The association between crashes and safety-critical events: Synthesized evidence from crash reports and naturalistic driving data among commercial truck drivers. (May 2021)

• Record Type: Journal Article
• Authors: Cai, Miao; Yazdi, Mohammad Ali Alamdar; Mehdizadeh, Amir; Hu, Qiong; Vinel, Alexander; Davis, Karen; Xian, Hong; Megahed, Fadel M.; Rigdon, Steven E.
• Is Part Of: Transportation research. Volume 126(2021)
• Publication Date: 2021-05
• Subjects: Truck -- Naturalistic driving studies -- Safety-critical events -- Crashes -- Injuries -- Fatalities; Transportation -- Periodicals; Transportation -- Technological innovations -- Periodicals; 388.011
• DOI: 10.1016/j.trc.2021.103016
• Languages: English
• ISSNs: 0968-090X

Abstract

Highlights: Safety-critical events are associated with the risk of crashes and injuries. The positive association held when stratified by business unit and driver type. We analyzed over 1.4 billion pings from 31, 828 commercial truck drivers. Rolling stability has the highest correlation with crashes. Our code, data, and analytic methods are available to promote future research. Abstract: The past decade has witnessed continuous growth of naturalistic driving studies (NDSs). When analysing NDS data, safety-critical events (SCEs) are commonly used as surrogates for safety, since actual crashes are very rare. However, the association between SCEs and crashes is not consistent in previous studies and has not been verified among commercial truck drivers. Based on routinely collected kinematic data from 31, 828 truck drivers in a large commercial trucking company, this paper examines the association between four types of SCEs (headway, hard brake, collision mitigation, and rolling stability) and crashes, as well as injuries and fatalities. Compared to existing studies on the subject, that are based on up to about 2 million miles driven, our study involves an estimated 2.3 billion miles driven. Bayesian negative binomial models were applied to examine the association between three outcomes (crashes, injuries, and fatalities) and the four SCEs. It was found that a unit increase in the number of any type of SCEs per 10, 000 miles was associated with 8.4% (95% credible interval (CI): 8.0–8.8%) increase in crashes per mile and 8.7% (95% CI: 4.8–13.6%) increase in the number of injuries per mile. The increase was different in different types of SCEs: 3.3% (95% CI: 2.6–4%) for headways, 8.1% (95% CI: 7.5–8.7%) for hard brakes, 50.4% (95% CI: 41.4–60%) for rolling stability, and 22.2% (95% CI: 19.8–24.5%) for collision mitigation. The results are consistent when stratified by different business units and driver types. This study provides statistically strong and robust evidence that SCEs are positively associated with crashes and injuries among commercial truck drivers. NDS and kinematic data routinely collected by trucking companies provide a promising opportunity for future data analytic research.