Impact of Primary Incident Spatiotemporal Influence Thresholds on the Detection of Secondary Crashes

Document Type

Article

Publication Date

1-1-2019

Abstract

Incident management agencies have been investing substantial amount of resources to devise strategies to mitigate secondary crashes (SCs). Nevertheless, detection of SCs is not a straightforward process, as the definition itself is subjective; identification of SCs depends on how the impact area of the primary incident (PI) is defined. Both static and dynamic methods, the two most common approaches used to define the impact area of the PI, have serious limitations that restrict their practical applications. Although the dynamic method is proven to yield accurate results, applying it requires real-time traffic data which are only available on limited locations. On the other hand, the static method’s one-size-fits-all approach of using fixed spatiotemporal thresholds does not yield reliable results. This study explored the impact of PI spatiotemporal influence thresholds on the detection of SCs. To implement the study objective, both static and dynamic approaches were developed. The static method was based on predefined spatiotemporal thresholds, and the dynamic method was based on prevailing traffic speed data from BlueToad® paired devices. Comparison of SC frequencies identified using the static and dynamic methods showed that the static method consistently under and overestimated SC frequencies for smaller and larger spatiotemporal thresholds, respectively. The prevailing traffic conditions were found to play a crucial role in instigating SCs, as more than 75% of SCs occurred during congested traffic conditions. Use of varying spatiotemporal thresholds depending on the prevailing traffic conditions is expected to reduce the biases associated with the subjective thresholds used in the static method.

Publication Title

Transportation Research Record

Digital Object Identifier (DOI)

10.1177/0361198119849058

ISSN

03611981

E-ISSN

21694052

Share

COinS