Conference PaperPDF Available

Agent-based modelling of Stakeholder Interaction in Transport Decisions

Authors:

Abstract and Figures

This is an abridged version of the paper presented at the conference. The full version is being submitted elsewhere. Details on the full paper can be obtained from the author. ABSTRACT Community Involvement, Public Engagement, Stakeholder Engagement, are all different ways to name the participation process of interested people to public decisions. In transport planning there are lots of decisions concerning several issues, with diverse stakeholders involved from organizations to citizens. Sometimes involvement is just a single, compulsory moment of the decision-making process and it lacks in its real purpose: engaging people to find the most shared solution in the shortest time, in order to make the process effective and (cost) efficient. The aim of this work is to improve the knowledge of the involvement process by building the network of relationships among stakeholders and analysing the opinion dynamics which leads to the final decision. The methodology proposed uses an agent-based simulation and a multi-state opinion dynamics and bounded confidence model as a basis to investigate the consensus formation phenomenon. It can be used as a tool both for a preventive analysis addressed to plan an effective participation process and to predict and foster the emergence of a coalition of stakeholders towards a shared decision.
Content may be subject to copyright.
Agent-based modelling of Stakeholder Interaction in Transport Decisions
LE PIRA Michela; IGNACCOLO Matteo; INTURRI Giuseppe; GAROFALO Cesare; Pluchino
Alessandro; RAPISARDA Andrea
AGENT-BASED MODELLING OF
STAKEHOLDER INTERACTION IN
TRANSPORT DECISIONS
Michela Le Pira, University of Catania, Italy, mlepira@dica.unict.it
Matteo Ignaccolo, University of Catania, Italy, matig@dica.unict.it
Giuseppe Inturri, University of Catania, Italy, ginturri@dica.unict.it
Cesare Garofalo, University of Catania, Italy, cesar egarofalo@yahoo.com
Alessandro Pluchino, University of Catania, Italy, alessandro.pluchino@ct.infn.it
Andrea Rapisarda, University of Catania, Italy, andrea.rapisarda@ct.infn.it
This is an abridged version of the paper presented at the conference. The full version is
being submitted elsewhere. Details on the full paper can be obtained from the
author.
ABSTRACT
Community Involvement, Public Engagement, Stakeholder Engagement, are all different
ways to name the participation process of interested people to public decisions. In transport
planning there are lots of decisions concerning several issues, with diverse stakeholders
involved from organizations to citizens. Sometimes involvement is just a single, compulsory
moment of the decision-making process and it lacks in its real purpose: engaging people to
find the most shared solution in the shortest time, in order to make the process effective and
(cost) efficient. The aim of this work is to improve the knowledge of the involvement process
by building the network of relationships among stakeholders and analysing the opinion
dynamics which leads to the final decision. The methodology proposed uses an agent-based
simulation and a multi-state opinion dynamics and bounded confidence model as a basis to
investigate the consensus formation phenomenon. It can be used as a tool both for a
preventive analysis addressed to plan an effective participation process and to predict and
foster the emergence of a coalition of stakeholders towards a shared decision.
Keywords: transport planning, stakeholder engagement, public engagement, agent-based
model, opinion dynamics, sustainable mobility
13th WCTR, July 15-18, 2013 – Rio de Janeiro, Brazil
1
Agent-based modelling of Stakeholder Interaction in Transport Decisions
LE PIRA Michela; IGNACCOLO Matteo; INTURRI Giuseppe; GAROFALO Cesare; Pluchino
Alessandro; RAPISARDA Andrea
INTRODUCTION
Community Involvement has become a relevant part of a decision-making process. The five
Public Engagement (PE) levels described by Cascetta and Pagliara (2011) (stakeholder
identification, listening, information giving, consultation, participation) are all linked with the
different phases of the “bounded rationality transportation planning process” and they refer to
levels of growing involvement. Social interaction is a key of success in transport planning,
because it fosters the emergence of coalitions facilitating the convergence of different
stakeholders to a shared solution. Therefore, planning becomes the management of a bi-
directional communication process and it requires specific programs and skills, able to
coordinate many players, conflicting interests and variables and anticipate problems. In this
respect the use of Decision Support Systems, based on quantitative methods (Cascetta, 2009),
can help to assess the outcome of different alternatives to increase the transparency and the
reproducibility of the decision process.
Community Involvement is an important part of the decision-making process according to
sustainability principles, as confirmed by the EU transport policy tendency. The Sustainable
Urban Mobility Plan (Buhrmann et al., 2011) and the Sustainable Urban Transport Plan
(Wolfram and Buhrmann, 2007) have become a reference point for a new way of transport
planning. Sustainable Urban Mobility Plans mean “Planning for the People” (Buhrmann et al.,
2011). They are the result of an integrated planning approach, with the aim to create a
sustainable urban transport system, also through a participatory approach. In Italy, public
participation in transport planning is required by law only for the Strategic Environmental
Assessment (Directive 2001/42/EC), and it must be carried out all along the planning process
from the beginning to the end.
Stakeholder theory and engagement
The concept of “stakeholder” was introduced by Freeman (1984) and it derives from
Economy, where there is a well-established literature affirming that the power of a company
lies on its relationships with them. Mitchell et al. (1997) report a chronology of the concept of
stakeholder and the key constructs in their theory of stakeholder identification and salience.
In transport planning there are lots of different stakeholders to be involved, e.g. citizens,
policy makers, public institutions, local communities, governmental organizations, NGOs,
public transport operators, experts, retailers, the private sectors and the third sector. For
example the authors, as partners of the PORTA project (www.porta-project.eu), supported by
the European Regional Development Fund within the MED Programme, are experimenting
the relevance of public participation of the diverse stakeholders involved in port planning and
in particular the relationships between Port Authority and city/citizens. The complexity of the
task requires specific tools; the methodology proposed in this work can help the knowledge of
the information exchange among the diverse stakeholders involved in transport planning.
