Citation

Osafune T, Takahashi T, Kiyama N, Sobue T, Yamaguchi H, Higashino T. Int. J. Intell. Transp. Syst. Res. 2017; 15(3): 192-202.

Copyright

(Copyright © 2017, Holtzbrinck Springer Nature Publishing Group)

DOI

10.1007/s13177-016-0132-0

PMID

unavailable

Abstract

Detecting aggressive driving behavior is essential for safe transport systems as it leads to the awareness of risks of accidents. Using smartphone-equipped sensors would be promising approach considering the penetration ratio to the consumers. In this paper, we have used a large dataset of accelerometer readings collected by smartphones of drivers. The experiment was performed to explore the accident risk indexes which statistically separate the safe drivers and risky drivers. By the statistical analysis, it is found that the frequency of acceleration exceeding 2.4 m/s2, that of deceleration exceeding 1.4 m/s2, and that of left acceleration exceeding 1.1 m/s2 separate the safe drivers and risky drivers. The classifier using these three criteria achieves 70 % classification accuracy and 83 % detection accuracy of risky drivers.

Language: en