & Krackhardt, 2000). However,
evidence remains mixed on which particular patterns
of inter-organizational relationships are
advantageous for enterprises. One of the key ideas
currently dominating the literature is Burt’s (1992)
open network perspective, according to which an
enterprise can obtain important performance
advantages when exploiting relationships to partners
that do not maintain direct ties among one another.
The absence of direct ties among a firm’s partners
(the presence of structural holes) indicates that these
partners are located in different parts of an industry
network, that they are connected to heterogeneous
sources of information, and that their invitations to
ICE-B 2009 - International Conference on E-business
168
jointly exploit business prospects present the focal
enterprise with access to diverse deal-making
opportunities (McEvily & Zaheer, 1999). Several
studies have shown that enterprises improve their
performance as a result of maintaining relationships
(e.g., Finlay & Coverdill, 2000; Hargadon & Sutton,
1997), whereas other studies have shown negative
performance effects of firms’ maintaining positions
in open networks (e.g., Ahuja, 2000; Dyer &
Nobeoka, 2000).
3 PROCESS INNOVATION
A business process is a set of logically related tasks
performed to achieve a defined business outcome
(Davenport and Short, 1990), e.g.: sequencing of
work routines, information flow and so on.
Process innovation is defined as “the
introduction of a new method of production, that is,
one yet tested by experience in the branch of
manufacture concerned a new way of handling a
commodity commercially” (Shumpeter 1911).
Archibugi et al. (1994) and Edquist et al. (2001)
define process innovation like the result in a
decrease in the cost of production. The drives of
process innovation are primarily reduction in
delivery lead time, lowering of operational costs,
and increase in flexibility (Boer and During 2001):
process innovations are a firm’s new way of design
or manufacturing existing or new products. While
newness on product innovation is defined at a macro
level (market, industry), newness of process
innovations is often defined at a micro level
(enterprise and business unit).
Meeus and Euist divide process innovations into
two categories: technological and organizational
innovations: technological process innovations
change the way products are produced by
introducing change in technology (physical
equipment, techniques, system); organizational
innovations are innovations in an organization’s
structure, strategy and administrative processes
(Damanpour 1987).
Process innovation can and should happen at
various levels within the organization as no
organization can depend solely upon innovation
occurring at one level only. Successful organizations
have an innovation process working its way through
all levels of the organization.
4 IMPACT ON THE NETWORK
Process innovation is a key factor for both
competing in a market and creating links with other
players. An enterprise owning a proprietary process
would in fact exploit it, by gaining a competitive
factor over those who do not possess it. On the other
hand, it could decide to share it with other
enterprises in exchange for money or, even better, in
exchange for other competencies it does not know.
This is the most important factor behind the creation
of what we here define “network for competences
exchange”, i.e.: a social network of enterprises,
where the ties semantically represent a synergy
among players exchanging process innovations or, to
a more general extent, competences.
Philippen and Riccaboni (2007) in their work on
“radical innovation and network evolution” focus on
the importance of local link formation and the
process of distant link formation. Regarding the
formation of new linkages Gulati (1995) finds that
the process of new tie creation is heavily embedded
in an actor’s existing network. This means that new
ties are often formed with prior partners or with
partners of prior partners, indicating network growth
to be a local process. Particularly when considering
inter-firm alliances, new link formation is
considered “risky business” and actors prefer
alliances that are embedded in a dense clique were
norms are more likely to be enforceable and
opportunistic behaviour to be punished (Gulati,
1995; Powell et al., 1996; Koka et al., 2006,
Granovetter, 1985). Distant link formation implies
that new linkages are created with partners whom
are not known to the existing partners of an actor. At
the level of the enterprise, (Burt 1992) shows that
distant linkage that serve as bridge between dense
local clique of enterprises, can provide access to new
source of information and favourable strategic
negotiation position, which improves the firms’
position in the network and industry.