There are several tools that can be used to engage: Roden (1984) suggests to develop a
“Community Involvement Plan”, the GUIDEMAPS Handbook (Kelly et al., 2004) reports the
different tools coupled with the phases of the involvement process, Whitmarsh et al. (2007)
propose a methodology divided into two phases (expert focus groups and questionnaires,
13th WCTR, July 15-18, 2013 – Rio de Janeiro, Brazil
2
Agent-based modelling of Stakeholder Interaction in Transport Decisions
LE PIRA Michela; IGNACCOLO Matteo; INTURRI Giuseppe; GAROFALO Cesare; Pluchino
Alessandro; RAPISARDA Andrea
citizen workshops and questionnaires), Mameli and Marletto (2009) propose a participate
procedure by involving experts, citizens and stakeholders to implicate in different ways with
“top-down” phases (results derived from the work of experts) and “bottom-up” phases (results
derived from the participation of citizens and stakeholders). It is clear that all the methods are
time-consuming and require money, so it is not easy to make a good involvement. Indeed
there are lots of negative examples where decisions failed because of lack of Community
Involvement (e.g. the High Speed Rail Turin-Lyon). In addition to the traditional tools,
having a clear insight of the actors who take part in the decision-making and predicting the
possible results of an interaction can be of great benefit for the planning process. In this
respect linking together stakeholders in a social network and simulating the communication
among them can help to improve the knowledge of the social interaction mechanisms.
Social network analysis and opinion dynamics models
The analysis of the network consists of finding properties which cannot be obtained by
visualization. Social Network Analysis (SNA) is a powerful instrument in doing so, because it
allows to measure the centrality of the different stakeholders and the potential problems due
to topology. The use of SNA in the field of Stakeholder Engagement can simply consists of
stakeholder mapping or it can include centrality measures.
Stakeholder engagement is a dynamic process and it is characterized by several reassessment
of the network. Together with the network analysis it can be helpful to simulate how the
opinions flow through the set of connections in order to improve the knowledge of the
involvement process at the earliest stage and to understand how to manage stakeholders. The
opinion dynamics which should lead to consensus can be reproduced through different
models. One of the most widely known is the Hegselmann and Krause (HK) compromise
model (2002), where the nodes form their actual opinion by taking an average opinion based
on their neighbours’ ones (i.e. the nodes connected with an edge). This leads to a dynamical
process which should flow into a consensus among all agents.
In general the opinion dynamics models consist of algorithms that can be analytically or
numerically solved; the dynamics is usually simulated by means of Monte Carlo algorithms.
Agent-based modelling is a powerful instrument in simulating the opinion dynamics for many
reasons, such as the relative easiness to represent a network of nodes (agents) linked together
with ties, the possibility to ask the agents (endowed with own properties) to have an opinion
and act according to simple behavioural laws, the power of visualization, that can help the
analysis, the opportunity to change the global variables, which makes generalization possible
and especially for the emergence of collective behavioural patterns which are not predictable
from the simple initial rules and that come out from simulations. For all these reasons, agent-
based modelling is suitable to represent the stakeholder network and to simulate the opinion
dynamics.
Therefore, in this work the focus is on a potential step of the participation process: the study
of how the network topology and the initial conditions can influence the final decision, by
simulating the opinion dynamics which takes place in the stakeholders’ network.
13th WCTR, July 15-18, 2013 – Rio de Janeiro, Brazil
3
Agent-based modelling of Stakeholder Interaction in Transport Decisions
LE PIRA Michela; IGNACCOLO Matteo; INTURRI Giuseppe; GAROFALO Cesare; Pluchino
Alessandro; RAPISARDA Andrea
METHODOLOGY
The need to include Public Engagement in transport decision-making process leads to the
effort to understand how to design and speed the process of taking a public decision and to
find out if the communication among stakeholders can influence the process of governance.
Social network analysis and opinion dynamics models can allow to know how the actors
involved in the planning process are linked together, how the network structure can enable or
limit a joint action and how the social and spatial architecture of the community network can
influence the outcome of the planning process. It is worth to make a distinction between the
two techniques:
SNA can be used to make static measures of the network, improving the knowledge
of the actors involved and helping to understand how a modified topology can foster
the emergence of coalitions towards a shared solution;
opinion dynamics models allow to make dynamic measures which can help to make
prediction about the final decision that might derive from interaction.
The methodology proposed is based on an agent-based simulation of the opinion dynamics on
a stakeholders’ network, through the implementation of a multi-state opinion dynamics and
bounded confidence model. It is not intended as an operative participative decision-making
tool, but as a strategic and preventive mean to plan an effective participation process and to
predict and foster the emergence of a coalition of stakeholders towards a shared decision.
We used NetLogo (Wilensky, 1999), a multi-agent programmable modelling environment
which can reproduce lots of characteristics of complex systems, following the time evolution
and the significant parameters real-time. NetLogo was previously used in transport modelling,
e.g. for the simulation of pedestrian behaviour (Camillen et al., 2009) and the impact of real
time information on transport network routing (Buscema et al., 2009).
