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

, a Multi Agent based Model.

In Proceedings of the International Conference on e-Business, pages 167-174

DOI: 10.5220/0002233001670174

 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 in “normal”

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

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