In order to examine and study how a process
innovation can spread and affect the network for
competences exchange, an agent based model is
used. The model is a comprehensive one, showing
the network dynamics for enterprises, and is
described in detail in the next paragraph.
5 THE E³ AGENT BASED MODEL
The model has been developed at the e-Business
L@B, University of Turin. It is built in pure Java,
thus following the Object Oriented paradigm. This is
IMPACT OF PROCESS INNOVATION ON ENTERPRISE NETWORKS FOR COMPETENCES EXCHANGE - E³, a
Multi Agent based Model
169
particularly suitable for agent based modelling, since
the individual agents can be seen as objects coming
from a prototypal class, interacting among them
basing on the internal rules (methods). While the
reactive nature of the agents may seem a limitation,
it’s indeed a way to keep track of the aggregate
behaviour of a large number of entities acting in the
same system at the same time. All the numerical
parameters can be decided at the beginning of each
simulation (e.g.: number of enterprises, and so on).
Everything in the model is seen as an agent; thus
we have three kinds of agents: Environment,
Enterprises and Emissaries (E³). This is done since
each of them, even the environment, is endowed
with some actions to perform.
5.1 Heat Metaphor
In order to represent the advantage of an enterprise
in owning different competences, the “heat”
metaphor is introduced. In agent based models for
Economics, the metaphor based approach
(Remondino, 2003) is an established way of
representing real phenomena through computational
and physical metaphors. In this case, a quantum of
heat is assigned for each competence at each
simulation turn. If the competence is internal (i.e.:
developed by the enterprise) this value is higher. If
the competence is external (i.e.: borrowed from
another enterprise) this value is lower. This is
realistic, since in the model we don’t have any form
of variable cost for competencies, and thus an
internal competence is rewarded more. Heat is thus a
metaphor not only for the profit that an enterprise
can derive from owning many competences, but also
for the managing and synergic part (e.g.: economy
of scale).
Heat is also expendable in the process of creating
new internal competences (internal exploration) and
of looking for partner with whom to share them in
exchange of external competences (external
exploration). At each time-step, a part of the heat is
scattered (this can be regarded as a set of costs for
the enterprise). If the individual heat gets under a
threshold, the enterprise ceases its activity and
disappears from the environment.
At an aggregate level, average environmental
heat is a good and synthetic measure to monitor the
state of the system.
5.2 Environment
The environment is regarded as a meta-agent,
representing the world in which the proper agents
act. It’s considered an agent itself, since it can
perform some actions on the others and on the heat.
If features the following properties: a grid (X,Y),
i.e.: a lattice in the form of a matrix, containing
cells; a dispersion value, i.e.: a real number used to
calculate the dissipated heat at each step; the heat
threshold under which an enterprise ceases; a value
defining the infrastructure level and quality; a
threshold over which new enterprises are introduced;
a function polling the average heat (of the whole
grid). The environment affects the heat dispersion
over the grid and, based on the parameter described
above, allows new enterprises to join the world.
5.3 Enterprise Agents
This is the most important and central type of agent
in the model. Its behaviour is based on the reactive
paradigm, i.e.: stimulus-reaction. The goal for these
agents is that of surviving in the environment (i.e.:
never go under the minimum allowed heat
threshold). They are endowed with a heat level
(energy) that will be consumed when performing
actions. They feature a unique ID, a coordinate
system (to track their position on the lattice), and a
real number identifying the heat they own. The most
important feature of the enterprise agent is a matrix
identifying which competences (processes) it can
dispose of. In the first row, each position of the
vector identifies a specific competence, and is equal
to 1, if disposed of, or to 0 if lacking. A second row
is used to identify internal competences or
outsourced ones (in that case, the ID of the lender is
memorized). A third row is used to store a value to
identify the owned competences developed after a
phase of internal exploration, to distinguish them
from those possessed from the beginning. Besides,
an enterprise can be “settled”, or “not settled”,
meaning that it joined the world, but is still looking
for the best position on the territory through its
emissary. The enterprise features a wired original
behaviour: internally or externally explorative. This
is the default behaviour, the one with which an
enterprise is born, but it can be changed under
certain circumstances. This means that an enterprise
can be naturally oriented to internal explorative
strategy (preferring to develop new processes
internally), but can act the opposite way, if it
considers it can be more convenient. While in the
present model the agents are stochastic (with a
different probability distribution decided at the
beginning of the simulation for the two agents’
classes), cognitive agents will be added shortly,
ICE-B 2009 - International Conference on E-business
170
using reinforcement learning techniques to optimize
their behaviour and make it more realistic.
