Content uploaded by Pierrick Duhamel
Author content
All content in this area was uploaded by Pierrick Duhamel on Oct 26, 2018
Content may be subject to copyright.
Learning Role-Playing Game Scenario Design for Crisis
Management Training: From Pedagogical Targets to
Action Incentives
Pierrick Duhamel1, Sylvain Brohez1, Christian Delvosalle1, Agnès Van Daele1 and
Sylvie Vandestrate1
1 University of Mons, Place du Parc, 20, 7000 Mons, Belgium
pierrick.duhamel@umons.ac.be
Abstract. Emergency and crisis management requires, from operatives and de-
cision-makers, specific knowledge that cannot be acquired through theoretical
course or real-life practice only [1]. Besides, developing practical exercises
adapted for agents and their needs is even more difficult when the system where
they operate is complex [2]. It is therefore necessary to develop such exercises
according to both rigorous and flexible methodology.
Since 2015, the Expert’Crise project, funded by the European Social Fund,
has organized seven exercises mainly in hazardous chemical companies. During
such exercises, trainees play their own role in their usual working place. Hence,
arrangements must be made [3] to isolate trainees from real environment and
establish exercise diegesis [4]. Through a trial and error experience, we devel-
oped a design methodology for crisis management Learning Role-Playing Game
[5] scenario.
Scenario design starts from trainees’ statement of requirements leading to ped-
agogical targets, chosen from an existing classification [6]. Then, because emer-
gency sequences are often similar, we developed a framework for our scenario
based on the narrative storyline of J. Campbell [7], describing the steps of an
emergency sequence. Nevertheless, because pedagogical targets change depend-
ing on exercise, this storyline varies and includes dedicated “situation-tasks” that
target competences previously identified. These situation-tasks aim to “force”
trainees to do actions under special circumstances through serious game interface
[8], and using its gameplay and diegesis. However, “situations” may not lead
obviously to a “task”, and incentives must be introduced to help trainees perform-
ing the task, unlike disturbances [9] that can also be added.
Keywords: Education and Training, Safety and Health, Crisis management, Se-
veso.
2
1 Introduction
1.1 Background
Despite the progress of major accident prevention since the 1970’s and accident such
as those of Flixborough or Seveso, there are still industrial disasters. Proof, if needed,
that prevention alone is not enough and safety needs emergency management when a
disaster occurs. Yet emergency and crisis management requires specific competences
that cannot be acquired through theoretical course or real-life practice only [10]. They
must therefore be trained through practical exercises adapted to their environment,
working habits, education, and professional needs. Moreover, European regulation
known as “SEVESO 3 Directive” requires hazardous companies to test their emergency
planning every three years, for instance with exercises.
Nevertheless, developing such exercise is a complex task that requires dedicated
skills [11]. This kind of training already exists but is mainly oriented to public organi-
zation [12] or outsourced to consulting firms. Therefore, because organizing exercise
in industrial plant is complicated, expensive, and comes with uncertain outcomes, com-
panies may be reluctant to organize such exercises [13]. To test the emergency man-
agement with exercises, SHE managers need a methodology to easily set up internal
exercises matching with their goals.
Launched in 2015 and funded by the European Social Fund, the Expert’Crise project
develops emergency and crisis management training for hazardous industries and crit-
ical infrastructures. These training are based on a pedagogy combining theoretical
courses and immersive practical exercises. Through a trial and error experience, and
the literature from different fields such as pedagogy, dramaturgy, game design, and
crisis management, a design methodology for crisis and emergency management Learn-
ing Role-Playing Game [5] scenario was developed.
1.2 Context of the Development of the Design Methodology
Eight walloon SEVESO companies have participated to one of the six exercises orga-
nized between 2015 and 2017. Eight other exercises are currently under preparation for
2018. Companies, their environment and the type of exercises [8] proposed vary from
an exercise to another and cover a wide scope of situation.
Exercises were held on industrial sites and used rooms, tools and communication
devices available to operatives and decision-makers both in their work-life and during
emergency. This configuration allows immersive situation for trainees without destabi-
lizing them [14]. In addition, it provides an emergency system test for companies in
accordance to European regulations [15]. Exercises are mainly functional [8], focused
on decision-maker functions [16]. The operational part is often simulated through the
control of information flow entering the crisis room. Therefore, these trainings target
members of the crisis management team and key persons in the warning chain.
