Assessing the impact of sleep time on truck driver performance using a recurrent event model

Abstract

Driver fatigue is a major safety concern for commercial truck drivers and is directly related to the total hours of sleep prior to a working shift. To evaluate changes in driving performance over a long on-duty driving period, we propose a mixed Poisson process recurrent-event model with time-varying coefficients. We use data from 96 commercial truck drivers whose trucks were instrumented with an advanced in situ data acquisition system. The driving performance is measured by unintentional lane deviation events, a known performance deterioration related to fatigue. Driver sleep time and other activities are extracted from a detailed activity register. The time-varying coefficients are used to model the baseline intensity and difference among three cohorts of shifts in which the driver slept less than 7 hours, between 7 to 9 hours, and more than 9 hours prior to driving. We use the penalized B-splines approach to model the time-varying coefficients and an expectation-maximization algorithm with embedded penalized quasi-likelihood approximation for parameter estimation. Simulation studies show that the proposed model fits low and high event rate data well. The results show a significantly higher intensity after 8 hours of on-duty driving for shifts with less than 7 hours of sleep prior to work. The study also shows drivers tend to self-adjust sleep duration, total driving hours, and breaks. This study provides crucial insight into the impact of sleep time on driving performance for commercial truck drivers and highlights the on-road safety implications of insufficient sleep and breaks while driving. KEYWORDSdriving fatigue, penalized splines, recurrent event models, time-varying coefficients, truck driving safety 4096

Bayesian criterion‐based assessments of recurrent event models with applications to commercial truck driver behavior studies

“…It has been shown that the off-duty sleeping groups have different SCE risk patterns over the 11-hour driving time. 9,10 The number of nondriving tasks had a mean of 2.86 (SD = 2.08) and a range of frequency from 0 to 13. We considered three shift starting time groups: 667 (36.1%) shifts started in the morning (5:00 AM-9:00 AM), 768 (41.6%) shifts started in the daytime (9:00 AM-8:00 PM), and 413 (22.3%) shifts started at night (8:00 PM-5:00 AM).…”

“…We adopt the B-splines with degree of 𝜈, which are widely used in recurrent event models with nonparametric intensity functions. 9,10,12 Let h l denote the number of knots and B lh (t) represent the value of the h th B-spline at t, where h = 1, … , H l and H l = h l + 𝜈. Then, the smooth function 𝛽 lj (t) is expressed by a linear combination of the B-spline basis functions B lh (t)'s given by 𝛽 lj…”

“…Dummy variables are created for the categorical variables, including the restarting group, the driver type, the starting time group, and the sleeping group. Following the findings from previous studies on the same subject, 9,10 we assume that each off-duty sleeping time group has different patterns of intensity functions of the SCE risk as well as the on-duty breaks incidence. Driving at 1:00 AM to 5:00 AM is considered as a time-dependent covariate in the model.…”

“…4 Thus, recurrent event models with time-varying coefficients are preferred to model complex transportation data. 9,10 The Bayesian joint modeling approach for multitype recurrent event data has been developed in the recent literature 9,11,12 due to its flexibility to handle the complicated data structures.…”

“…For instance, the impact of off‐duty sleeping time before a driving task might change over a driving shift due to increasing fatigue over time 4 . Thus, recurrent event models with time‐varying coefficients are preferred to model complex transportation data 9,10 . The Bayesian joint modeling approach for multitype recurrent event data has been developed in the recent literature 9,11,12 due to its flexibility to handle the complicated data structures.…”