Keywords
remote sensing, water risk management, process oriented system design, process networks, decision support systems, hardware software co-design, in-situ sensors, formal models, communicating sequential processes
Start Date
1-7-2008 12:00 AM
Abstract
Water monitoring and associated risk management requires a large number of in-situmonitoring stations. These stations measure water quality parameters and relay the measurementsto a data centre. From a data processing point of view, such systems constitute distributed processingnetworks. In present water monitoring systems and in a wide range of other environmentalmonitoring systems the network aspect stops at the node or in-situ monitoring station (IMS) level.This paper extends the networking idea from the system level to the IMS design level. We modelan in-situ monitoring station as a network of communicating sequential processes. For this networkwe establish the fundamental system properties such as: security, stability and functionality.In a second step we translate this functional model into an implementation model with the sameproperties. The designer has the freedom to translate particular network processes into either hardwareor software components. These components communicate over well defined interfaces, thisensures composability. In the practical part we follow the two step design approach to integratea new sensor into an existing IMS. First we create a process network with the desired properties.Then this process network is translated into hardware and software components for a hybridprocessor. Finally, the functionality of the resulting component network is tested.The process oriented design approach leads to more secure, reliable, and functional systems.Furthermore, this design method simplifies the integration of In-situ Monitoring Stations into thewater risk management system, because the complete system functionality is designed accordingto networking principles.
Towards Process Oriented Environmental Monitoring Systems for Intelligent Risk Management
Water monitoring and associated risk management requires a large number of in-situmonitoring stations. These stations measure water quality parameters and relay the measurementsto a data centre. From a data processing point of view, such systems constitute distributed processingnetworks. In present water monitoring systems and in a wide range of other environmentalmonitoring systems the network aspect stops at the node or in-situ monitoring station (IMS) level.This paper extends the networking idea from the system level to the IMS design level. We modelan in-situ monitoring station as a network of communicating sequential processes. For this networkwe establish the fundamental system properties such as: security, stability and functionality.In a second step we translate this functional model into an implementation model with the sameproperties. The designer has the freedom to translate particular network processes into either hardwareor software components. These components communicate over well defined interfaces, thisensures composability. In the practical part we follow the two step design approach to integratea new sensor into an existing IMS. First we create a process network with the desired properties.Then this process network is translated into hardware and software components for a hybridprocessor. Finally, the functionality of the resulting component network is tested.The process oriented design approach leads to more secure, reliable, and functional systems.Furthermore, this design method simplifies the integration of In-situ Monitoring Stations into thewater risk management system, because the complete system functionality is designed accordingto networking principles.
