A comparison of traffic crash and connected vehicle event data on a freeway corridor with Hard-Shoulder Running.

Abstract

Police crash reports have traditionally been the primary data source for research and development projects aimed at improving traffic safety. However, there are important limitations of such data, particularly the relative infrequency of crashes on a site-by-site basis in many contexts. Crash analyses often require multiple years of data and the use of such data for short-term evaluations creates challenges. Recently, connected vehicle (CV) event data have emerged as a promising means for addressing these limitations. CV events, which are reported when a vehicle engages in rapid longitudinal or lateral acceleration, can be obtained both at larger scale and in a timelier manner as compared to crash data. However, research as to the relationships between CV events and crashes is still in its nascent stages. This study examined the frequency of CV driving events and traffic crashes on a freeway corridor in Southeastern Michigan that operates with hard-shoulder running during periods of heavy congestion. This corridor uses the inside (left) shoulder as a temporary travel lane during peak periods and also provides dynamic advisory speeds based upon traffic congestion levels as monitored by microwave vehicle detection systems. Consequently, comparisons were made as to the general relationships of CV events and crashes with respect to traffic volumes, as well as whether the shoulder lane was open or closed. As the study was conducted during 2020, this also allowed for comparisons between each metric over the early stages of the COVID-19 pandemic. A series of analyses show strong correlation between traffic conditions along the corridor and the frequency of crash and CV driving events. Both crashes and CV events occurred more frequently during periods of congestion. However, significant differences were observed between crashes and CV events depending on whether the inside shoulder was open to traffic or not. Furthermore, the CV events were more reflective of changes in travel patterns that occurred following the introduction of travel restrictions in response to the COVID-19 pandemic.