More people are dying on Australian roads. This program could make drivers safer

Deaths on Australian roads have increased every year since 2020. This is despite the Australian government’s commitment to Vision Zero – having zero deaths or serious injuries on our roads by 2050.

Unfortunately, 1,317 road deaths were recorded in 2025, a 1.9% increase from 2024. Land transport accidents also remain a leading cause of death for children and young adults, and the third leading cause of injury hospitalisations.

To bring these stats down, we need to look at the entire system of road use – including the parts that don’t get benchmarked but perhaps should.

The ‘safe system’ principle

Part of Vision Zero is a stronger commitment to the Safe System approach. This means all parts of the road transport system work together to keep us safe. These include road users, vehicles, road quality and design, planning and speed.

But what exactly are “good roads”, “good vehicles” or “good drivers”? For some parts of the system, there are clear answers.

Vehicle quality and safety is benchmarked via the Australian New Car Assessment Program, ANCAP. Road safety is benchmarked via the Australian Road Assessment Program, AusRAP.

However, there’s no clear mechanism to benchmark human performance as road users. Sure, if we drive or ride a motorcycle, we must demonstrate certain competencies to be granted a licence. But afterwards, we don’t receive objective feedback on our performance as road users.

Our own judgements aren’t good enough. Many of us suffer to some degree from illusory superiority, and we have the general tendency to assess our own competencies on a task as “above average”. In one US study, 673 out of 909 participants (74%) thought they were better-than-average drivers.

Logically, most of us can’t be better than average at driving. This is where an assessment program for road users could come into play.

Towards a road user assessment program

Recent research from the Australasian College of Road Safety examined the novel proposal of a road user assessment program.

They suggested benchmarking – having a standard they can be measured against – should be available for road users as part of a safe system approach, just as it is for vehicles and roads.

Through interviews with road-safety experts (including two of us) and a forum of road safety researchers, professionals and advocates, the authors of the report identified five areas for feedback to road users:

• the road user’s skills and knowledge
• pre-trip preparation
• risk management (such as road positioning, speed, distraction, hazard perception and compliance)
• self-maintenance and monitoring, and
• what happens after an incident (that is, how we learn from crashes or near misses).

Do we need a separate program for this?

As drivers, we do already receive feedback from multiple sources. And several active safety systems exist in modern cars. Some of them, such as lane-keeping assistance, actively manipulate what the car does while we drive.

Such technologies are known as advanced driver assistance systems (ADAS). These can provide us with warnings on the road, or can automate some aspects of driving. Evidence shows ADAS can reduce the frequency of crashes. Moreover, autonomous emergency braking is now compulsory in new cars sold in Australia.

But many of us drive vehicles without these features. This strengthens the argument for a uniform and easy-to-use feedback mechanism available to all road users, to improve road safety.

However, such a benchmark would be complex to develop and put into place. Who would implement this system? Should modern technologies, such as artificial intelligence (AI), play a role? If the program was voluntary, how would we encourage people to take part?

For now, these big questions might seem insurmountable, but we have some recommendations.

So what might the program look like?

Guided by the five recommended areas for feedback to road users, we envision a benchmark program for typical car drivers could use advances in AI and telematics.

AI tools which can monitor driver behaviours already exist. Telematics uses information from sensors, GPS and other diagnostics, and can provide information on driving performance, such as speed and braking.

Indeed, the use of telematics is rapidly expanding in Australia for freight vehicles. While more data is needed to evaluate the impact of telematics on driving performance, the potential is there, especially in combination with other sources of feedback.

Using this data could allow for better trip preparation, also incorporating users’ driving history (such as driving skills, habits, knowledge and preferences), as well as traffic information and weather conditions. Telematics is achieved via a device placed in the driver’s vehicle. Perhaps a similar approach could be used here.

Noting the complexity in giving feedback to drivers, we also propose a shift from calling it a road user assessment to a road user “assistant” program. This would reflect that any such system is designed to support the road user. If feasible, it could be potentially adapted to other road users, such as cyclists and pedestrians.

The development of past benchmarking systems for roads and vehicles has increased safety on Australian roads. However, these only go so far.

The missing factor that will benefit from benchmarking is us as road users. Perhaps then we can get closer to the ambition of Vision Zero.

Acknowledgements: The authors would like to acknowledge collaboration with Roderick Katz of the Australasian College of Road Safety.