Finally, the enterprise keeps track of its
collaborators (i.e.: the list of enterprise with whom it
is exchanging competencies and making synergies)
and has a parameters defining the minimum number
of competencies it expects to find, in order to form a
joint. The main goal for each enterprise is that of
acquiring competences, both through internal (e.g.:
research and development) and external exploration
(e.g.: forming new links with other enterprises). The
enterprises are rewarded with heat based on the
number of competences they possess (different,
parameterized weights for internal or external ones),
that is spread in the surrounding territory, thus
slowly evaporating, and is used for internal and
external exploration tasks.
5.4 Emissary Agents
These are agents that strictly belong to the
enterprises, and are to be seen as probes able to
move on the territory and detect information about
it. They are used in two different situation: 1) if the
enterprise is not settled yet (just appeared on the
territory) it’s sent out to find the best place where to
settle. 2) If the enterprise is already settled and
chooses to explore externally, an emissary is sent out
to find the best possible partners. In both cases, the
emissary, that has a field of vision limited to the
surrounding 8 cells, probes the territory for heat and
moves following the hottest cells. When it finds an
enterprise in a cell, it probes its competencies and
compares them to those possessed by its chief
enterprise verifying if these are a good complement
(according to the parameter described in the
previous section). In the first case, the enterprise is
settled in a cell which is near the best enterprise
found during the movement. In the second case, the
enterprise asks the best found for collaboration). A
link is created among two enterprises if at least one
competence may be exchanged among them. Be
CM(a) the competences missed by enterprise a, and
CM(b) those missed by enterprise b, the exchanged
number of competences will be the minimum
between CM(a) and CM(b). The strength of the link
among two enterprises will be proportional to the
number of exchanged competences, and will vary
during the simulation (e.g.: after enterprise a
acquires a new competence that b is missing and
vice versa). While moving, the emissary consumes a
quantum of heat, that is directly dependant on the
quality of infrastructures of the environment.
5.5 Main Iterations
The main iterations for the simulation model are
described in this section.
At step 0, a lattice is created (X, Y). A number n
of enterprises are created, k of them internally
explorative and n-k of them externally explorative.
X, Y, n, and k are set by the user, before the
simulation starts.
At step 1, the environment checks if some
enterprise reached the minimum heat threshold; if
so, removes it from the world. After that, each
enterprise, if idle (not doing anything) decides what
behaviour to follow.
At step 2, all the enterprises that selected to be
EE move their emissary by one cell. All the IE ones
work on the R&D cycle (one step at a time).
At step 3, the EE enterprises check if the
emissary finished its energy and, in that case, ask the
best found enterprise for collaboration (they can
receive a positive or negative reply, based on the
needs of the other enterprise). The IE enterprises
check if R&D process is finished and, in that case,
get a competence in a random position (that can be
already occupied by an owned competences, thus
wasting the work done).
At step 4, the environment scatters the heat
according to its parameters. Loop from step 1.