Exercises rely on material arrangement and human organization. Human organiza-
tion refers to facilitators regulating the exercise [9] near trainees (or at distance) and
observers consigning what happened during the event. Experimental device is com-
posed of cameras, microphones, projectors and speaker that allow immersing trainee in
3
the diegesis and, on another hand, capturing multimedia flows that are lived-streamed
to distant facilitators to help them adjusting the scenario and saved to complete observ-
ers’ notes for later analysis.
The framework of the training offered remained the same throughout the project:
one to six theoretical course modules, briefing of some or all the trainees, simulation,
“hot” debriefing right after the simulation, analyse of the simulation [17]. and “cold”
debriefing. Nevertheless, inside this framework, each step have evolved to take previ-
ous mistakes, bias, limits and suggestions of improvement into account.
2 Methodology: from Pedagogical Target to Action
Incentive
Like games, Learning Role-Playing Games need rules, medium, context and players
[18]. Rules and medium are defined by the category of game but context and players
vary depending on the company. Therefore, before considering developing a scenario
for the exercise, these two elements must be properly defined with the company and
will allow to build the proper diegesis – i.e. what is true in the context of the simula-
tion – for the exercise. Then the methodology is split in two main parts. The first one
analyzes needs and wishes, and emergency system of the company to propose a peda-
gogical and organizational framework for the exercise. The second part designs exer-
cise content inside the previously defined framework.
These two parts are not isolated from each other, and the development may step from
one to the other depending on the progress of the pedagogical engineering and the re-
activity of industrial contact person. In addition, the methodology is iterative so some
steps are looped several times.
Fig. 1. Structure of the emergency exercise design methodology
2.1 The Framework: Defining Needs to Specify the Scope of LRPG
After the company agreed to participate to the training, a meeting is scheduled to deal
with needs and expectations of the firm. This first meeting aims to have a first scope
of the training company expectations including target audience identification, resources
that can be used and operational (e.g. evacuation or intervention exercise, warning
4
chain test…) and/or global (e.g. improving reactivity, internal or external communica-
tion…) objectives. These objectives formulated by the company must be interpreted to
fit emergency and crisis exercises pedagogical targets identified in literature [6].
The second step is about processing emergency plan and organization documents
into a mental picture of how the emergency system works. It appears emergency plans
from European Seveso companies are similar [3] with separated operational on site
management and strategic management in a crisis room. This common structure is con-
venient and allows building mental picture from an existing framework. A second meet-
ing with the contact person is then usually needed to clarify or verify some points. This
meeting may be held with other persons of the companies to have a better description
of their role and seize the difference between the prescriptive plan and what they would
do during an emergency according to their experience, their working procedures and
habits. Therefore, it is possible to have a representative picture of how the emergency
system could be expected to work during an emergency.
Once the company’s needs and its emergency system is understood, the third step
is about establishing borders of the sub-system tested during the exercise and figuring
out how this sub-system will evolve during the simulation and especially how it will
interact with the defined borders. This step lead to define, considering the target audi-
ence, a second category of audience: the peripheral audience. Those persons are not
directly aimed by the training but they play an important role in emergency plan and
are direct interlocutor to target audience so they have to participate, if possible. Never-
theless, because they are not directly targeted, facilitators mentor them by telling them
how to interact with the target audience. Therefore, they can be seen as an input/output
interface for facilitation. In the same way, indirect interlocutor such as medias, political
stakeholders, administrative authorities and emergency services are identified accord-
ing to company’s needs and facilitators simulate them during the exercise through
phone, mail or other means [9]. To this end, a sheet showing exercise’s phone numbers
and corresponding stakeholders is displayed in rooms where target audience is expected
to be. Explanations related to this sheet as well as other immersive device and interfaces
are given to trainees before the simulation or, if it is not possible, during the exercise
by facilitators.
Once all trainees are identified, areas where they are supposed to be can be consid-
ered. The crisis room, the disaster area, the guard post and the control room are usually
the main relevant areas for chemical industries exercises. Nevertheless, according to
activities of the company and its wishes, areas may be added or removed. In this last
situation, a specific interface is set up according to the target of the exercises defined
previously. It can be a reduced model simulating on-the-ground operations, leaded in-
terviews with intervention leader or sub-crisis unit managed by a facilitator. These sim-
ulations aims to give to the target audience an immersive experience including the cor-
rect information flow and realistic interactions with stakeholders as represented in fig-
ure 2.
5
Fig. 2. System borders audiences and others organizations simulation
An analogy can be made with theatrical scenography. Indeed, this step consist in build-
ing a scenery where trainees will evolve as actors and, the same way scenography use
cardboard environment to improve immersion, it may be relevant to use trick such as
smoke-producing device, alert horn or other sound and light device to simulate events
and strengthen realistic feeling of trainees, actors but also spectators of the simulation.
