The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B4. Beijing 2008
Mitigation
4.2 Interactions between DRA and SA
Figure 3. Creation of Disaster Management Knowledge Base
4. KNOWLEDGE-ORIENTED SENSOR WEB
Current research of sensor web is primarily focused on the
sensing and data collection. Specific to the domain of disaster
management, most of applications only use sensor web as a
sensing tool to collect disaster monitoring data and detect
disaster events. In fact, as a complete cycle of disaster
management and emergency response, sensing, monitoring and
detection are only parts of it. Response, countermeasures and
actions taken by disaster managers and emergency responders
play same critical roles for the success of disaster management.
In order to meet this need, in the following section, a
conceptual framework of knowledge-oriented sensor web for
disaster management is proposed.
4.1 Knowledge-Oriented Sensor Web: Conceptual
Framework
The aim of proposed framework (Figure 4. ) is to integrate
disaster sensing with disaster decision making, thus existing
sensing-based sensor web can be enhanced and upgraded into
knowledge-oriented sensor web.
The following abbreviations are used in Figure 4.
P: Prediction
ED: Event Detection
DRA: Disaster Response Agents
SA: Sensor Agents
SP: Sensors Platform
SM: Science Models
R&R: Retasking & Reconfiguration
A&R: Actions & Responses
D&H: Disaster & Hazard
When implementing a knowledge-oriented sensor web for
disaster management based on the proposed conceptual
framework, the most important aspect of this kind of sensor
web is the interactions between disaster response agents (DRA)
and sensor agents (SA). Without the implementation of the
interactions between them, a knowledge-oriented sensor web
for disaster management cannot be truly achieved. The details
of interactions between DRA and SA are shown in Figures 5.
Disaster Response Agents
Figure 5. Interactions between DRA and SA
The interactions between DRA and SA lay a solid foundation
for the implementation of knowledge-oriented sensor web. In
fact, from the technical perspective, the interactions between
them depend on the specific network environment. When
operating through proprietary network such as wireless sensor
network, the interactions can be carried out directly using