5.6 Parameters in the Model
At the beginning of a simulation, the user can
change the core parameters, in order to create a
particular scenario to study. Some of the parameters
are constituted by a scalar value, others are in
percentage, others are used to define stochastic
(normal) distributions, given their mean value and
their variance. Here follows a synthetic explanation
for the individual parameters:
Maximum number of steps: is the number of
iterations in the model. 0 sets the unbounded mode
Initial number of enterprises: is the number of
enterprise agents present at start-up (0 is random)
Initial heat for enterprise: a normal distribution
setting the initial energy for each enterprise, given
the mean and the variance
Number of competences: the length of the vector,
equal for all the enterprises (metaphorically
representing the complexity of the sector in which
they operate)
Competences possessed at start-up: a normal
distribution referring to how many processes an
enterprise owns internally, given the mean and the
variance
IMPACT OF PROCESS INNOVATION ON ENTERPRISE NETWORKS FOR COMPETENCES EXCHANGE - E³, a
Multi Agent based Model
171
Threshold for new enterprise to enter the market:
a delta in the average heat of the world, after which
a new enterprise is attracted in the market
Infrastructure quality: affects the cost of external
exploration
Minimum heat threshold: level under which an
enterprise cease
Minimum percentage of competences to share for
link creation: when asked for a competences
exchange, the other enterprise looks at this value to
decide whether to create a link or not
Emissary step cost: percentage of the heat
possessed by the enterprise spent for each step of its
emissary, during external exploration task
Internal exploration duration: quantity of steps
for internally developing a new competence
Internal exploration cost: percentage of the heat
possessed by the enterprise spent for each step of
internal exploration
Environment control cycles: quantity of steps for
sampling the average heat of the environment
Heat dispersion index: percentage of heat
evaporated at each step
Lattice dimension: the dimension of the grid
hosting the enterprise (i.e.: the whole environment)
Internal Exploration cost: una tantum cost for
setting up an emissary for external exploration
Propensity to External Exploration for new
enterprises: when a new enterprise enters the
market, it looks at the average number of links in the
network. If more than this value, it behaves as
externally explorative, otherwise internally
explorative
Number of initial enterprises doing external
exploration: variable to divide the initial behaviour
Value of internal/external competence: reward
(heat) given for each internal/external competence
possessed
6 QUALITATIVE RESULTS
While the main object of this paper is to present the
model itself as a tool for studying the effects of
process innovation on enterprise networks, in the
present paragraph some insights will be given about
preliminary results obtained from the model itself.
The presented ones will be mainly qualitative
results, although the model can give many
quantitative individual and aggregate results. In
particular, a “computational only” mode is present in
the model, allowing it to perform a multi-run batch
execution. This is done according to the theory
presented in Remondino and Correndo (2006): the
model is executed a defined number of times
(chosen by the user) and the different outputs are
sampled and collected at every n steps (again, n
decidable by the user) with the same parameters (in
order to overcome sampling effects that could be
caused by stochastic distributions) or by changing
one parameter at a time by a discrete step, in order to
carry on a coeteris paribus analysis on the model.
While this kind of analysis will be discussed in
detail in future works, here some qualitative and
semi-quantitative outputs will be discussed, obtained
from the model. The model can give the following
different kinds of outputs, when running innormal
mode: 1) a real-time graph, depicting the social
network, in which the nodes are the enterprises,
whose colour represent the behaviour they are
following at a given step, and the links are the ties
indicating two or more enterprises mutually
exchanging one or more competences. 2) A set of
charts, showing in real time some core parameters,
namely: average heat in the environment, number of
links (in the network), number of links (average),
number of enterprises doing internal exploration,
number of ceased enterprises since the beginning,
number of born enterprises since the beginning,
number of available competences (overall), total
number of skills possessed at the beginning,
obtained by external exploration, obtained by
internal exploration.
In figures 1, 2 and 3, the output graph is depicted
at times 0 (no links), 100 and 500. These pictures
belong to the same simulation, so the parameters are
the same for all of them, with the only variation of
time, giving a hint about the development of the
enterprise network. In figure 1 the initial state of the
network is shown, where no ties have been created,
yet. A total of 20 enterprises is on the territory, 10 of
which have an internally explorative behaviour and
the other 10 have an externally explorative mood.
