The City of Moreton Bay is preparing to trial an AI-powered traffic signal system at the Moreton Parade and Paper Avenue intersection in Petrie later in 2026. ABC News reports the $170,000 trial will use camera-derived traffic visibility and demand-based signal decisions rather than a fixed phase cycle, accounting for vehicles, buses, cyclists and pedestrians.
The City of Moreton Bay is preparing to trial an AI-powered traffic signal system at the intersection of Moreton Parade and Paper Avenue in Petrie, north of Brisbane, later in 2026. ABC News reports that the $170,000 project is Australia's first trial of the technology at an intersection, replacing fixed signal-phase combinations with decisions based on observed demand.
The system is intended to assess movements by cars, buses, pedestrians and cyclists and adjust traffic control accordingly. Queensland University of Technology professor Ashish Bhaskar told ABC News that cameras can provide greater visibility of intersection activity, while AI can use that input to improve the traffic-control algorithm. He said this could support more proactive decisions across a corridor or network rather than merely reacting at one junction.
From loops and phases to movement-level control
Conventional Australian intersections commonly use inductive loops embedded beneath the road to detect vehicles, then cycle through predefined phases. According to ABC News, the proposed system does not operate on a fixed cycle and can be tuned for local priorities, such as pedestrian movement in a CBD or vehicle throughput in high-traffic areas.
Drive reports that the Moreton Bay trial uses cameras and LiDAR sensors, with the traffic-control technology supplied by Austrian-headquartered road-infrastructure company SWARCO. LiDAR can generate three-dimensional spatial measurements, including in low-light conditions, complementing video-based detection.
A City of Moreton Bay spokesperson told Drive that the approach controls individual traffic movements independently rather than relying on the traditional phase-based model. The spokesperson also said comparable systems are operating in several European jurisdictions, although the approach has not previously been implemented in Australia.
What the trial needs to establish
The central question is not simply whether the system can shorten waits. The Conversation notes that an adaptive controller has to make distributional choices: a busy vehicle approach, a bus, cyclists and pedestrians may all compete for limited green time. System objectives and constraints therefore matter as much as prediction quality.
For data and ML practitioners, comparable intelligent-transport deployments generally require separate evaluation of perception, control and safety outcomes. Detection accuracy across daylight, darkness, weather and different road users is distinct from whether a controller reduces total delay. Aggregate travel-time gains can also obscure longer waits for pedestrians, cyclists, buses, or less-frequent side-street movements. The sources do not provide model architecture, training data, retention policies, latency targets, safety assurance methods or trial success thresholds. Those details would be material to assessing whether the system can operate reliably in a live public-road environment and whether any measured benefit is reproducible beyond a single suburban intersection.
Drive also reports that the council acknowledged ongoing costs associated with the AI-powered signal, which will be assessed during the trial. That makes the project a practical test of operating economics as well as traffic performance: sensing hardware, compute, maintenance and integration with existing signal infrastructure can determine whether an adaptive controller is viable at broader scale.
If the Petrie deployment produces independently measurable improvements while maintaining safe and equitable access, it could provide evidence for wider Australian testing. For now, the reported project remains a single-intersection trial rather than evidence that AI control reduces congestion across a road network.
Key Points #
- 1Moreton Bay's Petrie deployment tests demand-based, movement-level traffic control instead of conventional fixed signal-phase combinations.
- 2Cameras and LiDAR expand intersection sensing, but trial evidence must separately establish perception accuracy, control performance and safety.
- 3Adaptive traffic systems can reduce delay while creating allocation tradeoffs among drivers, transit riders, pedestrians and cyclists.
Scoring Rationale #
This is a real-world AI control trial in public transport infrastructure, with relevance to computer vision, sensor fusion and online decision systems. Its practitioner impact is currently limited by the single-intersection scope and the absence of disclosed technical specifications or results.
Sources #
Primary source and supporting public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.