Implementation of the multi-state opinion dynamics and bounded confidence
model
The implemented model is inspired to the majority rule (MR) model (Galam, 2002), where all
the agents at time t are endowed with binary opinions (+1, 1) and they can communicate
with each other. At each interaction, a group of agents is selected at random (discussion
group): as a consequence of the interaction, all agents take the majority opinion inside the
group. Our model can be considered a multi-state opinion model where agents are endowed
with one opinion among approval, disapproval or neutral, denoted by +1, 1 and 0
respectively. The neutral opinion is considered less significant and “contagious” than the two
others, so the latter were assigned with a double weight. Each node can change its opinion at
time t+1 based on its neighbours’ ones with a probability related to their influence. It is also a
bounded confidence model, because of the definition of a confidence bound which limits the
way a node can change its opinion: a node with +1 cannot directly change its opinion in 1
(and vice versa), but it must pass through the opinion 0 before. The activation of the
confidence bound depends on the node property influenceability, a random real number in the
range [0,1], which represents the probability that a node directly changes its opinion without
any confidence bound. If the parameter has a value close to 1, the probability to directly
change its opinion without passing through the neutral stance is high and vice versa when the
13th WCTR, July 15-18, 2013 – Rio de Janeiro, Brazil
4
Agent-based modelling of Stakeholder Interaction in Transport Decisions
LE PIRA Michela; IGNACCOLO Matteo; INTURRI Giuseppe; GAROFALO Cesare; Pluchino
Alessandro; RAPISARDA Andrea
value is around 0. In conclusion, each node is characterised by a certain influence (which
affects the neighbours’ opinions) and by a certain “influenceability” (which expresses to what
extent a node can be influenced by its neighbours).
The implemented algorithm consists of the creation, for each node, of a vector filled with the
weighted opinions of all the neighbours. Let xi (t) be the opinion of the node i at time t; the
opinion at time t + 1 will be:
where vi (t) is the vector of the neighbours’ opinions, which are repeated, for each neighbour,
a number of time related to the opinion weight, the influence and according to a belonging
factor, considering that there are more possibilities to interact within the same group:
with k = 1, +1, 0.
At each time an element of the vector will be randomly chosen, therefore the most numerous
opinion will be the most likely to be selected. At this point it is useful to distinguish “strong
ties” from “weak ties”, a standard description in community structure analysis for indicating,
respectively, links between nodes belonging to the same group and links between nodes
belonging to different groups. We call “degree” the total number of links (strong + weak) of a
given node and “z-out” the number of weak links of the same node.
In order to reproduce potential external influences to the opinions, we assumed that the
dynamics can be modified by means of Changing-Mind-Rate (CMR), a factor we introduced
to represent the probability that a given node would randomly change its opinion at a given
time. We considered a single event when, starting from a given distribution of opinions
among the agents, it ends with all agents converging towards the same opinion. We also
considered a multi-event version, with different (random) results related to the same initial
distribution of opinions.
The dynamics starts from a positive initial group, that is to say a group of nodes that initially
have the +1 opinion. Therefore, for what concerns the simulations, there are three main
elements that can be modified:
1. Topology (average degree, average z-out)
2. Initial conditions (positive initial group)
3. Opinion dynamics (CMR)
Considering N events for each simulation, we are interested into the following simulations’
results: the number of events ended with a complete consensus (all the opinions equal to +1)
or complete dissent (all the opinions equal to 1) and the average time for reaching consensus
or dissent. In order to convert the final outcome of the events into a unique index we
calculated the parameter W as the weighted average of the final network state, i.e. the net
frequency of the events which end with +1:
where Nk is the number of events ended with consensus (k = +1) or dissent (k = 1) and N is
the total number of events. W is included in the interval [1,+1], where the extreme values 1
and +1 represent, respectively, 100% of events ended with dissent or consensus. It represents
13th WCTR, July 15-18, 2013 – Rio de Janeiro, Brazil
5
Agent-based modelling of Stakeholder Interaction in Transport Decisions
LE PIRA Michela; IGNACCOLO Matteo; INTURRI Giuseppe; GAROFALO Cesare; Pluchino
Alessandro; RAPISARDA Andrea
a statistics of the events and it does not indicate the rate of agents which have the opinion +1
at a certain step of the simulation or the degree of sharing of a project. On the other hand, it is
an index which measures the tendency of the final state of the system towards the full
consensus or the full dissent, so it represents the final configuration of the opinions.
A time threshold was defined in order to exclude the cases in which the process took too long
time (t > 500) before reaching consensus (or dissent). Therefore, when time exceeds the
threshold without reaching any convergence of opinions, we say that the simulation outcome
is “no consensus/dissent”.
CASE STUDY
The decision-making process regarding transport planning is characterized by a high level of
complexity and it is not simple to be described with a model. Therefore, in order to apply our
methodology to a case study, we decided to represent a simple, real situation of a decision-
making process regarding transport issues. In particular we depicted a well-known situation of
a narrow and homogeneous community of people with the same interest, i.e. easy access to
the workplace. In particular, the case study of this work is about the idea of adopting parking
pricing inside the Campus of the University of Catania as one of the main transport policy for
sustainable mobility proposed by the mobility management office of the University. The topic
involves all the University staff, including full professors, associate professors and assistant
professors, while students are excluded because they cannot access those parking spaces.
Some observations carried out during several meetings on these issues, though not systematic
and statistically significant, were useful for the construction of the model. The network was
created according to relationships derived by roles and by department organization
(institutional relationships). Thanks to the knowledge of all the elements it was possible to
build the network and simulate the opinion dynamics which should lead to a
consensus/dissent.
Simulations and results
Taking into consideration topology, in order to reproduce realistic situations, two cases were
considered:
1. average degree 10, i.e. on average each node is connected with other 10 nodes;
2. average degree 20, i.e. on average each node is connected with other 20 nodes.
The simulations were performed by varying the number of weak ties, i.e. with a parameter z-
out ranging, in average, from 1 to 5 for degree 10 and from 5 to 10 for degree 20 (both degree
and z-out are extracted from normal distributions). We considered 10 different (random)
realizations of the initial distribution of opinions (multi-event version with N = 10). To
understand the impact of external influences on the final decision, a series of simulations was
made with average degree = 20, CMR = 0.5% and z-out varying from 5 to 10. The next tables
show some results in terms of the parameter W, as above defined.