Internal competences are rewarded 10% more than
external ones, but internal exploration strategy (e.g.:
research & development) is 30% more expensive.
After 100 steps (figure 2) some new players have
entered the market (an average of 1 new enterprise
each 10 steps), meaning that the average heat of the
system increased significantly; this can be thought as
a starting network, attracting new players thanks to a
good overall balance. Some ties have formed and
many new competences (the dimension of
enterprises) have been internally produced.
After the initial steps in which 50% of the
enterprise was doing internal exploration, now at the
100
th
step, only one third (i.e.: 33%) is doing that,
since almost all the smaller players are trying to
ICE-B 2009 - International Conference on E-business
172
outsource them from the bigger ones, in order to
gain some energy.
Figure 1: The network at time 0.
Figure 2: The network at time 100.
Unfortunately, many of these small enterprises
have no competence to give to the bigger one in
exchange for theirs. They will eventually die (ceased
enterprises) or try to change strategy, by starting an
internal exploration. That’s why at time 500 (figure
3) the total number of players increased again, but at
a lower rate (1 every 15 steps, as an average) and
now, in percentage, most of the survived enterprises
are doing external exploration (62% circa) and have
become quite big (many internal competences
possessed). Notice that in this experiment the
threshold under which an enterprise must cease is a
low value, meaning that few of them have to leave
the market. This was done intentionally to show how
enterprises can react and adapt their behaviour even
if they are modelled as reactive agents.
Figure 3: The network at time 500.
7 CONCLUSIONS AND
OUTLOOK
Process innovation is characterized by two important
aspect: one critical and typical aspect is the ability to
gather, develop and transform information and
knowledge in a potential competitive advantage. The
second aspect regards spending resources like time
and money: the development of process innovation
is usually time and resource consuming and is
difficult to attain, especially when referring to
radical cases. Though, process innovation is a key
factor for building a network for competences
exchange and a very important variable when
considering the strategies performed by an
enterprise; once possessed, the advantage can be
exploited or shared. In the first case, the enterprise
can gain customers and money, by being the only
one (or among the few ones) possessing it. But it
risks to lose its advantage as soon as other players
can develop it. Another strategy is that of sharing the
process innovation, in exchange for other
competencies and/or money.
An agent based model is introduced in this work,
aiming at capturing the dynamics behind the creation
and the following modifications of an enterprise
network for competences exchange, i.e.: a network
in which enterprises can internally develop and/or
share processes with other players. This is, by the
way, one of the focal points behind the creation of
industrial districts, enterprise clusters and so on. A
well established network of this kind can attract new
players, that will probably bring new knowledge and
competences in it.
The model is formally discussed in detail, and so
the agents composing it and its iterations. While
studying quantitative results is beyond the purpose
of this work, a qualitative analysis is described, and
IMPACT OF PROCESS INNOVATION ON ENTERPRISE NETWORKS FOR COMPETENCES EXCHANGE - E³, a
Multi Agent based Model
173
the network graph, one of the graphical outputs
supplied by the model, is analyzed: in order to show
how network dynamics emerge from the model and
its parameters, settable by the user.
At the beginning, when the enterprises have few
competences and high perception of how can be
difficult develop and innovation process, they try to
link with the enterprises that have already developed
innovative processes. That’s why, in an initial phase,
the number of enterprises doing external exploration
tends to increase. After some steps, the number of
enterprises choosing external exploration is lower
and lower and limited to the smallest players, or the
newly arrived ones. The reason is that at the
beginning, the enterprise’s capability are low and the
perception of the effort for developing a process
innovation is high. The enterprise at this phase
typically try to share and exchange competences
with others that have already developed the
innovative process, not having to face the risk of
inside developing, even if this can be more gainful
in the long run. As time passes by, the enterprises
start to become bigger and be more conscious about
their capabilities and knowledge, thus reducing the
perception of the effort to develop innovative
processes internally.
