An evaluation framework for traffic information systems based on data streams

Traffic information systems have to process and analyze huge amounts of data in real-time to effectively provide traffic information to road users. Progress in mobile communication technology with higher bandwidths and lower latencies enables the use of data provided by in-car sensors. Data stream management systems have been proposed to address the challenges of such applications which have to process a continuous data flow from various data sources in real-time. Data mining methods, adapted to data streams, can be used to analyze the data and to identify interesting patterns such as congestion or road hazards. Although several data stream mining methods have been proposed, an evaluation of such methods in the context of traffic applications is yet missing. In this paper, we present an evaluation framework for traffic information systems based on data streams. We apply a traffic simulation software to emulate the generation of traffic data by mobile probes. The framework is applied in two case studies, namely queue-end detection and traffic state estimation. The results show which parameters of the traffic information system significantly impact the accuracy of the predicted traffic information. This provides important findings for the design and implementation of traffic information systems using data from mobile probes.

► Framework for evaluation of traffic information systems based on data streams.
► Simulation of Car-to-X Communication to analyze benefit of Floating Car Data.
► Systematic evaluation of data stream mining algorithms for traffic applications.
► Two case studies: queue-end detection and traffic state estimation.
► Results reveal influential parameters for traffic information systems.