Heterogeneity in the Driver Behavior: An Exploratory Study Using Real-Time Driving Data

Table 1

Summary of the literature on driving style classification.

Studies	Objective	Instruments	Sampling frequency	Driving features	Nature of the data
Murphey et al. [9]	Driving style classification—calm, normal, aggressive, no speed	—	—	Acceleration and jerk features from 11 standard driving cycles	Pre-defined thresholds of safety critical events
Johnson and Trivedi [11]	Driving style classification—aggressive, nonaggressive	Smartphone sensors	1 Hz	Speed, acceleration, rotation
Eboli et al. [17]	Driver behavior classification—safe, unsafe	GPS	1 Hz	Speed, acceleration
Feng et al. [18]	Driver classification—aggressive, nonaggressive	CAN bus	10 Hz	Speed, acceleration, jerk
Mantouka et al. [20]	Trip classification—six categories from aggressive to nonaggressive	Smartphone sensors	1 Hz	Speed, acceleration, percentage of mobile usage while driving
Feng et al. [24]	Simulating driving styles—aggressive, normal, defensive	On board data logger (OBD-II), radar	1 Hz	Speed, headway, pedal movement, gear position
Chen et al. [25]	Driving style modeling—aggressive, moderate, careful	GPS	10 Hz	Speed, acceleration
Bagdadi [28]	Detection of safety critical braking events	GPS	10 Hz	Acceleration, jerk
Constantinescu et al. [10]	Driver classification—six categories from very aggressive to nonaggressive	GPS	1 Hz	Speed, acceleration	Abstract driving features
Hong et al. [15]	Driving style classification—Aggressive, nonaggressive	Smartphone sensors	2 Hz	Self-reports, speed, acceleration, engine RPM
Kalsoom and Halim [30]	Driver classification—slow, normal, fast	Driving simulator	0.1 Hz	Speed, acceleration, frequency of brakes, horns, gear, indicators
Chen and Chen [19]	Driving style classification—three groups	GPS	10 Hz	Speed, acceleration	Driving profiles
Chen et al. [21]	Driving style modelling—Individual driving habit graphs	Driving simulator	—	Speed, acceleration, steering wheel angle, throttle actuation
Chen et al. [22]	Identifying individual driving patterns	GPS	1 Hz	Speed, acceleration
Higgs and Abbas [23]	Driving pattern identification—30 patterns	GPS, cameras, radar	10 Hz	Speed, acceleration, yaw rate, range, lane offset
Li et al. [31]	Driving-style classification—low-risk, moderate-risk, high-risk	CAN bus, cameras	10 Hz	Speed, acceleration, headways, video data
Li et al. [32]	Identifying descriptive driving patterns—four patterns	CAN bus, cameras	10 Hz	Speed, acceleration, headways, video data