Using sensor ontologies to create reasoning-ready sensor data for real-time hazard monitoring in a spatial decision support system.

摘要

In order to protect at-risk communities and critical infrastructure, hazard managers use sensor networks to monitor the landscapes and phenomena associated with potential hazards. This strategy can produce large amounts of data, but when investigating an often unstructured problem such as hazard detection it can be beneficial to apply automated analysis routines and artificial intelligence techniques such as reasoning. Current sensor web infrastructure, however, is not designed to support this information-centric monitoring perspective. A generalized methodology to transform typical sensor data representations into a form that enables these analysis techniques has been created and is demonstrated through an implementation that bridges geospatial standards for sensor data and descriptions with an ontology-based monitoring environment. An ontology that describes sensors and measurements so they may be understood by an SDSS has also been developed. These tools have been integrated into a monitoring environment, allowing the hazard manager to thoroughly investigate potential hazards.