In the recent decades, we have witnessed an increase in the number of vehicles using the road infrastructure, resulting in an increased overload of the road network. To mitigate such problems, caused by the increasing number of vehicles and increasing the efficiency and safety of transport systems has been integrated applications of advanced technology, denominated Intelligent Transport Systems (ITS). However, one problem still unsolved in current road networks is the automatic identification of road events such as accidents or traffic jams, being inhibitor to efficient road management. In order to mitigate this problematic, this paper proposes the development of a technological platform able to detect anomalies (abnormal traffic events) to typical road network status and categorize such anomalies. The proposed work, adopts a complex event processing (CEP) engine able to monitor streams of events and detect specified patterns of events in real time. Data is collectively collected and analysed in real-time from loop sensors deployed in Slovenian highways and national roads, providing traffic flows. This prototype will work with a large number of data, being used to process all data, complex event processing tools. All the data used to validate the present study is based on the Slovenian road network. This work has been carried out in the context of the OPTIMUM Project, funded by the H2020 European Research Framework Program.
Approximate Semantic Event Processing in Real-time Linked Dataspaces
