E.I. Vlahogianni, G. Yannis, J.C. Golias

Pages: 107-120

Abstract
Data from a naturalistic Powered Two Wheelers (PTW) driving experiment are exploited using principal component analysis to identify the critical PTW driving characteristics and their combinations with the final aim to reveal useful insights on the PTW driving patterns emerging on the road. Data are collected in different road environments, ranging from urban road arterials to suburban roads. The analyses reveal three prevailing PTW driving patterns: i. Acceleration, ii. Maneuvering, and iii. Braking. These patterns are observed in both inside and outside urban areas and regardless of the lighting conditions (daylight, dusk, and night). Nevertheless, although acceleration and maneuvering patterns contain the same driving variables in all driving cases examined, braking is conducted in a varying manner with respect to the type of the area and the time of day.

Keywords: multi-dimensional signal processing; driving profile; naturalistic riding study; powered two wheelers; big data