The model is comprehensive and its scope is
wide. In future works other features will be
described in detail, and quantitative analysis will be
carried on in order to study real-world cases (e.g.:
existing industrial districts and so on) and the
underlying dynamics that lead to their creations.
Besides, a new feature will be implemented in
the model, referred to as “shock mode”, allowing the
user to stop the model at a given step, and change
some inner parameter. For example, it will be
possible to add a specific competence to one
enterprise only, so that it’s the only one in all the
network possessing it. In that way it becomes
possible to study how and based on which dynamics
this specific competence spreads on the network and
which kind of competitive advantage it gives, in
terms of central position in the network and
bargaining power to obtain other competences not
possessed internally.
ACKNOWLEDGEMENTS
The authors would like to gratefully acknowledge
the key support of prof. Anna Maria Bruno, and
prof. Gianpiero Bussolin, Full Professors of
Economia e Gestione delle Imprese” at the
University of Turin.
REFERENCES
Akbar Zaheer1 And Geoffrey G. Bell, Benefiting from
network position: firm capabilities, structural holes,
and performance. Strategic Management Journal,
2005
Balaji R. Koka John E. Prescott , Designing alliance
networks: the influence of network position,
environmental change, and strategy on firm
performance, Strategic Management Journal, 2008
Cohen WM, Levinthal DA. 1990. Absorptive capacity: a
new perspective on learning and innovation.
Administrative Science Quarterly 35: 128–152.
Cohen WM, Levinthal DA. 1994. Fortune favors the
prepared firm. Management Science 40(2): 227–251
Cohen, W.M., Levinthal, D.A., 1990. Absorptive capacity:
a new perspective on learning and innovation.
Administrative Science Quarterly 35, 128–152.
Coleman JS. 1988. Social capital in the creation of human
capital. American Journal of Sociology 94:
(Supplement): 95–120
Coleman JS. 1990. Foundations of Social Theory. Belknap
Press of Harvard University Press: Cambridge, MA
Frank Harary, Mathematical aspect of electrical network
analysis, 1969
Gulati R, Gargiulo M. 1999. Where do interorganizational
networks come from?. American Journal of Sociology
104(5): 1439–1493
Gulati R, Nohria N, Zaheer A. 2000. Strategic networks.
Strategic Management Journal 21(3): 203–215
Gulati R. 1998. Alliances and networks. Strategic
Management Journal, Special Issue 19(4): 293–317
Gulati R. 1999. Network location and learning: the
influence of network resources and firm capabilities
on alliance formation. Strategic Management Journal
20(5): 397–420
Nitin Nohria, The Differentiated Network: Organizing
Multinational Corporations for Value Creation, 1997
Phlippen S., Riccaboni M. Radical Innovation and
Network Evolution, 2007
Remondino, M. Agent Based Process Simulation and
Metaphors Based Approach for Enterprise and Social
Modeling, ABS 4 Proceedings, SCS Europ. Publish.
House – ISBN 3-936-150-25-7, pp.93-97, 2003
Remondino M., Correndo G., MABS Validation Through
Repeated Execution and Data Mining Analisys,
International Journal of SIMULATION: Systems,
Science & Technology (IJS3T), Volume 7, September
2006 – UKSim Publish. House, ISSN: 1473-8031
Sherer, Innovation and Growth: Schumpeterian
Perspectives, 2007
Thongchai Srivardhana ,Suzanne D. Pawlowskiv ERP
systems as an enabler of sustained business process
innovation: A knowledge-based view. Science Direct,
2007
Zhao Jing, Dang Xing-hua,
Research on the Evolution of
Enterprises’ Knowledge Space-time in the Techno-
innovation Process under the Network Environment,
International Conference on Management Science &
Engineering, 2008.
ICE-B 2009 - International Conference on E-business
174