Gorski Consulting has collected data regarding the change in volume and composition of traffic on Highway 401 since the COVID-19 pandemic has taken effect.
In the fall of 2018 Gorski Consulting conducted a number of 2-hour, videotaping sessions along Highway 401 between London and Tilbury, Ontario. Two of these sessions occurred on Tuesday, October 30th and Sunday, December 2nd.
Subsequently another videotaping session was conducted on Wednesday, March 25th, 2020 at a time when social distancing and various closures took effect throughout Ontario. In the 2018 sessions the numbers of vehicles travelling westbound past the Westminster Drive overpass were tabulated along with the numbers of large trucks and buses. This was also done during the March 25, 2020 session. The results of these three sessions are shown in the table below.
Previous articles were posted to the Gorski Consulting website discussing the 2018 videotaping sessions. One of the observations from these sessions was that the volume of heavy trucks on Highway 401 was greatly reduced on weekends and holidays. This reduction can be seen in the above table. The October 30th session occurred on a weekday (Tuesday) and this resulted in observations of 738 westbound trucks, whereas the December 2nd session occurred on a Sunday and it resulted in observations of only 417 trucks. This is a 45 % percent reduction in truck traffic over weekday totals.
What is interesting is the data for March 25th, which was a weekday (Wednesday) and therefore we should have observed heavy truck traffic similar to the Oct 30th data (738 total trucks). Instead we observed only 417 heavy trucks. This is particularly revealing because the March 25th data was obtained during rush hour, between 1530 and 1730 hours. The October 30th session commenced at about 1320 hours and thus should have been positioned at a time of day when traffic volumes should have been less than rush hour. So the finding of the reduced numbers of heavy trucks in the March 25th session is even more significant.
We can also observe the total number of westbound vehicles in the two 2018 videotaping sessions, 1827 and 1965. These totals include all westbound traffic regardless of whether it is heavy truck traffic or passengers, pick-up trucks, SUVs, vans, etc. Yet the total westbound vehicles in the March 25th session was only 1228. This is a reduction of about 33 to 37 %.
The US-Canadian border was closed about a week before the March 25th session. Only essential truck traffic was being allowed to pass through the border. Any personal trips by private citizens were disallowed. The Westminster Drive location on Highway 401 should represent all those heavy trucks that are travelling toward the US border at the Windsor-Detroit crossing. Furthermore the Westminster Drive location on Highway 401 is just west of the Highway 402 separation from Highway 401. So all those vehicles that might be travelling toward the Sarnia area and its border crossing to the US would be removed from the totals.
This dataset is very small compared to the vast data that is collected by the Ontario Ministry of Transportation (MTO) at its traffic counters which are located at every interchange of Highway 401. Yet the MTO data is not publicly available so the public remains in the dark about what effects are being experienced due to the COVID-19 pandemic. The Gorski Consulting data is a flashlight, shining into this enormous, dark cave.
At a time when citizens of Ontario and eastern Canada are shut down in their homes, they are reliant on the movement of goods to keep them fed and safe. When the volume of goods being carried by heavy trucks is reduced by 45% we need to consider what effects this may have over the long term. What goods are not being delivered and will there be shortages that could cause a chain reaction affecting other parts of the economy?
Certainly in the realm of road motor vehicle safety the reduction in all traffic volumes is likely to result in reductions in collisions. That may be so but there is one hiccup in this good news. When traffic volumes are low, those persons who like to travel above the speed limit are likely to increase their speed because there is less interference with their actions. Thus we could see an increase in speeds in that segment of the population of drivers and a possible increase in the number of high-speed, single-vehicle collisions. It remains to be seen whether those speed increases can be detected in any data and how this might change the collision statistics.
The superior design of expressways such as Highway 404 in Toronto should not result in a fatal impact to a bridge pillar. Yet, from the scant description provided by the OPP on their Twitter account this is what reportedly occurred. The collision likely occurred in the last few hours, perhaps in the early morning hours of March 26, 2020, and its specific location has not been identified.
