Compass Track Day 2023 - Displaying Connected Vehicle Capability

On Wednesday 3rd May, Compass held our annual Track Day at the RACQ Mobility centre, Brisbane. Over 90 transport professionals from across NSW, VIC, and QLD state government transport authorities and private sector spent the day participating in real-life road safety experiments and listening to presentations on how Compass Connected Vehicle data is being applied across Australia.

The event brought together innovative transport leaders from Suncorp, Transport Main Roads Queensland, UNSW, EY, IAG, Auto and General, Google Cloud, GHD, and Jacobs. Presenters included:

• Jeff Sharp, Associate Partner from EY
• Amanda Cochran, Strategic Initiatives from Transurban
• Eslam Hassan, Traffic Engineer and Road Safety Research Fellow from Bitzios Consulting
• Amir Soroush, Emerging Communications and IoT Lead from GHD - Digital
• Ahmad Tavassoli, Manager of Transport Analytics from Transurban

Key Learnings from the Road Safety Experiments

Attendees participated in 5 experiments that imitated sections of the road network under different speeding and braking conditions:

1. a Roundabout
2. Measuring g-forces around sharp bends
3. High-speed curves
4. Braking event - time to collision (stopping distance)
5. change in speeds and g-forces around bends

Participants received personalised results via their phone for their overall trip results and the results of each individual experiment

‍

Aggregate Results: G-forces

We've focused on G-forces for the aggregate results of Track Day.

G-forces at each site

What stands out is the 90+ percentile values. These can be looked at in relation to each other, but also compared to the median values of the same experiment.

Looking at the 90+ percentile values, the roundabout and braking experiment reported significantly higher values. Considering that G force is a measure of acceleration, this makes sense. Roundabouts tend to report higher g forces than regular turning events due to the combinations of vectors (acceleration forwards, and accelerations sideways). With braking, it's easier to decelerate from a set speed than it is to accelerate to it.

In terms of assessing risky driving behaviour, it can be interesting to look at 90+ percentile value G forces relative to 50 percentile G forces for the same site. When we look at these gaps, we see that both the roundabout (0.7) and the braking (0.64) also had the highest variance between 90th and 50th percentiles. Aggressive driving behaviour was more pronounced in these kinds of contexts.

G-force by driver (random sample of 20 drivers)

For the most part,  people stayed in their own "tier" for most experiments. That is, people that experienced high g forces on one event, were more likely to experience high g forces in another. This is because higher g forces are linked more to driver behaviour than they are experiment context.

The most notable exception to this was the braking experiment. If a driver was going to drive more aggressively to generate more G forces in a higher tier, it would be during the braking experiment. This is likely because people feel a lot more comfortable during sudden braking than they do going around a corner at high speed and because it was the last experiment on the track.

The braking experiment also had the largest standard deviation for drivers. This indicates that the braking experiment is the most likely experiment for drivers to drive more aggressively, regardless of if their driving behaviour was classified as "safe" or "aggressively" based on the other experiments.

G-force by Vehicle

This is a distribution chart of G forces pulled over 0.2 Gs.

• Vehicle make and model greatly impacts the experience of G force on a vehicle. A heavier vehicle will experience greater G forces since it has a higher mass.
• The other factor that will have impacted these charts is the braking and accelerating capabilities of a car. For example, vehicles with a stronger braking system will have resulted in higher g forces for many of the drivers that slammed on the brakes in the same way.

‍