Nevertheless, these tricks do not have to be fully realistic with a homogenous diegesis
and can only be representative or symbolic with a heterogeneous diegesis [4], the same
way an accessory – such a hat or glasses – is enough to understand a same actor plays
different character.
The last border, which should be defined, is the duration of the exercise. They usu-
ally last between one and three hours for logistic reasons but longer exercise could be
planned.
Fig. 3. Audience, simulation device and information flow
Therefore, once interfaces between trainees, environment and facilitators are de-
fined, context and players are set and the LRPG is operable. Therefore, an exercise draft
can be submitted and the scenario development initiated.
6
2.2 The Scenario : Designing an Interactive Story to Achieved
Pedagogical Target
Once the framework is defined, scenario can be developed, it is the fourth step. A
scenario explains the diegesis of the exercise and develops the sequence of input to the
target audience, which should lead it to pedagogical targets [19]. Based on targets ex-
pressed by the company, knowledge and competences [20] are selected from an existing
list [6]. At this point, situations staging competences are considered. The relevance and
the ease to stage of each situations are then evaluated to choose the more adapted situ-
ation for the exercise. This first association between knowledge and situation are proto-
“situation-task” [21]. A situation-task are the central part of these exercises and aims to
“force” trainees to do an action (the task) under special circumstances (the situation)
through serious game interface and using its specific gameplay. The task is a mean to
involve trainees in a reasoning process harnessing knowledge targeted [22]. Once all
knowledge, skill, and competences that will be aimed by the exercise are identified,
they are grouped into pedagogic bloc associated with a situation, eventually including
first information input ideas. These pedagogical blocs are integrated in a framework of
the evolution of a crisis in industrial environment inspirited by the hero’s journey [7]
in the same way as in game design or scriptwriting methodologies [23]. The hero’s
journey is organized around the transition from the common world to the unnatural/un-
common world where the protagonist of the story becomes a hero. A parallel can be
made with the transition from normal to crisis mode. Therefore, the different step of the
hero’s trip can be adapted to fit with crisis situation process as seen in table 2.
Pedagogical blocs with prototype idea of scripting will be integrated in this frame-
work and, based on the length of the exercise, will give a first view of how the exercise
will process. The arrangement of each blocs is important because it will influence the
dynamics and the stress of the exercise in addition to encourage the resolution of some
problematics before other [19]. Further developments of “situation-task” is not possi-
ble before defining the “plot” of exercise, i.e. the accidental sequence.
The fifth step consist to select, with the contact person, the most adapted accidental
sequence for the exercise. The accidental sequence means the causes of the accident,
the accident, dangerous phenomenon associated, and people, environment, equipment
or structure affected [24]. Accidental sequence should lead to “situation-tasks” and
should be justifications for inputs helping trainee to do the wanted task.
At the sixth step, based on the hero’s journey chronology, exercise is divided in 15
minutes sequences assigned to one “situation-task” where input leading to the task will
be added as shown in table 1. Some sequence may not be related with a “situation-task”
especially in introduction, transition, conclusion or build-up period.
Table 1. Block chronology of the exercise
Time block
Situation-task
Pedagogical target
Incentive /
Perturbation
7
Table 2. Hero’s journey step and crisis management process comparison
Hero's journey step
Crisis management process [25]
Departure
The call to adventure
First step of the warning chain.
Warning
Refusal of the call
Disbelief or minimization of the crisis. Fear to leave
the normal operation mode.
Supernatural aid
There is no supernatural aid in crisis process. Nev-
ertheless, during exercises, facilitation can play this
role.
The crossing of the first
threshold
Awareness of the gravity of the situation. Emer-
gency plan "engagement" and reflex procedures.
Initiation
The belly of the whale
Information and action flooding. Difficulty to pic-
ture correctly the situation.
Crisis management
The road of trials
First decision-making and awareness of operational
difficulties.
The meeting with the
goddess
Meeting with emergency services and information
exchange.
Woman as a temptress
Temptation to not act anymore, letting all actions to
emergency services.
Atonement with the fa-
ther
Communication with authorities, medias and higher
hierarchic level.
Apotheosis
Expectation of change, improvement or the end of
crisis. Domino effect if any.
Return
Refusal of the return
Expectation of recurrence, domino effect or unex-
pected consequences. Stay in alert.
Back to the normal
The magic flight
Last communications to authorities and media with,
eventually press conference. Checklist verification.
