Localized Processing and Analysis of Accelerometer Data in Detecting Traffic Events and Driver Behaviour

Bratislav Predic (University of Nis, Serbia)

Dragan Stojanovic (University of Nis, Serbia)

Abstract: Recent advancements in sensor technologies resulted in the development of sensors with small dimensions and with power consumption that is low enough to be embedded in various mobile devices and which is widely integrated in vehicles. Such sensors can be extensively used to detect real-time traffic events and situations of user/vehicle in context-aware mobile applications. This paper explores the usage of a large number of anonymous mobile devices already involved in the road navigation function as mobile sources of traffic information. Apart from collecting location and speed data, which is extensively used today to calculate average trip time per road segment, we are exploring possibility of using an acceleration sensor integrated with a mobile device in order to efficiently and timely detect critical traffic events and redistribute this information to other drivers through proactive traffic information system. Such a system would be capable of warning drivers of 'near-accident' situations enhancing their situational awareness and general safety.

Keywords: accelerometer data, extended floating car data, traffic information systems, vehicle tracking

Categories: H.4.3, H.1.2, H.3.5