Citation

Xu W, He J, Jiang B. Transp. Res. F Traffic Psychol. Behav. 2024; 103: 273-289.

Copyright

(Copyright © 2024, Elsevier Publishing)

DOI

10.1016/j.trf.2024.04.009

PMID

unavailable

Abstract

The impact of roadside greenness on driving fatigue in real urban settings has been insufficiently investigated, presenting a critical knowledge gap for researchers, policymakers, professionals, and the public. In this onsite driving experiment, 34 urban residents completed seven driving tasks on different urban road routes in a randomized order with one-day intervals. A total of 238 tasks were conducted, each lasting an hour, assessing psychophysiological, visual, and muscular fatigue. A cardiovascular activity monitor (BioHarness) continuously measured the driver's heart rate, with lower rates indicating reduced psychophysiological fatigue. Visual and muscular fatigue were self-reported using a Visual Analog Scale questionnaire administered before, at the midpoint, and after completing the driving task. Deep transfer learning semantic segmentation analyzed road landscape characteristics and traffic conditions recorded from the drivers' view. Statistical analysis demonstrated that higher mean and variation in greenness significantly predicted lower psychophysiological fatigue after adjusting for multiple covariates. These results indicate that enhancing both the quantity and diversity of green landscapes along urban roads is vital for reducing driver's psychophysiological fatigue. This study reveals that roadside landscapes in urban settings are not trivial decorations, and they should be considered an essential component of transportation infrastructure.

Keywords

Deep learning technology; Driving fatigue; Green infrastructure; On-site experiment; Quantity and diversity; Urban transportation