This is the final in a series of four articles dealing with the death of a retired professor who died when attempting to cycle across Gainsborough Road in London, Ontario on September 5, 2020. Data has been collected to document the volume and speed motor vehicles, cyclists and pedestrians. The finalized tables are now available below.
The first table (below) shows the average speed of 585 eastbound vehicles that were videotaped while travelling within a 200 metre distance approaching the multi-sue path. The data is broken down into six 20 minute segments. These speeds include the large number of turning vehicles that existed in the first hour of documentation
We can compare the numbers of vehicles travelling straight through the site by looking at the last column “EB Straight Thru Vehs As % of All Vehs”. This shows that in the first hour the data included many vehicles that were turning onto Gainsborough from two driveways on the north side of the road. As noted previously there was some kind of commercial event taking place at an establishment on the north side of the road and this was drawing a large number of vehicles in and out of the driveways. This is why the percent of vehicles travelling straight through the site was only in the range of 65 to 80 in the first hour. However in the second hour we can see that the percentage of straight-through vehicles rose to about 92 to 98. This shows that the commercial event was completed and vehicles were no longer exiting the driveways in the second hour of videotaping. This is also reflected in the smaller number of total eastbound vehicles (91, 63 and 85) in the last hour versus the larger numbers (118, 101 and 127) in the first hour.
In terms of average speeds of eastbound vehicles, these speeds were reduced as vehicles approached the cycling path. There was also a trend for higher speeds in the second hour as the interference caused by the turning vehicles was removed.
We can recall that on the eastward approach to the site the maximum posted speed limit was 60 km/h. As eastbound vehicles approached the 200 metre marker a sign was posted indicated that the maximum speed was reduced ahead to 50 km/h. Then just before reaching the cycling path another sign indicated the maximum posted speed was 50 km/h. So drivers travelling through the site should have been reducing their speed from 60 to 50 km/h. We can see in the above table that average speeds were substantially higher than those posted maximums.
There is further reason for concern when we remove those observations where an eastbound vehicle was interfered with by other vehicles. An interfering vehicle was defined as any vehicle that was within five seconds ahead of the subject vehicle. Thus it was reasoned that a driver following such an interfering vehicle was not free to chose an independent speed. When all those observations of interference were removed from the above table the adjusted results are noted in the table below.
Comparing the data in the two tables the average speed of vehicles was increased when those 232 vehicles that travelled straight through the site and experienced no interference. This is particularly obvious in the first hour of observations where the interference caused by left-turning vehicles was present. If we look at the distance of 200 to 150 metres west of the cycling path the average speeds in the first hour were 66.14, 68.55 and 62.79 km/h as noted in the first table. But in the second table those average speeds rose to 68.08, 71.24 and 68.98 km/h. Similar results are apparent in the other distance segments.
Interference must be considered an influencing factor in the speeds of eastbound vehicles on Gainsborough Road. It is more likely that a cyclist or pedestrian might try to cross Gainsborough at times when there is no interference. In other words, it is unlikely that cyclists or pedestrians would try to cross when the gap between eastbound vehicles was five seconds or less. Thus it is important to examine the average speed of non-interfered vehicles rather than examining the total population of eastbound vehicles.
The numbers and percentages of eastbound speeding vehicles for the full two hours of documentation are summarized in the table below.
The above table shows that, out of the 232 eastbound vehicles that travelled without interference the numbers and percentage of speeders is as follows:
130 vehicles travelled at 72 km/h or greater, or 56.03 % of total
44 vehicles travelled at 80 km/h or greater, or 18.97 % of total
4 vehicles travelled at 90 km/h or greater, or 1.72 % of total
Even if the maximum posted speed was 60 km/h throughout the full distance of 200 metres west of the cycling path this would still indicate that about 19 % of non-interfered drivers chose to drive at 20 km/h or greater above the posted speed. But we know that these drivers were informed that they were approaching a 50 km/h zone thus the speed difference is even greater than 20 km/h.
These results should not be surprising. The eastbound vehicles had been travelling on a rural highway just before reaching the collision site where the posted maximum speed was 80 km/h. The collision site is just at the edge of the built up area of the City of London. The maximum speed of 60 km/h is posted at a distance of approximately 500 metres west of the cycling path. Then at approximately 250 metres a sign indicates that a maximum speed of 50 km/h is just ahead. Then within 20 metres west of the cycling path a sign indicates the 50 km/h maximum speed. However the erection of such signs does not guarantee that drivers will follow that requirement. Indeed numerous traffic studies conducted by Gorski Consulting, based on videotaped motions of vehicles, have shown that drivers do not travel at or below the speed posted on a maximum speed sign. The degree to which drivers speed through a site changes from one location to another. This is why detailed studies are needed to determine where speeding is a greater problem and if further actions to reduce speeds are necessary.
The following series of five images shows what can happen when vehicles are travelling too quickly in a area where visibility is poor. These images are frames taken from a video camera positioned on the east side of the cycling path on the north side of Gainsborough Road. The two cyclists in the images are travelling in the same (southbound) direction as the retired professor and his partner. As can be seen the female in the lead does not look toward the west while the male cyclist looks twice. Both times there is no eastbound vehicle in view. Then, as the male rider is looking forward in the last frame, a red eastbound vehicle appears just past the edge of the trees/bushes.
It is unknown how the actual collision occurred in which the professor was killed. However this is an example where the combined effect of high motor vehicle speed and poor visibility could be major causal factors.