13th WCTR, July 15-18, 2013 – Rio de Janeiro, Brazil
6
Agent-based modelling of Stakeholder Interaction in Transport Decisions
LE PIRA Michela; IGNACCOLO Matteo; INTURRI Giuseppe; GAROFALO Cesare; Pluchino
Alessandro; RAPISARDA Andrea
Table I - Parameter W with random initially positive nodes (av. degree = 10, CMR 0.0%).
number of
random
positive
1 2 3 4 5
0 -1.0 -1.0 -1.0 -1.0 -1.0
50 no consensus/dissent -1.0 -1.0 -1.0 -1.0
100 no consensus/dissent no consensus/dissent -1.0 -0.8 -0.8
150 no consensus/dissent no consensus/dissent 0.4 -0.6 0.2
200 no consensus/dissent no consensus/dissent 0.8 0.8 0.6
250 no consensus/dissent 1.0 1.0 1.0 1.0
300 no consensus/dissent 1.0 1.0 1.0 1.0
350 no consensus/dissent 1.0 1.0 1.0 1.0
400 1.0 1.0 1.0 1.0 1.0
W
average degree = 10, CM R = 0.0
average z-out
Table II - Parameter W with initially positive groups (av. degree = 10, CMR = 0.0%).
average
influence
number
of nodes
head of
department
full
professors
associate
professors
assistant
professors
5
W
random nodes
1 department
2 departments
3 departments
positive initial group
4 departments
1
2
3
4
average z-out
average degree = 10, CMR = 0.0
Whatever the initial conditions are, it is clear that a too small number of weak ties critically
slows down the information exchange; actually, when a node has on average 10 links, it is
evident that we need more than 2 weak ties in order to reach convergence of opinions.
Furthermore, the parameter W is minimum when the positive initial nodes are heads of
departments (a minority, but very much influent) or assistant professors (more numerous, but
less influent), that is to say that it is very difficult to reach consensus when only one of these
groups is originally positive about the given topic (in our case the parking pricing). On the
other hand, higher W values are achieved with entire positive departments. In Table I it is
useful to make comparisons by column, in order to notice the change from total dissent (i.e.
100% of events ended with dissent) to total consensus (i.e. 100% of events ended with
consensus) as the number of initially positive nodes increases. Analysing the results by row in
Table I and Table II it appears that, in the transition phase (and in particular in proximity of
the critical threshold), which is an area of “turbulence”, there are fluctuations in the results
(e.g. for 150 random positive nodes) also due to the limited number of simulations with the
same starting conditions. This result is also visible if we study the behaviour of the parameter
W versus an increasing number of randomly chosen initially positive nodes (ranging from 0 to
13th WCTR, July 15-18, 2013 – Rio de Janeiro, Brazil
7
Agent-based modelling of Stakeholder Interaction in Transport Decisions
LE PIRA Michela; IGNACCOLO Matteo; INTURRI Giuseppe; GAROFALO Cesare; Pluchino
Alessandro; RAPISARDA Andrea
400), where a transition from dissent (W = 1) to consensus (W = +1) clearly appears in
correspondence of around 150 positive nodes and can be appreciated plotting the parameter W
within a scatter diagram ( Figure 1). Indeed, all the events end with dissent up to 50, then
there is a transition phase with some events ended with dissent and some others with
consensus (from 50 to 250 nodes) and where the lines for different z-out can intersect, whilst
all the events end with consensus when there are more than 250 (randomly chosen) initially
positive nodes.
Figure 1 - The parameter W as a function of the number of random positive nodes on varying z-out (av. degree =
10, CMR = 0.0%).
For what concerns the average time to reach the final decision, it is possible to plot it as a
function of the number of random positive nodes and for several values of z-out (Figure 2).
It results that the convergence time presents a peak exactly in correspondence of the transition
from total dissent to total consensus. Such a peak is much more pronounced for smaller values
of the average z-out, i.e. when the small number of weak ties does not allow the positive
opinions to spread over the entire networks.
Figure 2 - Average convergence time as a function of the number of random positive nodes on varying z-out (av.
degree = 10, CMR = 0.0 %).
If we increase the number of links (average degree = 20) the results are similar. The greater
number of links improves the communication among nodes, so consensus/dissent is always
reached, even when the number of weak ties is small. If we consider the presence of external
influences, represented by non zero values of the CMR indicator (CMR = 0.5%) in general it
produces an increase in convergence time but does not significantly affect the transition from
13th WCTR, July 15-18, 2013 – Rio de Janeiro, Brazil
8
Agent-based modelling of Stakeholder Interaction in Transport Decisions
LE PIRA Michela; IGNACCOLO Matteo; INTURRI Giuseppe; GAROFALO Cesare; Pluchino
Alessandro; RAPISARDA Andrea
dissent to consensus, which occurs between 150 and 200 initially positive (randomly chosen)
nodes.
CONCLUSIONS AND DISCUSSIONS
Transport planning is mainly a complex decision-making process, with many actors involved
and different conflicting objectives and opinions. In this paper we propose an agent-based
model that can simulate the opinion dynamics on a network of stakeholders involved in
transport planning, in order to support the decision-making process. The presented model is a
multi-state opinion model with 3 different opinions. It is also a bounded confidence model
because of the presence of a confidence bound which limits the opinion changing from
approval to disapproval (and vice versa) by means of the neutral opinion. We applied our
model in a very simple case study, both to test the model and to capture the intrinsic essence
of the complex phenomena of social interaction. The decision-making process regards the
adoption of a new parking pricing system inside a University Campus, where a well-known
situation of a narrow and homogeneous community of people (professors) with the same
interest, made quite reasonable the opinion dynamics model we implemented. For what
concerns topology, many links help the communication among nodes and it takes few time to
reach the final decision, while few links slow down the process and sometimes it requires too
much time to reach consensus or dissent. Choosing random initial positive nodes, there is a
transition from dissent to consensus within which the time required for the convergence of
opinions has a peak. Introducing external influences which affect the dynamics, the process
slows down and it requires more time to reach a decision.
