Inside a Transurban trial car looking out the windscreen; Inside a Transurban trial car looking out the windscreen;

Connected and Automated Vehicle trials, New South Wales

Informing the planning of our roads today for a connected and automated vehicle tomorrow

Self-driving cars currently rely heavily on ‘reading the road’ to get from point A to B – with their complex array of cameras, sensors, GPS and other technologies detecting and responding to other vehicles, lights, signs and road markings. So, is our existing road infrastructure giving clear and safe directions for new ‘driver assistance’ features? And what do we need to implement now and in the future for a fleet of increasing numbers of connected and automated vehicles (CAVs)?

Arup was engaged by Transurban, in collaboration with the NSW Government, to establish and run a trial program of partially automated vehicles, observing their interactions with Sydney’s orbital network road infrastructure. This expansive network covers some of Sydney’s busiest roads – the Lane Cove Tunnel, the Hills M2 Motorway, Westlink M7, the M5, Eastern Distributor and the Sydney Harbour Bridge and Tunnel.

Project Summary


7,110km travelled

10vehicles tested

4,125driver observations

The observations collated as part of this trial, aim to inform not only our future road infrastructure and regulations, but ongoing upgrades and maintenance and improvements in the technology of the vehicles driving these roads.

Working closely with Transurban and vehicle manufacturers, Arup provided support in project management, planning and operational readiness advice for the trial. In addition, Arup people were observers in selected trial vehicles, in an extensive range of scenarios, environmental conditions and locales. The team was also involved in processing footage and data collected from the trials. 

This project follows earlier successful trials conducted with Transurban and the Victorian Government in Melbourne. 

Sydney’s unique trial findings

Sydney presented some new challenges for partially automated vehicles as we prepare our roads for a CAV-driven future. Ten latest model vehicles with partial automation features were trialled – focusing on their ability given road conditions, to steer and keep in their lane, manage their speed relative to other vehicles and recognise the speed limit.

The inconsistency of our road surfaces – from concrete to asphalt – with bitumen sealed cracks and signs painted on the pavements impacted the ability of vehicles to identify lanes and in some cases disengaged lane keeping features. Sydney’s tunnels proved problematic too, with line markings flush against the tunnel wall or faded and dirty lines impeding lane visibility.


Identifying speed limits amongst the visual stimuli of the city presents issues for the best of us. Trials highlighted that this is equally confusing for vehicles, with speed signs on buses, parallel roads and exit ramps and conditional speed signage being identified by vehicles as applying to the motorway. Vehicles also contended with the electronic signage, with flashing signs read more reliably than the continuous signs.

Other factors which impacted automated features were sharp curves, bends and dips and permanent bollards that interfered with lane keeping, and environmental factors such as rain, glare or shadows.

Read the detailed findings and recommendations in the Transurban report.

What constitutes a CAV-friendly road?

The trial recommended items which would make roads easier to navigate for future CAVs:

  • Well-marked roads, and an alternate sealing method for repairing cracks, to enable lane keep to assist to function more effortlessly.

  • Clear and consistent placement, readability, visibility and accuracy of speed signs and supporting map data

  • Wide shoulders in tunnels, with increased frequency of cleaning and re-marking lines inside the tunnel

  • A standardised approach to electronic signage

  • Demarcation between objects, such as bollards, and changes in road surface and markings.


The study also recommends drivers are informed of the capabilities and limitations of driver assistance features to ensure public safety.

As these automated features rapidly improve and become more commonplace, trials like this one will continue to inform a collaborative approach to road infrastructure delivery, improvements and maintenance – ensuring the clear direction and safer operation of CAVs in our communities.