Creating Emergency Management Training
Simulations through Ontologies Integration
Regina B. Araujo1, Rafaela V. Rocha1, Marcio R. Campos1, Azzedine Boukerche2,
1Programa de Pós-Graduação em Ciência da Computação - Universidade Federal de São Carlos (UFSCar)
São Carlos, SP, Brasil
2SITE, University of Ottawa
Ottawa, Canada

Abstract
Training simulations, which involve the collaboration of
multiple users (represented as avatars), sharing a common
virtual environment, are difficult to build, control and manage.
This paper describes an architecture to support non-
programmer emergency management trainers to rapidly create
different instances of powerful and complex training
simulations. The novel aspects of this architecture, that makes
it different from other related systems, are the innovative
techniques and concepts that are used. Events collected from
sensor networks deployed on physical environments subject to
emergency situations can be added to the simulation scenarios
being created. A set of ontologies was devised to create
powerful training simulation instances, such as different fire
classes, different fire fighting techniques, specific rescue
tactics, etc. A case study was implemented to validate the
architecture. The results show that this system can be a
powerful tool for the creation of complex training simulations.

1. Introduction

Simulation modeling, particularly those aimed at training
situations, such as emergency actions by Special Forces (e.g.,
police and fire fighters) are not trivial to build. By having
control over a simulation, in real-time, a specialist user, such
as a fire fighter commander, can add and/or remove existing
objects in the simulation, as well as change existing object
properties, making the application a flexible environment to
train human capabilities on different equipments and
situations. Response to emergency standard tactics must also
be followed by trainees.
In order to make it easier and faster for safety
specialists/trainers to model virtual environments for training
simulations, without the need for programmers, a supporting
architecture was devised, which uses a set of ontologies to
instantiate different training scenarios. The scenarios include
training on life saving, assets saving and equipment use in
different types of emergency: fire, explosions, gas or toxic
substance leaking. The ontologies are fed by safety specialists
– they are useful to structure the simulation entities (objects
that are important to the application) and the simulation model
itself. Real events can be used as part of the scenarios. The
events are collected from physical environments, which are
continuously monitored by sensor networks. These events are
submitted to different interpretation levels (aggregation, co-
relation, incident and accident interpretations) and recorded in
a database to be used for at least two purposes: post evaluation
of incidents and accidents and for the creation of simulation
scenarios as close to reality as possible.
This paper describes the devised supporting architecture
and the respective ontologies. The paper is organized as
follows: Section 2 introduces ontology concept and its use for
simulation modeling and analysis. Section 3 presents the five
emergency ontologies devised for training simulation creation
and execution. A description of our architecture to support the
creation and execution of training simulations is given in
Section 4. A case study is described in Section 5, followed by
related work in Section 6. Section 7 describes the current
status of the work, followed by Conclusions and Bibliographic
References.

2. Using Ontologies for Simulation Modeling
and Analysis

According to Chung [1], training simulations are also models
of existing or proposed systems, but contrary to typical
simulation models, the resources and operational policies
decision making are not made beforehand. Outputs are
observed not only at the end of the simulation run, but also
during it. This allows the users to see the impact of their
decisions on the simulation environment in real-time. In the
simulations created by our architecture, any safety specialist
(trainer) will be able to create simulation scenarios. Avatars,
objects and events can be added to the scenario so that the
trainees can be able to practice their knowledge making
decisions through their avatars. For that, specific emergency
preparedness and response rules for an emergency scenario
have to be followed by the trainees. In order to create these
scenario for different emergency situations (e.g., different fire
fighting techniques), an architecture was devised that uses a
set of ontologies as the knowledge base for the simulations.
According to Pérez [3], ontologies provide a domain
common vocabulary and define the meaning of its terms and
the relationship among them. In ontologies, the knowledge is
The 11th IEEE International Conference on Computational Science and Engineering - Workshops
978-0-7695-3257-8/08 $25.00 © 2008 IEEE
DOI 10.1109/CSEW.2008.52
373