Browsing by Subject "Floating Car Data"

Driving speeds regardless of vehicle type are a part of almost everyone’s daily lives. The subject has been widely studied and many algorithms for determining optimal routes exist. A novel data source for this type of research is GPS-collected Floating Car Data. As positioning enabled devices have become increasingly abundant, the collection of huge amounts of data with locations, speeds and directions has become vastly more common. In this master’s thesis, I examine a type of Big Data -set of car speeds within the Helsinki area through three different viewpoints. First, I examine the driving patterns described by the distribution of data on different kinds of roads and time periods. Second, I focus on one variable, intersection density, and determine the effect it has on the change in speed and whether it is possible to conduct statistical analysis for the data. Last, I analyze the steps needed to take in order to fully utilize the variables of the data within the road network system. The results indicate that while there are clear differences in changing speed within road classes, the differences are not as clearly described by road class as they are by speed limit. Also, time of day has a clear effect where times of congestion are distinguishable. While among all road classes the mean driven speed is below the speed limit, on larger roads the mode is above the speed limit. I prove that it is possible to find numerous variables that depict speed change through novel Floating Car Data. Focusing on intersection density, the result is that at highest, within the Helsinki area, intersection density represents around eight per cent of change in speed compared to speed limit. As a final result, a method to viably use linear Floating Car Data to research intersection density and its effects is developed. As a mediate step and a side result, a workflow of modifying road network layers into segments between intersections is produced.