Further research will tend to modify the opinion dynamics, for instance increasing the number
of possible opinions or changing the model from a discrete choice model to a continuum one,
or including the possibility that the stakeholders could change their mind by policy persuasion
or awareness raising. Indeed, our model considers that some people can have a greater weight
than others through the parameter influence, but we are neutral about the result. For what
concerns the stakeholder network, it would be useful to see how the geographical distance and
the department affinity influence the topological distance of the nodes, affecting the
information exchange. Moreover, in order to calibrate the model, it would be helpful to
compare the results of the simulations with a real situation with systematic observations to see
if the model results are in agreement with reality.
In conclusion, Stakeholder Engagement is an integral part of the transport planning process. It
involves all the stakeholders from the very beginning of the planning process, with different
levels of involvement during the planning phases. Its aim is to foster the emergence of
coalitions among stakeholders towards a shared solution.
Our model can be useful to the design of the stakeholder involvement at an early stage of the
planning process, because it can predict and, therefore, raise the awareness of the possible
results of interaction; consequently it allows to set up the priority for information and it helps
to understand how to improve the linkages among stakeholders in order to facilitate the
involvement process; moreover it investigates the probability that external influences can
modify the convergence towards a shared solution. Therefore, studying the stakeholder
network and the opinion dynamics can help to understand how to make a good involvement
13th WCTR, July 15-18, 2013 – Rio de Janeiro, Brazil
9
Agent-based modelling of Stakeholder Interaction in Transport Decisions
LE PIRA Michela; IGNACCOLO Matteo; INTURRI Giuseppe; GAROFALO Cesare; Pluchino
Alessandro; RAPISARDA Andrea
process and can be helpful to make the planning process transparent, effective and (cost)
efficient.
REFERENCES
Buhrmann, S., F. Wefering and S. Rupprecht (2011). Guidelines. Developing and
Implementing a Sustainable Urban Mobility Plan. ELTISplus, Working document.
Rupprecht Consult, Germany. In:
http://www.sustainable-urban-mobility-plans.org/docs/SUMP_guidelines.doc
Buscema, D., M. Ignaccolo, G. Inturri, A. Pluchino, A. Rapisarda, C. Santoro and S. Tudisco
(2009). The impact of real time information on transport network routing through
Intelligent Based Simulation. Proceedings to TIC-STH Symposium on Human Factors
and Ergonomics. IEEE Xplore, ISBN: 9781424490943.
Camillen, F., S. Caprì, C. Garofalo, M. Ignaccolo, G. Inturri, A. Pluchino, A. Rapisarda and S.
Tudisco (2009). Multi agent simulation of pedestrian behavior in closed spatial
environment. Proceedings to TIC-SENCS Symposium on Engineered and Natural
Complex Systems - Modeling, Simulation and Analysis. IEEE Xplore, ISBN :
9781424438785.
Cascetta, E. (2009). Transportation System Analysis: Models and Applications. Springer,
New York.
Cascetta, E. and F. Pagliara (2011). Public Engagement and Transportation Planning: Some
evidences from Italy. SIDT, National Conference and International Seminar, Venice.
In: http://www.sidt.org/sidt2011/wp-content/uploads/SIDT-2011_PL2_02_Public-
Engagement-and-Transportation-Planning.pdf
Freeman, R. E. (1984). Strategic management: A stakeholder approach. Pitman, Boston.
Hegselmann, R. and U. Krause (2002). Opinion dynamics and bounded confidence: models,
analysis and simulation. Journal of Artificial Societies and Social Simulation, 5(3).
Kelly, J., P. Jones, F. Barta, R. Hossinger, A. Witte and A. C. Wolf (2004). Successful
transport decision-making. A project management and stakeholder engagement
handbook. In: http://www.civitas-initiative.org/docs1/GUIDEMAPSHandbook_web.pdf
Mameli, F. and G. E. Marletto (2009). A participative procedure to select indicators of
sustainable urban mobility policies. CRENoS-CUEC, Cagliari.
Mitchell, R. K., B. R. Agle and D. J. Wood (1997). Towards a theory of Stakeholder
Identification and Salience: Defining the Principle of Who and What really counts.
Academy of Management Review, Vol. 22, No. 4, 853-886.
Roden, D. B. (1984). Community Involvement in Transportation Planning. Technical Report
Series 37. North Central Texas Council of Governments, Arlington, TX.
Whitmarsh, L., A. Haxeltine and M. Wietschel (2007). Sustainable transport visions: expert
and non-expert stakeholder perspectives on sustainable transport. International
Conference on Whole Life Urban Sustainability and its Assessment, Glasgow. In:
http://download.sue-mot.org/Conference-2007/Papers/Whitmarsh.pdf
Wilensky, U. (1999). NetLogo. Center for Connected Learning and Computer Based
Modeling. Northwestern University, Evanston, IL. In:
http://www.ccl.northwestern.edu/netlogo
13th WCTR, July 15-18, 2013 – Rio de Janeiro, Brazil
10
Agent-based modelling of Stakeholder Interaction in Transport Decisions
LE PIRA Michela; IGNACCOLO Matteo; INTURRI Giuseppe; GAROFALO Cesare; Pluchino
Alessandro; RAPISARDA Andrea
Wolfram, M. and S. Buhrmann (2007). Sustainable Urban Transport Planning SUTP Manual.
Guidance for stakeholders. Rupprecht Consult, Germany. In:
http://www.ccl.northwestern.edu/netlogo
13th WCTR, July 15-18, 2013 – Rio de Janeiro, Brazil
11
... It has only been in more recent years that the stakeholders have become more engaged in making the appropriate transportation decisions and invited to give their ideas and express their needs and concerns [15]. The awareness of including the stakeholders in the transportation decision-making process is a consequence of the failure of many projects because of lack of consensus building [3]. ...