The OPP description noted “SB vehicle entered centre ditch, went airborne and struck bridge support”. On a modern expressway that should contain large traffic volumes, Highway 404 should contain barriers that control the motion of vehicles entering a “centre ditch”. Such barriers and other roadside features should not cause a vehicle to go airborne. And those who designed Highway 404 should have created protections from impacts to immovable objects such as bridge pillars. This is an unsafe situation that should be documented by the OPP no different than if they had encountered an impaired, speeding or distracted driver.
Ensuring the public’s safety does not mean that we only capture armed bank robbers in green shirts. We capture all bank robbers regardless of what colour of shirt they are wearing because the colour of the shirt should be irrelevant. On that basis the OPP need to be protecting the public from all road safety risks, not just selected ones.
Gorski Consulting has now completed the documentation of 560 instances of driver response to a red signal turning green at numerous intersections throughout London, Ontario. A dashcam was used in a vehicle that was positioned behind the drivers who were stopped at a red traffic signal. When the signal turned green the delay in time was noted when the brake lights of the vehicles extinguished.
The overall average delay for all 560 observations was 0.70 seconds. However there were a number of additional facts to consider. For example, 110 drivers, or 19.6 %, reacted by more than 1.0 seconds. And 86 drivers, or 15.4 %, reacted in 0.4 seconds or less. It is generally understood that a person should not be able to react to a simple stimulus such as a green light in less than 0.40 seconds therefore this suggests that many of these drivers were already releasing their brake pedal before the signal turned green. This was evidenced further by noting that vehicles were observed to crawl forward while the signal was still red even though the brake light was still illuminated. Thus this confirmed that these drivers were releasing their pressure on the brake pedal before the signal turned green, likely in anticipation of the signal turning green.
As some drivers took as much as 3.0 to 4.0 seconds before extinguishing their brake lights this affected the overall average of 0.70 seconds. Similarly some vehicles could be seen extinguishing their brake lights 3.0 to 4.0 seconds before the signal turned green and this also affected the overall average. In an attempt to clarify the data, we conducted a frequency count to see where the most observations were clustered and this is shown below.
Response greater than 0.2 but less than 0.3 seconds = 11 observations
Response greater than 0.3 but less than 0.4 seconds = 31 observations
Response greater than 0.4 but less than 0.5 seconds = 44 observations
Response greater than 0.5 but less than 0.6 seconds = 54 observations
Response greater than 0.6 but less than 0.7 seconds = 55 observations
Response greater than 0.7 but less than 0.8 seconds = 37 observations
Response greater than 0.8 but less than 0.9 seconds = 23 observations
Response greater than 0.9 but less than 1.0 seconds = 17 observations
Thus, looking at the above breakdown, the most common response delay is between 0.5 and 0.7 seconds.
It needs to be noted that this response time is different from the response time to start a vehicle in motion, which should take a longer time. Thus, once a brake signal is extinguished the driver’s foot needs to lift completely off the brake pedal and transition over to the accelerator pedal. Then the driver needs to depress the accelerator pedal and a further delay in the vehicle mechanical and electronic systems occurs before the vehicle can begin moving forward. Other research has indicated that delay to start a vehicle in motion is likely in the range of 2.0 seconds.
This research shows how much difference exists in drivers, some who are very conscious of the traffic signal and want to begin accelerating forward as quickly as possible, versus those drivers who do not appear to be attentive to the status of the traffic signal and wait a substantial time before accelerating forward. Those late responders, approaching 20% of the total, may be an indication of the extent of driver lack of attention to basic cues around their driving task. Although there is much discussion about driver distraction by in-vehicle instrumentation and cell phones, there is no way tell, from this research, why these late delays occurred. It might even be an indication of the general level of inattention that exists in this 20% of the overall population of drivers.