Rescue from without
Other stakeholders close their crisis units and emer-
gency services leaves plant.
The crossing of the re-
turn threshold
Report and debriefing
Master of the two worlds
Crisis unit closing
Freedom to live
End of the sequence
For each time sequences, situation-tasks and pedagogical targets related are reported.
In this way, the purpose of each sequence and reasons why specific task were chosen
is not forgot. Then inputs leading trainees to do the task are chosen. Nevertheless, the
choice of relevant and efficient inputs is not easy. Indeed, they must lead to the task to
do but in a realistic and non-obvious way to keep trainees focused and in a “flow” state
8
[26]. It is the reason why proto-situation task should integrate staging ideas in the very
first step of scriptwriting. Each situation-task stages the resolution of a problem by par-
ticipants and can lead to different solutions as shown in figure 4. The situation is usually
implied by dangerous phenomenon or its consequences such as a wounded person or
damaged equipment, and the trainee’s reasoning process can be helped by incentive
(e.g. municipal authority asking a press statement) but can also be slowed down by
perturbation [21] keeping the simulation challenging for trainees.
Fig. 4. Flow state diagram of a situation-task reasoning process
Once inputs are chosen, the way they will be injected into the system is defined. It could
be trough a call, a direct interaction with someone from the peripheral audience or the
use of an immersing device such as a sound speaker playing firefighters’ siren. Direct
interactions between target audience and facilitation are carefully managed because fa-
cilitator may influence target audience and decrease relevancy of observation. Never-
theless, in some situation, it is not possible to fully separate facilitation and target au-
dience. In that case, interactions are limited – with only some fact presentation for in-
stance – and are taken into account into further analysis.
In the end of this step, a precise timing with a 5 minutes meshing is established and
inputs, recall, expected trainee’s reaction and facilitators recommended reaction are
specified in a table as shown in table 3. There is no need to over saturate trainees with
inputs if it is not the purpose wanted. Indeed, communication between participants dis-
patched inside the organization will be important independently of the number of input.
This precise table is then used to write the global script, which can be exposed to com-
pany’s contact person for validation.
Once the global script written, come the seventh and last step of the design method-
ology. Each input is integrated into the corresponding facilitator’s sheet with details
related to the medium and the context of injection. These sheets can be split in two
groups: those dedicated to mentoring peripheral public and give it a consistent – but
partial – view of what happen and how they are supposed to react, and those dedicated
to distant facilitation, simulating different stakeholder and controlling immersing de-
vices. Facilitation sheets can also include a question/answer part to help facilitator to
answer to possible questions from trainees. Nevertheless, these sheets cannot anticipate
9
all questions and reaction of trainees and facilitators should adapt the scenario to train-
ees’ reactions.
Table 3. Precise chronology and input timeline
3 Conclusion
This methodology allows to design efficiently emergency and crisis exercise for
chemical industries. Nevertheless, it still needs a high level of competences to choose
and arrange situations-tasks and therefore cannot be used by SHE manager. The next
step of development of the methodology will consist in identifying a limited number of
situation-task that can be staged in exercise to offer an user-friendly method for SHE.
References
1. Lagadec, P. : La gestion de crise - Outils de réflexion à l’usage des décideurs. MacGraw
Hill, Paris, France (1991).
2. Saad Noori, S., Comes, T., Schwarz, P. and Wang Y. : Behind the scenes of scenario-based
training: understanding scenario design and requirements in high-risk and uncertain envi-
ronments. In: Proceedings of the 14th ISCRAM Conference, 976 - 987. Albi, France (2017).
3. Duhamel, P., Brohez, S., Delvosalle, C., Dubois, L.-A., Van Daele, A., and Vandestrate, S. :
Le projet Expert’Crise ou la formation à la gestion de crise en milieu industriel par des exer-
cices de mise en situation : premiers résultats. In : Récents Progrès en Génie des Procédés,
110 (2017).
4. Morten, G. : Interaction: The Key Element of Larp. In :G. Morten, L. Thorup and M. Sander
(eds.), When Larp Grows Up, p.66-71. Frederiksberg: Projektgruppen KP03 (2003).
5. Mariais, C., Michau, F. and Pernin, JP. : A Description Grid to Support the Design of Learn-
ing Role-Play Games. Sage Publications, 43 (1), 23-33 (2012).
6. Lapierre, D. : Methode EVADE : Une approche intégré pour l’Evaluation et l’Aide au DE-
briefing. PhD thesis, Université de Nîmes, Nîmes, France (2016).