Article
Full-text available
Including an active participation of stakeholders along the transportation decision-making process is increasingly recognized as a necessary condition for reaching successful and high-quality decisions. This paper presents a framework for deciding on the appropriate transportation strategy for a supply chain from a multistakeholder perspective. It consists of three steps: (1) defining the transportation-strategy decision-making context and the objectives that must be achieved; (2) analyzing the actual transportation strategy regarding its three components: transportation network; transportation mode; and transportation insource/outsource, as well as identifying the stakeholders interested in the study; and (3) conducting a group decision making regarding each transportation strategy’s component, while involving the key stakeholders and taking into account the specificities of transported products. The proposed framework is then applied to a real case of the Moroccan public pharmaceutical supply chain, which has different features that distinguish it from other supply chains including its importance, urgency, and regulation. We employed the DELPHI method to determine the key stakeholders that should be involved in the decisional process. After that, we applied the group AHP method for selecting the appropriate transport-network design option while involving the identified key stakeholders.
... Their work demonstrates that the technique used helps collect public opinions more accurately that can represent the public interest in the selection of the best development and design scheme for the territory, which would not be corrupted by the "dominating opinions". According to La Pira et al. (2013) the social interaction, between people that have a common interest, is a key of success in transport planning because it fosters the emergence of coalitions among stakeholders towards a shared solution. Also, it can help to understand how to make a good involvement process and can be helpful to make the planning process transparent, effective and cost-efficient. ...
Article
Full-text available
Diverse stakeholder participation in transport planning is beneficial but difficult to achieve, as it deals with various levels of government, operators, users, and other interested parties. Furthermore, such planning is confined by issues of geography, economics and human demand, and in the case of transport it must integrate with other territorial constraints. The authors had an opportunity to carry out an accompanying research while municipalities or regional agencies establish a widened stakeholder involvement framework for municipalities including their surrounding ones, known as Functional Urban Area (FUA). In this process, aiming at optimizing commuter traffic, participating institutions are trying to set up a coordinating structure among various stakeholder in each FUA. To optimize the commuter traffic, a various new types of mobility services are considered as potential implementations such as ride sharing, car sharing, bike sharing, etc., as well as classical types of transport services such as public transport. In this paper, we present what kind of role each stakeholder can play for different types of measures towards environment-friendly commuter traffic. Following this, we present an assessment about how it may change along with the penetration of higher-level autonomous vehicles (AVs). It will add some extra roles to public authorities compared to today, especially as regulators and financers. Stakeholder involvement to address questions arising with the penetration of AVs onto the street will have to be carried out in a step-wise manner, starting with those having with endogenous motivation for sustainable mobility, and then being extended to further stakeholders.
... The new approach of participation in decision-making process involves several actors with different roles and it is inspired to some basic concepts, i.e: (i) levels of growing involvement, as represented by the "ladder of citizen participation" (Arnstein, 1969); (ii) a clear classification of the main actors involved, i.e. experts, stakeholders and citizens that contribute with different degrees of competence and interest to the decision-making process (Le Pira et al., 2013); (iii) an integrated transport planning decision-making model (Cascetta et al., 2015), where a "cognitive decision-making" is bounded with stakeholder engagement and quantitative analysis. In the framework of participatory decision-making process in transport planning, planners and experts define the plan structure for the final technical evaluations, stakeholders and citizens are involved in all the planning phases for the definition of objectives, evaluations criteria and alternatives and decision-makers are in charge of the final decision, supported by a performance-based ranking and a consensus-based ranking of plan alternatives (Le Pira et al., 2015b). ...
Article
Full-text available
Stakeholder engagement is a key issue for sustainable transport planning. Appropriate methods and tools are needed to support an efficient participation process. This study presents a combination of Analytic Hierarchy Process (AHP) with the Delphi method as a useful support for participatory decision-making processes aimed at consensus building. A case study will be presented and the results will be analysed also via an agent-based model (ABM), used to reproduce the same process of convergence of opinions, with the aim to understand the role of network topology, stakeholder influence and other sensitive variables on the emergence of consensus.
... In this respect, Le Pira et al. [7] propose a new participatory transport planning framework ( Fig. 1) able to include (a) different levels of involvement (as presented in [8]), (b) different actors, i.e. experts, stakeholders and citizens that contribute with different degrees of competence and interest (as explained in [9]) to the "bounded-rationality" decision-making process, where quantitative evaluations of the planning alternatives are integrated within a participation process (as introduced in [2]). ...
Conference Paper
Full-text available
Including an active participation of citizens and stakeholders from the beginning of transport decision-making processes is widely recognized as a precondition to avoid the failure of projects/policies/plans as a consequence of a lack of con-sensus. Appropriate methods and tools are needed to support participation pro-cesses towards well-thought and shared solutions. In this paper quantitative methods, stakeholder interaction and simulation models are used to guide and reproduce a participatory experiment aimed at consensus building about mobili-ty management strategies. Analytic Hierarchy Process (AHP) has been used to elicit stakeholder preferences, different voting methods have been used to ag-gregate the individual preferences, group interaction has been performed via a facilitated dialogue to reach a consensus among stakeholders and an agent-based model (ABM) has been used to simulate the same consensus building process. Besides the social network of stakeholders has been analyzed to gain insights on its influence on the consensus formation. The results of this integrated procedure, applied in a pilot experiment with University students as stakeholders, provide useful suggestions on how to use different methods and guide effective and efficient participation processes aimed at consensus building.
... In previous studies (Le Pira et al., 2013;, an ABM was already implemented to simulate the opinion dynamics on a particular stakeholder network where a binary decision has to be taken about a single transport policy measure: the various agents interact with each other and the conditions leading to the convergence of opinions according to a majority rule were investigated. In this paper, a new ABM is presented to reproduce interaction in stakeholder networks with different topologies, where each stakeholder has an individual preference list over a set of (more than two) alternatives, and a collective preference list has to be found (see also Le Pira et al., 2015a). ...
