The research subject of this dissertation is using traffic participants (drivers, vehicles and mobile devices inside vehicles) as sources of traffic data further used in providing dynamic navigation service and traffic management. Motivation for this approach to detecting and disseminating traffic data is found in inefficiency of adequatelly covering large road network with traditional traffic sensors. Traditional traffic sensors include inductive loops, cameras etc. which are static and permanently integegrated with traffic infrastructure during construction.
Participatory sensing concept is not new and is applied in the traffic domain form some years now. Widelly popular social network found their way into traffic domain too, and today there are manz commercially available social networks gathering drivers and allowing them to exchange information. This model of information exchange usually amounts to manual input of messages which are redistributed to other drivers by the system.
...Floating car Data (FCD) concept eliminates the driver from the information input loop. By anonymously collecting data about current speeds of a large number of vehicles information about traffic congestions can be constructed and used in various types of information systems. Extended Floating car Data (XFCD) concept, this dissertation deals with, expands dataset collected from traffic participants with data from various sensors inside vehicle. This dissertation especially focuses on one particular type of sensor commonly found today integrated with mobile devices used by drivers for navigation, accelerometer. Most important characteristic of this type of sensor is its capability to detect relevant vehicle maneuvers by recording force patterns that act on the vehicle (mobile device) during these maneuvers. GPS receiver traditionally used in FCD systems is typically not capable to detect these types of maneuvers. This dissertation demonstrates accelerometer data analysis methods localized on mobile device tasked with detecting relevant traffic events (vehicle maneuvers). This information collected from participating vehicles is further used in implemented prototype of proactive traffic information service that uses this data to timely warn drivers about hazardous traffic events and conditions existing on their navigation route. Localised analysis of accelerometer device on mobile devices used by the drivers is conditioned by the characteristics of accelerometer data, its sheer volume, and implemented prototype of proactive traffic information service acts preemptively by increasing drivers’ situational awareness of traffic conditions ahead of them.
Faculty:
University of Niš, Faculty of Electronic Engineering