7. Campbel J. : The Hero with a Thousand Faces. Novato (CA), USA: New World Library
(1949).
8. Tena-Chollet, F., Tixier, J., Dandrieux, A. and Slangen, P.: Training decision-makers: Ex-
isting strategies for natural and technological crisis management and specifications of an
improved simulation-based tool. Safety Science, 97, 144-153 (2016).
Time
block
Situation-
task
Pedagogi-
cal target
Incentive /
Perturbation
Facilita-
tor sheet
Precise timing
14h00-
14h15
Warning
chain. Crisis
unit estab-
lishment
Proce-
dures
check
Imprecise alert
Weak signals
No feedback
from operators
Inter-
vention
chief
14h00: Small event
14h00/5: Information
check not successful
14h15: Warning con-
firmation. Major acci-
dent
10
9. Fréalle, N. and Tena-Chollet, F.: The key role of animation in the execution of crisis man-
agement exercise. In: Proceedings of the 14th ISCRAM Conference, 944 - 956. Albi, France
(2017).
10. Lagadec, P. : Les exercices de crise: pour des ruptures créatrices. In: La Lettre des Cindy-
niques, 34, 5-9 (2001).
11. Limousin, P. Contribution à la scénarisation pédagogique d’exercices de crise. PhD thesis,
Ecole Nationale Supérieure des Mines de Saint-Etienne, Saint-Etienne, France (2017).
12. Alberta Emergency Management Agency: Exercise design 100 (2012). http://www.aema.al-
berta.ca/documents/exercisedesign100sel63227.pdf last accessed 2018/04/09
13. European Commission’s Joint Research Centre and the Dutch Ministry of Social Affairs and
Employment: Enforcement of Seveso II: An Analysis of Compliance Drivers and Barriers
in Five Industrial Sectors: Key Points and Conclusions. (2008)
14. Cook, D.: Flight Operations (2015). https://www.tc.gc.ca/eng/civilaviation/publica-
tions/tp185-4-07-operations-4034.htm#instructor, last accessed 2018/04/09.
15. European Parliament and Council : Directive 2012/18/EU of the European Parliament and
of the Council of 4 July 2012 on the control of major-accident hazards involving dangerous
substances, amending and subsequently repealing Council Directive 96/82/EC (2012).
16. Lagadec, P. : Cellules de crise - Les conditions d'une conduite efficace. Les Editions d'Or-
ganisation, Paris, France (1995).
17. Vandestrate, S., Dubois, LA., and Van Daele, A.: Crisis Management and Simulation Train-
ing: Using Activity Chronicles to Analyse Crisis Managers’ Behaviours. Communication
presented at the 20th Congress of the International Ergonomics Association, Florence (2018)
18. Klabbers, J.H.G.: The gaming landscape: A taxonomy for classifying games and simula-
tions. Proceedings of the 2003 DiGRA International Conference: Level Up, 2, 54-68 (2003).
19. Limousin, P., Chapurlat, V., Tixier, J., and Sauvagnargues, S.: A new method and tools to
scenario design for crisis management exercise. Chemical Engineering Transaction, 53, 319-
324 (2016)
20. Winterton, J., Delamare-Le Diest, F. and Stringfellow, E.: Typology of knowledge, skills
and competences. Cedefop, Luxembourg (2005).
21. Tena-Chollet, F.: Elaboration d'un environnement semi-virtuel de formation à la gestion
stratégique de crise basé sur la simulation multi-agents. PhD thesis, Ecole Nationale Supé-
rieure des Mines de Saint-Etienne, France (2012).
22. Pastré P., Mayen P. and Vergnaud G. : La didactique professionnelle. Revue française de
pédagogie, 154, 145-198 (2006).
23. Schell, J.: The art of game design. Elsevier, Burlington (MA), USA (2008).
24. Debray, B. and Salvi, O.: ARAMIS Project: an integrated risk assessment methodology that
answers the needs of various stakeholders. In: WIT Transactions on The Built Environment,
82, 265-275 (2005).
25. Dautun, C.: Contribution à l’étude des crises de grande ampleur: connaissance et aide à ladé-
cision pour la sécurité civile. Sciences de l’environnement. PhD thesis, Ecole Nationale Su-
périeure des Mines de Saint-Etienne, Saint-Etienne, Franc (2007).
26. Schmidt, J.A.: Flow in Education. In: Peterson, P., Baker E., and McGaw, B. (Eds.), Inter-
national Encyclopedia of Education, 3ième ed., Elsevier Ltd, 605-611. (2010)
