A Web portal to support Water Quality Monitoring through Wireless Sensor Networks has been developed as part of the InWaterSense project. It is essentially an integrated portal since it supports management of both, the observational stream data on water quality coming from wireless sensors – dynamic data, as well as of the data describing the wireless sensor network itself, its devices and the corresponding site allocation data – static data. Access is given to a wide range of individuals, from water experts to WSN engineers, to general public. Experts’ module infers statistics about water parameters given the experts’ data, as well as the expert’s rules evaluated by the portal. Furthermore, the portal is distinguished for its level of scalability: it allows adding with ease new components/functionalities, e.g., add certain new regulatory documents for water quality, and directly compare through the portal the data measured by sensors with the quality standard values as defined by those corresponding regulations.
An ontology for water quality management has been developed, named InWaterSense ontology. It is an OWL 2 ontology which bases on the SSN ontology developed by W3C Semantic Sensor Network Incubator Group (SSN-XG), and is publicly available here.
The ontology, as depicted in figure below, consists of four modules:
Ontology framework modules
Two use cases were approached to illustrate the usability of InWaterSense ontologies. In particular, a stream data scenario from the domain of surface water quality management and static data scenario from the domain of drinking water quality management. A simulated rivers quality data sample can be found here while a drinking waters quality sample can be found here .
The InWaterJessSense expert system to support ranking and identification of potential polluters has been developed. It leverages ontology modules in conjunction with Jess rules (http://www.jessrules.com/) to support water bodies classification based on different regulation authorities e.g., Water Framework Directive (WFD). As a production rule system, Jess supports closed-world and non-monotonic reasoning.
The Jess implemented architecture of our system and its related components for reasoning over ontologies are presented in the figure below. Namely, input data coming from sensors in their available format, say SQL, are transformed into RDF streams using D2RQ tool (http://d2rq.org/). RDF data streams are next imported into the core ontology. The set of rules for water quality classification based on WFD regulations are defined and may run against the knowledge base. Moreover, a set of rules for investigating sources of pollution by observing if eventual critical events appear are defined and may be activated. A simple user interface was developed using Java Swing (http://openjdk.java.net/groups/swing/), which offers a user to monitor water quality based on the WFD regulations and to eventually find the possible sources of pollution.
Jess implemented architecture for WQM
For instance, as a usage scenario, to classify water statuses, one may select the regulation authority i.e., WFD, select the water quality parameters which are to be monitored, and then press the button “Classify”. The JTextArea below the “Output” label serves for printing the classification results.
Jess implemented architecture for WQM
A software component of the Web portal to run on a smartphone and support registration and maintenance of devices in the sensing station Is being developed.