South Australia will trial five artificial intelligence systems designed to improve traffic flow, incident response and road safety across the state’s transport network.
The State Government has committed almost $500,000 to the trials through the Office for Artificial Intelligence’s $35 million AI Proof of Value program.
The technology will use existing roadside cameras and traffic infrastructure to detect crashes, monitor queues and adjust signal timing in response to changing conditions.
What technology will be tested?
One trial will use artificial intelligence and existing roadside CCTV cameras to automatically detect traffic crashes.
The system would alert Traffic Management Centre operators, potentially reducing the time required to identify an incident and coordinate a response.
Another trial will test adaptive pedestrian signals.
Pedestrian green time is typically fixed at five seconds. The proposed system would detect how many people are waiting and extend the crossing time when required.
AI will also be used to identify traffic queues and reallocate green time at signalised intersections.
How could AI reduce congestion?
Two trials will focus specifically on queue management.
One system will detect congestion approaching signalised intersections and provide additional green time where needed.
Another will identify vehicles overflowing from dedicated right-turn lanes into through-traffic lanes.
Right-turn overflow can restrict traffic movement and increase congestion across an intersection. Earlier detection could allow traffic controllers or automated systems to adjust signal phases before queues worsen.
For road freight operators, more responsive traffic signals could improve travel-time reliability through busy urban corridors and reduce unnecessary idling.
Could the trials reduce transport costs?
The government said smoother traffic flow could reduce travel times and lower fuel consumption.
Data from South Australia’s Traffic Management Centre estimates that forcing 5,000 motorists to wait an additional 20 minutes during the morning peak costs the state economy more than $33,000.
The calculation reflects the broader productivity effect of congestion, including lost time and higher vehicle operating costs.
Heavy vehicles can face even greater fuel costs when repeatedly stopping, accelerating and idling in congested traffic.
While the trials are focused on the wider road network, successful systems could also benefit freight and service vehicles operating through Adelaide.
How will the technology improve road safety?
The fifth trial will use artificial intelligence to detect cyclists and determine their direction of travel.
That information could help traffic signals respond earlier to cyclists approaching an intersection and reduce conflicts with other road users.
The crash-detection system could also improve safety by reducing the time between an incident occurring and Traffic Management Centre operators becoming aware of it.
Faster detection can support earlier warnings, traffic controls and emergency responses.
Is AI already managing Adelaide traffic?
South Australia already uses smart-camera technology on some of Adelaide’s busiest roads.
Those systems monitor traffic flow and automatically adjust signals according to demand.
The government said technology installed near the Heaslip Road exit from the Northern Expressway had contributed to a significant reduction in collisions.
The new trials will test whether similar data-driven approaches can be expanded to other road safety and congestion problems.
If the trials prove successful, the systems could support a more responsive transport network that adjusts to incidents, pedestrians, cyclists and changing traffic volumes in real time.
