1079 
KNOWLEDGE-ORIENTED SENSOR WEB FOR DISASTER MANAGEMENT: FROM 
SENSING TO DECISION MAKING 
Xiangyu Si a „ Jonathan Li a , * , Zhijun Wang b 
department of Geography and Environmental Management, Faculty of Environment, University of Waterloo 
200 University Avenue West, Waterloo, Ontario, Canada, N2L 3G1 - (xsi, junli)@fes.uwaterloo.ca 
b XEOS Imaging Inc. 2750 rue Einstein, Quebec City, Quebec G1P 4R1, Canada - zhijun.wang@xeosimaging.com 
Commission IV, WG IV/8 
KEY WORDS: Sensor Web, Disaster Management, Emergency Response, Decision Support System 
ABSTRACT: 
The shocking consequences of disasters and natural hazards have already proven the urgency and necessity of disaster management 
and emergency response. Timely and updated information reflecting the newest disaster situation is extremely important for 
effective emergency response and efficient actions. While significant research has been demonstrated on the current research of 
disaster management, the problem of how to offer timely, even real time data and information to disaster managers and emergency 
responders is still far from solution. The emergence of sensor web does provide a promising way to solve this problem. However, 
most of current sensor web researches are mainly focused on data collection, sensing and real time monitoring of the disaster 
targeted, little research has been done on the process of decision making during disaster management and emergency responses. 
Especially, for improving the strategies of disaster management, in the context of sensor web, the inherent relationship between 
sensing and decision making really needs to be explored deeply. For an operationally effective disaster management, its sensing, 
monitoring, detection and decision making must be integrated seamlessly. In order to tackle with the limitations above, this paper 
presents a conceptual framework of knowledge-oriented sensor web for disaster management. Through the proposed architecture, 
disaster sensing and disaster decision making can be integrated closely by using multi-agent system and rule-based expert system. 
Thus, the existing sensor web can be potentially extended and greatly enhanced from mere disaster sensing and monitoring to a 
higher level, i.e., knowledge-oriented decision making, which will better support disaster management and emergency response. 
1. INTRODUCTION 
Disasters and hazards are absolutely unavoidable, though a 
great amount of efforts have been made in the domain of 
disaster management and emergence response (Kwan and Lee, 
2005; McQuaid, 2006; Napier, 2003). Every year, all kinds of 
disasters and hazards definitely bring great damage and huge 
loss to human lives and properties. No matter they are either 
nature hazards or man-made disasters (such as terrorism). The 
shocked aftermath of each hazard has already proven the 
urgency and necessity of disaster management. Timely and 
updated disaster information is extremely important for 
effective emergency response and efficient actions. Timely 
disaster data and information will absolutely help disaster 
managers make better decisions and take actions in time. 
Unfortunately, for current existing disaster management, though 
significant research and interesting results have been 
demonstrated, the problem of how to offer timely, even real 
time data and information to disaster managers and emergency 
responders, is still far from solution (Erharuyi and Fairbaim, 
2003; Fiorucci et al., 2005). 
With the rapid development of sensors technology, the 
emergence of sensor web does provide a promising way to 
solve this problem. Although the current research of sensor 
web is still at its early stage, the practical deployment and 
applications of sensor web have already shown its great 
potential (Delin, 2004; Gibbons et al., 2003; Liang et al., 2005; 
Paul, 2001). A certain fact is that more and more attentions 
have been paid into the research of sensor web (McCarthy et al., 
2008). 
It has been known that, for an operationally effective disaster 
management, sensing, monitoring and decision-making should 
be integrated seamlessly. However, most of current sensor web 
researches are mainly focused on data collection, real time 
sensing and monitoring of the disasters targeted, little research 
has been done on the process of decision making during disaster 
management and emergency responses. Especially, for 
improving the strategies of disaster management, in the context 
of sensor web, the inherent relationship between sensing and 
decision making really needs to be explored deeply. The reason 
of above limitations lies in the following facts: 1) the research 
of sensor web is still at its infancy stage, 2) disaster decision 
making involves complex relationships and interactions among 
a variety of disaster response entities. These entities include 
human and organizations, each of them owns a heterogeneous 
role in support of disaster management. For the research of 
disaster decision support, there are two fundamental questions 
need to be answered. The first question is how to represent 
disaster response entities and the interactions among them. The 
second question is how to capture and represent the expertise 
and professional knowledge owned by those entities. 
¥ 
In order to answer above questions, in the paper, a conceptual 
framework of knowledge-oriented sensor web for disaster 
management is proposed and discussed. For representing 
heterogeneous disaster response entities and their interactions, 
* Corresponding author: Jonathan Li, Associate Professor from Department of Geography, University of Waterloo, Canada.