Driving performances assessment based on speed variation using dedicated route truck GPS data

Ying Li, Li Zhao, Laurence R Rilett

Research output: Contribution to journalArticle

Abstract

It was hypothesized that a driver is not safe when travel speed is too high and also not necessarily safe when travel speed is too low. Based on this hypothesis, this paper studied the risky driving performances by measuring speed variations of a driver's recurrent trips in two perspectives: 1) driver profiles, which scored the risk on-road driving of each driver and 2) driving patterns, which reflected the risk speed patterns of a type of drivers. The proposed method was tested on a 30-day global positioning system (GPS) dataset, collected from 100 trucks. The study first split the raw dataset into trips and finds the most repeatedly traveled route. Next, the frequency and amplitude of the speed variations from trips of each truck are calculated to establish driver profiles. A risk score is used to rank the truck drivers, i.e., a higher score indicates that the truck driver is more likely to conduct risky driving performances. All trucks are featured in four pre-defined driving patterns according to the different types of speed variations. The geospatial speed distribution of several trucks is manually examined from the raw dataset to verify the results. The contribution lies in providing a method to evaluate a driver's risk performance through mass truck GPS data. The proposed method would help for monitoring on-road risky driving performances in large fleet management and also providing knowledge about driving styles among drivers which would be beneficial in study driver assistant system.

Original languageEnglish (US)
Article number8682043
JournalIEEE Access
Volume7
DOIs
StatePublished - Jan 1 2019

Fingerprint

Trucks
Global positioning system
Truck drivers
Monitoring

Keywords

  • dedicated route
  • Driving pattern
  • global positioning system
  • risky driving performance
  • speed variation
  • trajectory

ASJC Scopus subject areas

  • Computer Science(all)
  • Materials Science(all)
  • Engineering(all)

Cite this

Driving performances assessment based on speed variation using dedicated route truck GPS data. / Li, Ying; Zhao, Li; Rilett, Laurence R.

In: IEEE Access, Vol. 7, 8682043, 01.01.2019.

Research output: Contribution to journalArticle

@article{8384106cf91547a29464c28c853f23e0,
title = "Driving performances assessment based on speed variation using dedicated route truck GPS data",
abstract = "It was hypothesized that a driver is not safe when travel speed is too high and also not necessarily safe when travel speed is too low. Based on this hypothesis, this paper studied the risky driving performances by measuring speed variations of a driver's recurrent trips in two perspectives: 1) driver profiles, which scored the risk on-road driving of each driver and 2) driving patterns, which reflected the risk speed patterns of a type of drivers. The proposed method was tested on a 30-day global positioning system (GPS) dataset, collected from 100 trucks. The study first split the raw dataset into trips and finds the most repeatedly traveled route. Next, the frequency and amplitude of the speed variations from trips of each truck are calculated to establish driver profiles. A risk score is used to rank the truck drivers, i.e., a higher score indicates that the truck driver is more likely to conduct risky driving performances. All trucks are featured in four pre-defined driving patterns according to the different types of speed variations. The geospatial speed distribution of several trucks is manually examined from the raw dataset to verify the results. The contribution lies in providing a method to evaluate a driver's risk performance through mass truck GPS data. The proposed method would help for monitoring on-road risky driving performances in large fleet management and also providing knowledge about driving styles among drivers which would be beneficial in study driver assistant system.",
keywords = "dedicated route, Driving pattern, global positioning system, risky driving performance, speed variation, trajectory",
author = "Ying Li and Li Zhao and Rilett, {Laurence R}",
year = "2019",
month = "1",
day = "1",
doi = "10.1109/ACCESS.2019.2909572",
language = "English (US)",
volume = "7",
journal = "IEEE Access",
issn = "2169-3536",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - JOUR

T1 - Driving performances assessment based on speed variation using dedicated route truck GPS data

AU - Li, Ying

AU - Zhao, Li

AU - Rilett, Laurence R

PY - 2019/1/1

Y1 - 2019/1/1

N2 - It was hypothesized that a driver is not safe when travel speed is too high and also not necessarily safe when travel speed is too low. Based on this hypothesis, this paper studied the risky driving performances by measuring speed variations of a driver's recurrent trips in two perspectives: 1) driver profiles, which scored the risk on-road driving of each driver and 2) driving patterns, which reflected the risk speed patterns of a type of drivers. The proposed method was tested on a 30-day global positioning system (GPS) dataset, collected from 100 trucks. The study first split the raw dataset into trips and finds the most repeatedly traveled route. Next, the frequency and amplitude of the speed variations from trips of each truck are calculated to establish driver profiles. A risk score is used to rank the truck drivers, i.e., a higher score indicates that the truck driver is more likely to conduct risky driving performances. All trucks are featured in four pre-defined driving patterns according to the different types of speed variations. The geospatial speed distribution of several trucks is manually examined from the raw dataset to verify the results. The contribution lies in providing a method to evaluate a driver's risk performance through mass truck GPS data. The proposed method would help for monitoring on-road risky driving performances in large fleet management and also providing knowledge about driving styles among drivers which would be beneficial in study driver assistant system.

AB - It was hypothesized that a driver is not safe when travel speed is too high and also not necessarily safe when travel speed is too low. Based on this hypothesis, this paper studied the risky driving performances by measuring speed variations of a driver's recurrent trips in two perspectives: 1) driver profiles, which scored the risk on-road driving of each driver and 2) driving patterns, which reflected the risk speed patterns of a type of drivers. The proposed method was tested on a 30-day global positioning system (GPS) dataset, collected from 100 trucks. The study first split the raw dataset into trips and finds the most repeatedly traveled route. Next, the frequency and amplitude of the speed variations from trips of each truck are calculated to establish driver profiles. A risk score is used to rank the truck drivers, i.e., a higher score indicates that the truck driver is more likely to conduct risky driving performances. All trucks are featured in four pre-defined driving patterns according to the different types of speed variations. The geospatial speed distribution of several trucks is manually examined from the raw dataset to verify the results. The contribution lies in providing a method to evaluate a driver's risk performance through mass truck GPS data. The proposed method would help for monitoring on-road risky driving performances in large fleet management and also providing knowledge about driving styles among drivers which would be beneficial in study driver assistant system.

KW - dedicated route

KW - Driving pattern

KW - global positioning system

KW - risky driving performance

KW - speed variation

KW - trajectory

UR - http://www.scopus.com/inward/record.url?scp=85066959239&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85066959239&partnerID=8YFLogxK

U2 - 10.1109/ACCESS.2019.2909572

DO - 10.1109/ACCESS.2019.2909572

M3 - Article

VL - 7

JO - IEEE Access

JF - IEEE Access

SN - 2169-3536

M1 - 8682043

ER -