Showing papers by "Jarrod Trevathan published in 2015"
Journal Article•
TL;DR: This paper describes how the SWE framework can be extended to integrate disparate WSN sys-tems and to support standardized access to sensor data and introduces a web-based data visualiza-tion and statistical analysis service for data stored in the Sen-sor Observation Service by integrating open source technologies.
Abstract: Wireless Sensor Networks (WSNs) have been used in numerous applications to remotely gather real-time data on important environmental parameters. There are several projects where WSNs are deployed in different locations and operate independently. Each deployment has its own models, encodings, and services for sensor data, and are integrated with different types of visualization/analysis tools based on in-dividual project requirements. This makes it difucult to reuse these services for other WSN applications. A user/system is impeded by having to learn the models, encodings, and ser-vices of each system, and also must integrate/interoperate data from different data sources. Sensor Web Enablement (SWE) provides a set of standards (web service interfaces and data encoding/model specifications) to make sensor data publicly available on the web. This paper describes how the SWE framework can be extended to integrate disparate WSN sys-tems and to support standardized access to sensor data. The proposed system also introduces a web-based data visualiza-tion and statistical analysis service for data stored in the Sen-sor Observation Service (SOS) by integrating open source technologies. A performance analysis is presented to show that the additional features have minimal impact on the sys-tem. Also some lessons learned through implementing SWE are discussed.
7 citations
TL;DR: Results suggest the combination of online services and interoperability between disaster portals; and social networks can further enhance disaster management initiatives as 70.5% of the users were able to estimate the correct location of a disaster e.g. fallen power lines, fire.
Abstract: Disaster management that uses web-based technology to enhance user collaboration around disasters is an emergent field. A number of dedicated 'disaster portals' exist but they do not integrate large social networks such as Twitter and Facebook. These social networking sites can facilitate the analysis and sharing of collective intelligence around disaster information on a far greater scale by increasing accessibility to, and the use of, a disaster portal. This paper presents the 'Riskr' project, which applies a low-technological solution to creating a disaster portal fed by social networking messages. The system has been implemented using Twitter and tested by users to determine the feasibility. Results suggest the combination of online services and interoperability between disaster portals; and social networks can further enhance disaster management initiatives as 70.5% of the users were able to estimate the correct location of a disaster e.g. fallen power lines, fire.
6 citations
Journal Article•
TL;DR: In this article, the authors describe a practical problem where a relational database must record a series of events that occur over a given time span, without any modification to the underlying database management system.
Abstract: Current database practitioners cannot easily store information about transient phenomena using existing commercial off-the-shelf tools, because vendors do not support temporal database characteristics. This paper describes a practical problem where a relational database must record a series of events that occur over a given time span, without any modification to the underlying database management system. The database must be able to perform queries on the temporal information of an event occurrence. However, the database must also manage overriding events which take precedence over previously scheduled events, while maintaining knowledge of the overridden event. A basic approach to the problem is discussed, which is conceptually simple but cannot efficiently record the required information, nor can it cope with complex overriding event scenarios. We propose a second approach that models the events using time spans. The time span approach is then refined to create a third scheme that can efficiently handle complex overriding event scenarios. The two time span approaches result in significant storage gains compared to the first approach. However, the storage gains come with a trade-off in efficiency for data manipulation queries. The proposed approach is implementable with current relational standards (and SQL), and does not have the overhead and complexity of a full-scale temporal database.