When drivers fail to detect the activation of a green signal while stopped the opposite might also be true: that these same drivers might not be attentive enough when they are approaching a green signal to detect when the signal has turned to amber and red. Hopefully drivers are a little more attentive when their vehicle is in motion and they are approaching a traffic signal but how much more attentive they are needs further study. It is likely that many drivers may not purposely travel through a red traffic signal but that their lack of attention may be the cause of that action. Thus certain penalties for such infractions may be of lessened effect when the solution might be in studying a driver’s level of vigilance and ability to focus their attention. Such deficiencies are not generally tested in new drivers and certainly not in drivers who have been on the road for many years.
Although the most severe, head-on collisions also provide the greatest opportunity to use the structure of vehicles in riding down the destructive forces of that impact. This was clearly shown in the OPP photo shown above, relating to a serious head-on that occurred on Hwy 86 in Huron County, northwest of Listowel, Ontario on March 19, 2020.
It is helpful that the OPP frequently provide a photo or two during their notifications of serious collisions in Ontario. While a single photo is far from adequate, it is better than none, as was frequently the case only a few years ago. Even the availability of a single photo can say a lot about what has taken place. While we do not have the advantage of examining the vehicles, the collision site, or even additional crucial photos, we have the advantage of having examined hundreds of similar collisions over the past four decades.
Firstly, in earlier times, before the advent of electronic stability control (ESC) and similar technology, the patterns of vehicle crush could help the investigator in determining the pre-crash circumstances of a head-on collision. As someone who spent many long hours in a towing compound measuring the crushed state of vehicles, and then returning to the office to creature large-scale diagrams of those crushed shapes, the patterns of vehicle crush and specific “points-of-mutual-contact” said a lot about what happened in the few seconds before impact. Now so-called “experts” simply examine the contents of an event data recorder (“black box”) and they know precisely what happened because the box tells them so. But they need to know nothing about the physical evidence on the vehicle. That places a lot of blind faith in the validity of the numbers in the box.
In the past, many head-on collisions could be grouped into two classes: 1) Those with crush patterns that were identical, look book-ends and 2) Those with one vehicle exhibiting additional crush along its side, primarily its right side.
In the first instance the book-end patterns of crush (exhibited in the two vehicles shown in the above photo) occurred in passing situations where a driver misjudged a return to their own side of the highway resulting in an off-set impact with maximum crush at the left (driver’s side) of each vehicle. The offset involved direct contact to 50 % or less of the left portion of the front ends. This would result in counter-clockwise rotation of both vehicles and spin-off to final rest, generally on opposite sides of the highway. When these collisions approached high severity the probability of structural intrusion into the driver’s space increased making it more difficult to protect drivers, even when seat-belted. As time as moved on Federal compliance tests and even those by the insurance industry (Insurance Institute for Highway Safety – Highway Loss Data Institute) caused improvements to be made in such offset frontal impacts resulting in less structural intrusion and better ride-down as a “system” of safety features worked together. Air bags were a giant improvement when they became depowered in 1998. So the seasoned investigator who came upon a collision site could have an instant understanding, although preliminary, of what might have led to this type of crash. But now-a-days things are a little different.
In the second instance, the pattern of crush involving one of the vehicles with direct damage to its side, was indicative of a loss-of-control situation by the vehicle with the damage to its side. Very often such events occurred not to far from a highway horizontal curve and sometimes the drop-off of the right side wheels off of the right pavement edge, before commencing a counter-clockwise rotation into impact. While many such collisions involved this single and direct rotation, many of these events involved multiple rotations, clockwise and counter-clockwise, as the driver made attempts to regain control before finally sliding into the front end of another vehicle in the opposing lane. Unfortunately the cause of these events was difficult to locate because the initiation of the loss-of-control occurred very far from the actual impact and because there was no physical evidence generated the substantial distance in the early portion of the driver response. Very often signs of a vehicle travelling onto the right shoulder were destroyed as good intentioned passers-by stopped on the shoulder and destroyed that evidence with their tires. The evidence was destroyed further by police and other emergency vehicles as they came to a stop on the shoulder to attend to the victims.
