Unintended veering out of a travel lane is rarely documented with sufficient precision. Yet this could be a cause of many collisions such as the multi-fatal collision on I-75 near Gainsville Florida that occurred on January 3, 2019. Using multiple, synchronized video cameras Gorski Consulting has documented such details at four sites along Highway 401 in Southwestern Ontario.
Preliminary data was provided in a news item posted on December 20, 2018 on this Gorski Consulting website. That data was based on videotaping at three sites. A second videotaping session was conducted at one of the sites (Westminster Drive) thus resulting in four sessions and three sites. The original table is reproduced below.
Since that posting a fifth videotaping session was conducted at a fourth site (Kenesserie Road). Additional processing of the data along with the addition of the fifth videotaping session has led to adjustments to the original table. These adjustments involved removal of some observations where the veering was simply too minimal. For example, the criterion for inclusion of a veering occurrence was with the observation of the right front tire and whether the full width of that tire was observed to move outside of the white-painted edge line of the lane. Upon closer viewing of the video we observed several instances where the veering did not reach that threshold. Thus those observations were excluded. This has resulted in the adjusted table of data below.
The original impetus to gathering this veering data was because, in our initial session at Westminster Drive on October 30, 2018, we were drawn to the number of vehicles that appeared to be veering to the right, out of the right lane. Now looking at the table above and comparing that session to the others, the percent of veering vehicles (2.13%) is higher than the others.
Another interesting result is that two sessions were conducted outside of the normal work week. The November 23rd session at Dillon Road was on the Friday of the US Thanksgiving Day weekend when most truckers would be spending the holidays at home. This session produced 1.62 percent veering.
The December 2, 2018 session at Westminster Drive occurred on a Sunday and it produced 1.32 percent veering, which is lower than the original session that occurred on a weekday. A further anomaly is that 50% of the veering observations (13) originated from non-trucks. This is much higher than in the other four sessions where non-trucks produced 2, 2, 4 and 0 % of the veering observations. While the number of data points is small it leads one to consider what might have caused this result other than random variance.
A factor to consider is the varying volume of heavy trucks in the five videotaping sessions. This variance in volume is shown in the table below.
It can be seen that the two non-working day sessions of November 23 and December 2, 2018 the percentage of heavy truck traffic was 25.8 and 20.6 respectively and this is far below the percentages for the other three sessions that occurred on working days. The average truck traffic percentage for those other three working day sessions (Westminster on Oct 30th = 40.4 %, Graham Road on November 5th = 48.6 % and Kenesserie Road on December 19th = 48.0 %) was 45.7%. In other words, the percentage of trucks during non-working days was about half of what it was during working days. If veering is more prominent with heavy trucks then it should affect the results of veering frequency between the non-working days and the working days.
So comparing the data between the two sessions at the Westminster site, the Oct 30th session was a working day whereas the Dec 2nd session was a non-working day. So we should expect to see less truck veering observations in the Dec 2nd session and this is what occurred. In the Oct 30th session there were 37 observations of veering trucks whereas in the Dec 2nd session there were only 13 such observations.
The explanation for the larger number of non-truck veering in the Dec 2nd session could be that there were substantially more non-trucks on the road. However it is too early to tell.
The seemingly larger number of veering trucks at the the Dillon Road session of November 23rd is not readily explainable at this time. That session was part of the Thanksgiving Day period and it can be seen that the percentage of trucks on the road was low (25.8%) and the overall travel volume was low (only 1234 vehicles). Again, these are very low numbers of data points and the results may simply be due to random variance. Never-the-less with further work and larger data some conclusions could be made in the future.
The procedures described here represent a methodology where details of vehicle motions can be documented and evaluated. The Ontario Ministry of Transportation has maintained a number of traffic counting stations for many years along Highway 401. These magnetic loop devices should be capable of providing 24-hr, 7-day-a-week data on traffic volumes, truck versus non-truck composition, gaps between traffic units and vehicle speeds. However it is highly unlikely that these stations would be capable of documenting the numbers of vehicles veering out of their lanes, how those veering actions take place and what circumstances exist before, during and after those veering actions. The methodology used by Gorski Consulting is capable of documenting such details and more. This may be particularly important when there is a need to understand how some types of collisions occur on the highway and this could lead to changes that could reduce the severity and frequency of those instances.