Abstract
Traditional approaches to evaluating and predicting safety issues in traffic systems are via crash records. However, considering the characteristics of scarcity, inconsistency, inaccuracy, and incompleteness of crash records, conclusions and recommendations drawn purely based on crashes have limitations. Tire skid marks are considered an indication of some safety hazards, and it could have good potential to be used as surrogates for crashes. By collecting and analyzing the data based on selected arterial and freeway segments in the Reno-Sparks area in northern Nevada, a methodology was developed to categorize different tire skid marks. Sliding window and linear regression techniques were applied to determine any correlation between tire skid marks and crashes. The analyses indicated that there was a relatively strong linear correlation between skid marks and crashes on freeway segments.
Highlights
Journal of Advanced Transportation probability of higher-than-average crash rates is likely
Preventative actions could be taken when the surrogate measures indicate a high likelihood of future crashes. e main advantage of using surrogates is that they occur more frequently than crashes; the data on surrogate measures can be collected over a shorter time frame than crashes but with similar statistically significant measures as crash data
Saunier and Sayed [11] extracted vehicle trajectory data from videos recorded at intersections and quantified the probability of collision using traffic conflicts identified by time to collision (TTC)
Summary
In the event that a video skid mark did not touch the bottom of the screen, either the lower side edges of the video were used or the approximate location of the time point (based upon pavement markings or other identifiable features in adjacent lanes) was used to determine the time index. Smaller skid marks that appeared on video for less than one second were recorded with the same start and stop indices. It should again be noted that these calculated locations are estimates only, based on time-space relationships derived from the study videos. Because crashes are more likely to occur at the end of a skid mark, the time index where the trailing end of the skid mark appeared was the point used as the basis for estimating the location of each mark.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