Severe, head-on collisions are also characterized by short, post-impact travel distances to final rest, often less than 10 metres. This is differentiated from angle impacts at intersections where those travel distances may be in the range of 25 to 50 metres.
Obviously there are many varieties of head-on collisions besides the two that have been mentioned. A detailed exploration of crush patterns is rarely a part of a police reconstructionist’s duties as now-a-days independent reconstructions requested by insurers and other agencies are a rarity. As mentioned earlier, reconstructions have become more and more reliant on the downloading of event data from a vehicle’s “black box” and little focus on collecting and understanding of the details of the physical evidence.
It the advent of electronic stability control (ESC) and other advances, the motions of a vehicle approaching impact are increasingly controlled by vehicle sensors and computer logic and thus the two patterns of head-on collisions become murky in their interpretation. The unfortunate reality is that, while in most instances the computer logic leads to an improvement in safety, at times that logic is limited such at collisions become more severe. The photo above provides an opportunity to discuss this problem even though it may not have occurred in this specific instance.
The oddity about the evidence shown in the above photo is that the direct contact damage is almost fully across the front ends of both vehicles. This is odd because, in the past, it was difficult to achieve. Such occurrences were created when one driver or both made very poor judgments about how to avoid the collision. This would occur for example when one of the driver’s judgments was badly impaired or in instances of poor visibility or poor road surface conditions. Most head-on collisions occurred with 60% or less of frontal offset and it was difficult to find a collision with more. So in the present instance, if this had occurred 30 years ago, we would view it as rare. But those vehicle sensors and computer logic changes that.
Now, as drivers attempt to introduce massive steering, braking and acceleration inputs, the vehicle computers increasingly become the “Big Brother” making the final decision as to how the vehicle’s pre-impact motion will be altered. The computers believe that the greatest benefit to the driver is to keep the vehicle pointing in the direction where the vehicle is travelling. And that makes since if the collision is inevitable and we want to use the full length of the vehicle’s front distance to crush and protect the occupant. But in this attempt some collisions may become more severe as there is greater overlap between vehicle front ends and the impact force is closer to each vehicle’s centre-of-gravity. This is the Princess Diana syndrome that has been discussed before, wherein the force of the tunnel pillar impact was substantially toward her vehicle’s centre-of-gravity. So what we see in the above photo is that, with the very large overlap there is very little post-impact rotation and very little post-impact motion to final rest. Essentially all of the pre-impact kinetic energy was dissipated by the vehicle crush and none was dissipated by post-impact factors.
Even the most inexperienced novice could recognize that the centre of the debris located between the vehicle front ends is where the point of impact occurred. General observations that the two vehicles were of relatively similar mass, indicate they were likely travelling at similar speeds. The deployment of side curtain air bags would suggest some lateral force which may not be unusual given that such impacts cannot be precisely at 180 degree closing velocities.
So was this collision affected by the vehicle’s sensors and computers? An assessment of the recorded data would provide that answer but we will never be provided with that information. While this is kept secret amongst the very few police and insurance investigators officially assigned to the case, the public also has a stake in this. It is the public who will be involved in the next fatality-producing impact yet the public is not entitled to know what it is that will kill them. Oh yes, your doctor will tell you the drug he is giving you will make you better but you are not entitled to know when that drug contains a poison.
It is hard to have traffic fatalities when there is no traffic. This may be what is in store in the short term as the Corona virus turns society on its head.
A domino effect can be initiated when the Corona virus response causes large numbers of businesses and events to close down. This also results in heavy hits to the economy that could be the start of permanent shut downs in jobs, defaults on debt and similar catastrophes that occurred after the 1929 stock market collapse.
We may be seeing fewer motor vehicle collisions in the foreseeable future but that may not be a good thing.