An Event Driven Architecture for Decision Support
0 likes1,075 views
The document discusses an event-driven architecture for decision support related to water quality monitoring. It describes a use case of monitoring for cyanobacterial blooms in a reservoir in France. The architecture includes sensors to collect pH values, an event processing layer to analyze the data as it arrives, and a decision maker module to notify authorities if pH variation exceeds 0.5 in 24 hours, indicating a bloom. The proposed architecture aims to process measurements quickly as they arrive to support timely decision making.
An Event Driven Architecture for Decision Support
- 1. An Event Driven Architecture for Decision Support Thomas Everding & Theodor Foerster Geoinformatik 2011, Münster, Germany
- 2. GENESIS GENeric European Sustainable Information Space for environment FP-7 Project 2008 – 2011 Building generic collaborative information networks Based on ISO, OGC, W3C and OASIS standards http://www.genesis-fp7.eu/
- 3. Problem Decision making relies on (environmental) monitoring Public health affected Pure information access not sufficient Phenomena have to be detected and processed as fast as possible Event Driven Architecture data acquisition pre-processing dissemination
- 4. Use Case Water quality monitoring Drinking water supply and recreational area Cyanobacterial blooming in Summer
- 5. Villerest reservoir, France Images: Google Maps
- 6. Bloom Detection Cyanobacterial bloom is detected using remote imagery Indicator: variation of pH value in 24 hours modified from GENESIS D3100.3, p. 62f. Number Name Origin Units Data values Threshold 3 pH Local measurement pH units 0 to 14 See below 3b pH variation Internal pH units / day -14 to 14 Abs(pH variation) >= 0.5 3c pH status Internal Normal | abnormal If (3b >= 0.5) -> abnormal, else normal
- 7. Architecture Gateway Integration Layer Event Processing Raster Processing Raster Data Store Decision Maker pH values Sensor Alert Service Enterprise Service Bus Sensor Event Service Web Processing Service Web Coverage Service
- 8. Indicator: pH value variation >= 0.5 within 24 h ESP Algorithm Sensor Event Service 24h max - min low high >= 0.5 < 0.5 pH value max min dif dif high low critical pH variation
- 9. Benefits On the fly processing Measurements are processed as soon as they arrive Decision makers are notified Pluggable architecture Data & processes Enterprise Service Bus
- 10. Thanks for your attention! Thomas Everding & Theodor Foerster {everdingt; theodor.foerster}@uni-muenster.de ifgi.uni-muenster.de swsl.uni-muenster.de www.genesis-fp7.eu