The slope of the roadside at the path is another factor that could influence the cause of a cyclist collision. It was observed that a substantial downslope existed on the south side of Gainsborough at the roadside between the sidewalk and the edge of pavement. Slopes are an issue for recreational cyclists because they can increase or decrease a cycle’s speed. A cyclist whose physical abilities may be limited can have difficulty travelling up slopes and this can reduce their speed. Similarly down-slopes can push a cyclist into an unwanted position or make the cycle more difficult to stop.
The photos below show how slope measurements were taken on both sides of Gainsborough Road along the crossing at the cycling path. A carpenter’s level attached to a digital level was used to measure the slope in percent. In the three photos below the measurement being taken is at the south sidewalk and this gave a reading of 1.6 % downslope to the north (i.e. toward the road).
The next three photos show a similar measurement being taken of the south roadside located between the sidewalk and the edge of the roadway pavement. This reading showed a downslope of 11.6% to the north (i.e. toward the road).
As a comparison, roadway geometric design manuals for the province of Ontario recommend a maximum downslope of 12 percent for secondary roads. Thus the roadside slope is at the top of that maximum recommendation for roads driven by motor vehicles. However, cycles are not motor vehicles and they do not possess the “low rear” that can help a braking system to slow a vehicle’s motion along a downslope. The only way to reduce a cycle’s speed on a downslope is to use the cycle’s brakes. This is not ideal when approaching a busy roadway where it is important for a cyclist to stop before entering the roadway.
Northbound cyclists, such as those from the video frame shown below, were observed to stop at the top of the downgrade, on the sidewalk, to make traffic observations before descending the downslope and crossing the road. When starting their motions from a distance away from the road they increase their travel distance and time and make themselves more vulnerable to problems if they misjudge the speed of approaching vehicles.
Various riders exist on a cycling path, some of them may be elderly, some may have physical limitations and some may be small children who may be inexperienced in handling their cycle. Such limitations need to be kept in mind when considering the safety of an intersection of a cycling path with a busy road.
Similar slope measurements were taken on the north side of Gainsborough Road and these revealed less of a concern. The sidewalk on the north side was almost at the edge of the travelled surface of the road. The surface at the path gate was essentially level whereas the slope at the north edge of the sidewalk was 6.1%, as shown in the photos below.
The speed of pedestrians or cyclists crossing Gainsborough Road can be estimated from various research. However it is also helpful to obtain data from the actual site. The following table provides the time taken (in seconds) for cyclists and pedestrians to cross the 11.7 metres of pavement between the north and south curbs of Gainsborough.
A typical, 2-lane roadway might be about 8 metres wide. The extra road width at the site was due to the presence of a centre, mutual-turn lane as shown in the westward view below.
While the average crossing time of the 23 observations in the above table was about 6.7 seconds there were a number of observations of much longer times. The previously mentioned 2 adults pushing the two toddlers on mini-cars (Observation #13) involved a crossing time of almost 14 seconds. An eastbound motor vehicle travelling at 80 km/h will travel about 22.2 metres per second. Thus in 14 seconds such a vehicle would travel about 311 metres. However, we must also remember that a crossing pedestrian would need additional time to survey the roadway, and then move forward to the curb before that 14 second count-down would begin. Thus one can see that the parents (?) in Observation 13 could be making a judgment to cross the road when a speeding vehicle could be well over 300 metres away, or even 400 metres away. In such long distances a pedestrian would not be capable of detecting the travel speed of a vehicle. For example it is well-established from past research that, barring additional ques, an observer would have difficulty detecting the speed of an approaching vehicle, using the expanding size of the vehicle in the observer’s field of view, until that vehicle is about 175 metres away. If there are sufficient numbers of cyclists and pedestrians who take such long times to cross this roadway then a reasonable solution is to install a signalized, pedestrian crossing at such a site.
The safety of the Gainsborough site is compromised by the lack of any signage warning drivers on Gainsborough Road of the presence of the multi-use (cycling) path. However signage is not always the best improvement as most drivers ignore their warnings.
Judging by the data collected on the Sunday afternoon of September 13, 2020, Gainsborough Road is well populated by motor vehicle traffic. The City of London indicates an Average Annual Daily Traffic (AADT) of 7500 vehicles for Gainsborough. Given what has been observed in our study and considering the rapid development in that edge of the City, it is likely that the present traffic volume is higher than what has been indicated in the City data. Residents who spoke with us during the video documentations advised that we should come to the site on a weekday afternoon peak-hour as the traffic volume would be much higher. This is a correct advisement as we were aware of that fact. The opportunities for users of the path to cross Gainsborough Road would likely be more limited on a weekday peak hour than what has been shown in our data.
The City of London is likely no different than many cities in Canada and North America where a rapid transformation is taking place from a transportation system designed for dominance by motor vehicles to one that includes active transportation. During this transition infrastructure modifications have involved compromises where certain developments are less than ideal from a safety viewpoint. While many new developments have the luxury of “starting from scratch” and thus including well-developed facilities for active transportation, older developments remain challenging. How do you insert a new, multi-use path system into an old traffic system that did not envisage such a need? At newer developments, how do you maintain safety until sufficient in-fill occurs and funds become available to complete certain safety projects that would otherwise remain incomplete? These are complicated and difficult decisions. Gorski Consulting continues to examine and report on cycling paths and facilities to illuminate where improvements need to be made.