The aim of this Diploma Thesis is to identify the critical driving parameters affecting speeding using data from smartphones. The data used for the analysis were collected from two databases over a period of 6 months, containing routes of more than 200 drivers in naturalistic driving answers to a questionnaire from 100 of the above drivers. Subsequently, Poisson and log-linear regression models were developed correlating the driving characteristics and the answers to the questionnaire with speeding. Finally, four models were developed for speeding percentage prediction; one general model and three model for each type of road (urban, rural and highway). The results demonstrate that the number of harsh accelerations, the percentage of mobile use and the distance of the trip affect the speeding percentage and are correlated with the aggressive behavior of the drivers. Furthermore, increased average acceleration does not always lead to breaking speed limits. Finally, male drivers tend to drive faster in comparison with women.
| ID | ad104 |
| Presentation | |
| Full Text | |
| Tags | big data, driver behaviour, naturalistic driving, speed, statistical modelling, telematics |