Article
Abstract We address the problem of a participatory decision-making process where a shared priority list of alternatives has to be obtained while avoiding inconsistent decisions. An agent-based model (ABM) is proposed to mimic this process in different social networks of stakeholders who interact according to an opinion dynamics model. Simulations’ results show the efficacy of interaction in finding a transitive and, above all, shared decision. These findings are in agreement with real participation experiences regarding transport planning decisions and can give useful suggestions on how to plan an effective participation process for sustainable policy-making based on opinion consensus.
... The latter are also widely used for applications to social and economic systems and they are also used as a tool to foster stakeholder engagement and collaboration (Voinov and Bousquet, 2010). In a previous study (Le Pira et al., 2013), an ABM was already implemented to simulate the opinion dynamics on a particular stakeholder network where a binary decision has to be taken about a single transport policy measure: the various agents interact with each other and the conditions leading to the convergence of opinions according to a majority rule were investigated. In this paper, a new ABM is presented to reproduce interaction in stakeholder networks with different topologies, where each stakeholder has an individual preference list over a set of (more than two) alternatives, and a collective preference list has to be found (see also Le Pira et al., 2015a). ...
Article
Full-text available
We address the problem of a participatory decision-making process where a shared priority list of alternatives has to be obtained while avoiding inconsistent decisions, related to the “Condorcet paradox”. An agent-based model (ABM) is proposed to mimic this process in different social networks of stakeholders who interact according to an opinion dynamics model. Simulations’ results show the efficacy of interaction in avoiding the paradox and finding a transitive and, above all, shared decision. These findings are in agreement with real participation experiences regarding transport planning decisions and can give useful suggestions on how to plan an effective participation process for sustainable policy-making based on opinion consensus.
... The new approach of participation in decision-making process involves several actors with different roles and it is inspired to some basic concepts, i.e: levels of growing involvement, as represented by the "ladder of citizen participation" (Arnstein, 1969); the main actors involved, i.e. experts, stakeholders and citizens that contribute with different degrees of competence and interest to the decision-making process (Le Pira et al., 2013); the overall transport planning decision-making model (Cascetta et al., 2015), where a "cognitive decision-making" is bounded with stakeholder engagement and quantitative analysis. In the framework of participatory decision-making process in transport planning, planners and experts define the plan structure for the final technical evaluations, stakeholders and citizens are involved in all the planning phases for the definition of objectives, evaluations criteria and alternatives and decision-makers are in charge of the final decision, supported by a performance-based ranking and a consensus-based ranking of plan alternatives (Le Pira et al., 2015b). ...
Article
In this study a consensus building process based on a combination of Analytic Hierarchy Process (AHP) and Delphi method is presented and applied to the decision-making process about alternative policy measures to promote cycling mobility. An agent-based model is here used to reproduce the same process of convergence of opinions, with the aim to understand the role of network topology, stakeholder influence and other sensitive variables on the emergence of consensus. It can be a useful tool for decision-makers to guide them in planning effective participation processes.
... the "ladder of citizen participation" (Arnstein, 1969) and different levels of growing involvement (Kelly et al., 2004); the "participation pyramid" (Le Pira et al., 2013), where experts, stakeholders and citizens contribute with different degrees of competence and interest to the decision-making process; the transport planning "bounded-rationality approach" (Cascetta and Pagliara, 2013), where quantitative evaluations of the planning alternatives are integrated with a participation process. ...
Article
Full-text available
The aim of this work is to give a contribution in the field of stakeholder engagement in order to reduce the widespread conflicts arising when transport plans have to be implemented and understand the role of quantitative methods to support shared decisions. We present the results of a participation experiment, with university students as stakeholders, where the AHP method was applied to derive individual priority vectors, on the basis of their judgments of preference between all couples of alternatives regarding the mobility management of their university. The aggregation of the individual judgments was done by using different methods, some derived from AHP and other derived from voting methods, such as Pairwise Majority Rule (PMR). A discussion about the results of the different methods, before and after stakeholder interaction, and from an agent-based simulation in terms of respect of the consistency condition and degree of consensus of the collective decision will provide some recommendations that can be useful to guide effective and efficient participation process.
Thesis
Full-text available
The aim of the research is to give a contribution and an insight on the complex field of stakeholder involvement in transport planning, by analysing the role of decision-support methods and agent-based modelling in guiding a participation process. The approach is twofold: from one side it is about to deeply understand the process of making a collective decision, by studying how the interaction among different actors can lead to a convergence of opinions towards a shared collective decision. From the other side, it is based on finding appropriate decision-support methods to help the group decision-making process. Agent-based modelling and simulations have been used, in order to guide real participation processes and predict the results of an interaction process, and group multi criteria decision-making methods, to help taking consistent decisions based on several judgment criteria. The results of the research should help decision-makers and practitioners in dealing with multiple stakeholders and complex decisions and guiding the participation process.
Article
Full-text available
In transport planning several actors with conflicting objectives are involved in the decision-making process. Though public participation is fundamental to legitimate a transport plan, some inconsistencies may arise when individual preferences are aggregated into a collective decision. In this work, we reproduce the process of collective preference ranking among plan alternatives using agent-based simulations of the opinion dynamics on groups of stakeholders linked in typical social networks. The results show the efficacy of interaction and the relevance of the network topology to find a transitive and shared collective preference ranking.
Article
Full-text available
Stakeholder theory has been a popular heuristic for describing the management environment for years, but it has not attained full theoretical status. Our aim in this article is to contribute to a theory of stakeholder identification and salience based on stakeholders possessing one or more of three relationship attributes: power, legitimacy, and urgency. By combining these attributes, we generate a typology of stakeholders, propositions concerning their salience to managers of the firm, and research and management implications.
