Analysing near-misses for back-of-queue protection

Goal

Identify spillback on highway off-ramps with the Near Misses tool. We’ll learn this by completing the Transurban use case.

Context

Transurban, one of the world's largest toll operators, found near-misses were occuring on a portion of the M7 due to spillback from the off-ramp. After changing the traffic signal timing at the top of the off-ramp, they wanted to see if this helped reduce spillback and, thus, near-misses.

Do

Let’s re-create this use case, and understand how to download near-misses for analysis.

1. Copy and paste the coordinates “-33.740576, 150.847464” into the Search Bar in the top left of the platform Home Page.
2. Choose the Near Misses tool from the Tool Selection panel.
3. Choose the following dates in the Date Selection panel:
   • 1 July-31 December 2020
   • 1 July-31 December 2021
4. Select a polygon around the Rooty Hill Rd off-ramp that looks a little like this:
   • From the exit of the Westlink M7 down to Lamb Street, or

• Paste this into the WKT selection tool and disable road-matching:
  • POLYGON ((150.84767826430271 -33.73490856101763, 150.84880808234004 -33.7345530455431, 150.84667058875664 -33.74502743728327, 150.846243090041 -33.744837005212915, 150.84778513898232 -33.737574880117634, 150.84767826430271 -33.73490856101763))

Further Links

• Spillback and back-of-queue protection use case
• Near-Miss - How to use the near-miss tool
• How do I interpret G-force Movements, Speed, and Braking and Accelerating graphs?
• Understanding Results from Compass

00:00

Let's use the Road Intelligence platform to understand spillback on highway off ramps. We'll pretend we're part of Transurban in New South Wales, Australia. And we want to know if changing traffic signal timing on a part of the M7 will reduce spillback and near misses.

00:16

To start this study, first we'll go to the tools and make sure we have Near Misses selected. Next we'll copy and paste these coordinates into the Search Bar. You can find them in the lesson notes. But just in case you can't here it is on the screen.

00:32

Next we'll put in the date ranges for before and after the signal change. We'll go to the Date Selection tool and click on this existing date. In the calendar, we'll go to 2020 and select the 1st of July to the 31st of December that year. That's before the signal change.Next we'll select a similar period after the change, we'll go to 2021 and select the 1st of July to the 31st of December.

01:11

As for the zones, we'll draw this shape by clicking around the exit of the offramp. If you're familiar with the WKT tool in Advanced Mode, you can also get the polygon from the lesson notes.

01:34

To see if the traffic light changes had any effect, we need to find out what happened before and after the changes, and what kind of near misses were there.

01:44

So what happened before and after the changes? To answer this, let's examine the near misses on the map. We'll make sure the near misses layer is already turned on. You can do that by clicking on the bottom right corner. If the dots aren't already showing.

01:57

In the Date Selection panel, we'll hide the 2021 date range and view just 2020 data on the map. You should see about six near miss points going north down in the polygon.

02:09

Now toggle the date ranges to view just 2021 data on the map. You'll notice fewer new miss points are in the polygon. In this version, we can see three. We can see a reduction in near misses after the signal was changed. The change in phasing gave drivers more time to exit the off ramp and reduce spill back.

02:28

Now that we figured out the effects of the changes, let's see what kind of near misses were occurring around the off ramp. When you click on a near miss dot, you can get a brief overview in the pop up on the top right corner.

02:42

But to save clicking on each point for more details, we can download the raw near miss data and view everything on a CSV file. All you have to do is click Download Results.

02:55

Because we have two date ranges. We should have two separate files, one for 2020 and one for 2021. We'll open the files up, and you can see in 2020 there were six breaking near misses. And in 2021 there were three breaking near misses. G-forces in 2020 were higher, and they also had higher speeds.

03:18

And that's it. Through this analysis, we identified a reduction in near misses, causes and factors of those near misses, and identified a change in the speeds and g-forces of near misses. Just like Transurban, we can see how traffic light changes reduce spillback by doing this quick before-and-after analysis.

03:37

Now you can apply these skills to your own studies or try following along with this video. Just know that the platform is constantly updating its data set so results may vary, but the steps are generally the same. We'll see you in the next practice lesson.