Conference Paper
Full-text available
Agent-based simulations show their potential in many context of transport management in presence of unusual demand, such as airport passenger terminals, railway stations, urban pedestrian areas, public buildings, street events or open space exhibitions, where management or control by related authorities and public safety are strongly affected by spatial geometry and crowd behavior. We illustrate these ideas with an example based on the simulation of people visiting and evacuating a museum, which offers an excellent test environment for simulating a collective behavior emerging from local movements in a closed space. The model we apply is developed within a programmable modeling environment, NetLogo, designed for simulating time-evolution of complex systems. We verify the existing emergency plan for building evacuation, for different demand patterns such as visiting group size and inter-arrival times, and we compare it with alternative evacuation strategies looking for the optimal one. In this respect, we further demonstrate the effectiveness of agent-based simulations in finding emergent results difficult to be predicted.
Article
Full-text available
Starting from an original framework based on four dimensions and thirteen objectives of sustainable urban mobility policies, this paper advocates the selection of a core set of performance indicators founded on a participative procedure. Citizen participation and stakeholder involvement is made possible by a national sample survey and a deliberative multi-criteria analysis, respectively. Such a procedure is applied to the Italian case and it shows that the set of indicators based on citizen evaluations radically differs from that based on stakeholders’ opinions: citizens are more oriented towards reducing private transport costs, air pollution and traffic accidents; stakeholders are more in favour of improving car-free accessibility and reducing the consumption of land and public space generated by urban mobility. For further testing at a local scale, a more articulated procedure is proposed in order to increase the role of citizens and to help generate unequivocal results.
Article
Full-text available
When does opinion formation within an interacting group lead to consensus, polarization or fragmentation? The article investigates various models for the dynamics of continuous opinions by analytical methods as well as by computer simulations. Section 2 develops within a unified framework the classical model of consensus formation, the variant of this model due to Friedkin and Johnsen, a time-dependent version and a nonlinear version with bounded confidence of the agents. Section 3 presents for all these models major analytical results. Section 4 gives an extensive exploration of the nonlinear model with bounded confidence by a series of computer simulations. An appendix supplies needed mathematical definitions, tools, and theorems.
Article
Stakeholder theory has been a popular heuristic for describing the management environment for years, but it has not attained full theoretical status. Our aim in this article is to contribute to a theory of stakeholder identification and salience based on stakeholders possessing one or more of three relationship attributes: power, legitimacy, and urgency. By combining these attributes, we generate a typology of stakeholders, propositions concerning their salience to managers of the firm, and research and management implications.
Book
This book provides a comprehensive and systematic presentation of the mathematical models for the simulation of transportation systems and the methodologies for the analysis and design of these systems. Theoretical and operational aspects are presented in a rigorous and exhaustive framework, addressing a broad range of applications performed by researchers and practitioners. Topics are presented with an increasing level of detail and complexity. In this 2nd edition the author extends and generalizes the contents of the previous edition entitled "Transportation Systems Engineering: Theory and Methods" published in 2001. In addition to entirely new material dealing with the recent developments in the field, the text has been revised to simplify the presentation. The Bibliography has also been expanded significantly. All the topics are presented with simple numerical examples as well as with references to real-life applications. The material is structured so as to offer different levels of complexity and different reading paths based on the reader’s needs. It is suitable for graduate level courses on transportation analysis and planning. Prerequisites include a basic knowledge of calculus, optimization techniques, probability theory and statistics. Due to the breadth and depth of topics covered, the book will also serve as an excellent reference guide for researchers, teachers and practitioners.
Public Engagement and Transportation Planning: Some evidences from Italy In: http://www.sidt.org/sidt2011/wp-content/uploads/SIDT-2011_PL2_02_Public- Engagement-and-Transportation-Planning Strategic management: A stakeholder approach
  • E Cascetta
  • F Pagliara
Cascetta, E. and F. Pagliara (2011). Public Engagement and Transportation Planning: Some evidences from Italy. SIDT, National Conference and International Seminar, Venice. In: http://www.sidt.org/sidt2011/wp-content/uploads/SIDT-2011_PL2_02_Public- Engagement-and-Transportation-Planning.pdf Freeman, R. E. (1984). Strategic management: A stakeholder approach. Pitman, Boston.
Sustainable transport visions: expert and non-expert stakeholder perspectives on sustainable transport International Conference on Whole Life Urban Sustainability and its Assessment
  • L Whitmarsh
  • A Haxeltine
  • M Wietschel
Whitmarsh, L., A. Haxeltine and M. Wietschel (2007). Sustainable transport visions: expert and non-expert stakeholder perspectives on sustainable transport. International Conference on Whole Life Urban Sustainability and its Assessment, Glasgow. In: http://download.sue-mot.org/Conference-2007/Papers/Whitmarsh.pdf Wilensky, U. (1999). NetLogo. Center for Connected Learning and Computer Based Modeling. Northwestern University, Evanston, IL. In: http://www.ccl.northwestern.edu/netlogo LE PIRA Michela; IGNACCOLO Matteo; INTURRI Giuseppe; GAROFALO Cesare; Pluchino Alessandro; RAPISARDA Andrea
Successful transport decision-making. A project management and stakeholder engagement handbook
  • J Kelly
  • P Jones
  • F Barta
  • R Hossinger
  • A Witte
  • A C Wolf
Kelly, J., P. Jones, F. Barta, R. Hossinger, A. Witte and A. C. Wolf (2004). Successful transport decision-making. A project management and stakeholder engagement handbook. In: http://www.civitas-initiative.org/docs1/GUIDEMAPSHandbook_web.pdf
Sustainable Urban Transport Planning SUTP Manual. Guidance for stakeholders. Rupprecht Consult
  • M Wolfram
  • S Buhrmann
Wolfram, M. and S. Buhrmann (2007). Sustainable Urban Transport Planning SUTP Manual. Guidance for stakeholders. Rupprecht Consult, Germany. In: http://www.ccl.northwestern.edu/netlogo