Episode transcript · Auto-generated · The Low Divide
← Listen to this episodeAI Deep Research Report
On 18 June 2025, Perth recorded 188% peak congestion — the worst acute 15-minute traffic collapse of any Australian city. At 5:00 PM, the average Perth driver could travel just 4.0 kilometres in 15 minutes. That is less than the distance from the Perth CBD to East Perth.
Perth's traffic performance across every major index:
| Metric | Perth Value | Australian Rank |
|---|---|---|
| Annual rush-hour time lost (TomTom) | 62 hours/year | 8th — but concentrated into severe peaks |
| Annual commuter delay (INRIX) | 42 hours/year | 4th worst in Australia |
| Peak-hour network congestion | 72.2% | Evening rush is twice as bad |
| Worst-day acute collapse | 4.0 km in 15 min | #1 worst in Australia |
| Avoidable congestion cost | $2.1 billion/year | — |
| Road Network Health Score (2026) | 55.4 / 100 | One freeway breakdown paralyses the city |
| Projected score without action (2046) | 31.8 / 100 | Matching Los Angeles |
Without intervention, Perth will hit 55.2% average congestion and 110 hours lost per driver per year by 2046 — matching the congestion conditions that make Los Angeles a global by-word for gridlock.
The political conversation jumps immediately to tunnels. The engineering literature does not. Every fix on this page can be implemented with paint, software, and traffic signal hardware. None require acquiring a single property.
Eight interventions. Zero new tunnels. 25.6 minutes saved per commuter per day (conservatively discounted by 18% so overlapping fixes aren't double-counted).
| The Fix | Mechanism | Daily Saving |
|---|---|---|
| 1. Hold green till the last car clears | Install dual radar loops 65–130 m upstream of 950+ SCATS signals; hold green until last car in platoon clears (MOVA control) | 3.8 min/day |
| 2. Show drivers the “green wave” speed | Broadcast optimal cruise speeds to connected vehicles on 26 major arterials; eliminates stop-go accordion waves (C-ITS GLOSA) | 3.6 min/day |
| 3. Meter CBD entry before gridlock hits | Throttle private car inflow to CBD, West Perth, Subiaco before gridlock; 100% bus signal preemption at cordon (MFD perimeter gating) | 3.2 min/day |
| 4. Switch on Smart Freeways for PM peak | Extend Active Traffic Management to Mitchell NB and Kwinana SB; add AI radar incident detection — 12 min faster response (Bidirectional Smart Freeway) | 5.0 min/day |
| 5. Paint central right-turn lanes | Convert 4-lane commercial strips to 5-lane with central painted turning lane — Scarborough Beach Rd, Albany Hwy, Guildford Rd, Canning Hwy, Stirling Hwy (Flush median / TWLTL) | 4.8 min/day |
| 6. Extend clearways & tow peak blockers | Expand clearway hours to 06:30–09:30 and 15:30–18:30; raise WA fine from $150 to $350; pre-position tow units (ANPR enforcement) | 2.8 min/day |
| 7. Flip bridge lanes for peak direction | 4:2 reversible lanes on Causeway and Mounts Bay Rd at peak; convert Albany Hwy/Liege St to 2-phase MUT geometry (Dynamic tidal flow / Median U-Turn) | 5.0 min/day |
| 8. Cut off-peak fares & mandate flex work | Mandate QEII-style workplace travel plans at major hubs; 30% peak/off-peak Transperth fare spread (Institutional TDM) | 3.0 min/day |
| Network-adjusted total | Zero new tunnels | 25.6 min/day · 98 hours/year |
98 hours returned per commuter per year. That is the entire INRIX and TomTom delay penalty — eliminated — without a metre of new road.
Perth's 950+ traffic signals run on SCATS — a system that measures traffic at the stop line. It cannot see a platoon of 12 cars approaching 100 metres back. So it regularly cuts the green phase 3 seconds before the last car arrives, forcing those cars to sit through a full 120-second red cycle.
The fix: Install two radar or inductive loops upstream — one at 130 metres, one at 65 metres. The controller holds green until the last car in the approaching platoon clears, then ends the phase immediately. No wasted green time. No premature red.
Saving: 1.9 minutes per one-way trip across a typical 14-intersection Perth commute.
Green Light Optimal Speed Advisory broadcasts a cruise speed to approaching vehicles: "travel at 52 km/h and you'll hit the next three lights green." At 25% connected-vehicle penetration — already achievable via navigation app probes — this system eliminates the accordion effect where one heavy braker triggers a chain of unnecessary stops across an entire arterial.
Saving: 1.8 minutes per one-way trip on the 26 primary arterial corridors.
The Macroscopic Fundamental Diagram — the equation that governs how many vehicles a road network can productively handle — shows that once a network exceeds critical density, adding more vehicles reduces total throughput. Perth's CBD regularly crosses this threshold at peak. The result: more cars entering produces slower movement for everyone already inside.
The fix: Install dynamic cordon signals at Charles Street, Thomas Street, the Causeway, and Beaufort Street. When CBD density approaches critical, these signals hold a green wave for buses and slow private car inflow by 10–15%. Queues are stored on wide multi-lane boundary approaches, not inside the CBD core.
Saving: 16–30% increase in CBD network discharge; 11–28% reduction in bus travel times.
Main Roads WA's Kwinana Smart Freeway (northbound, launched August 2020) saved 6–10 minutes per AM peak trip and absorbed a 36% post-COVID volume surge without catastrophic gridlock. The Mitchell Freeway Southbound (December 2024) saves up to 7 minutes in the AM peak.
The gap: Neither freeway has equivalent outbound PM protection. At 5:00 PM, when 72.2% average congestion strikes and 188% on worst days, there is no coordinated ramp metering, no variable speed harmonisation, and no Hard Shoulder Running northbound on Mitchell or southbound on Kwinana.
The fix: Mirror the existing Smart Freeway architecture onto the outbound evening directions. Add AI/radar Automatic Incident Detection (AID) — 50–60% of freeway delay is non-recurring, caused by breakdowns. AID detects incidents 12 minutes earlier, enabling upstream diversion before the shockwave propagates.
Saving: 5.0 minutes per day for the 38.2% of Perth trips that use the freeway network.
This is the simplest intervention on the list. It is also the one with the largest single-corridor impact.
When a driver turning right into a Scarborough Beach Road car yard has to wait 35 seconds for a gap in oncoming traffic, every vehicle behind them is blocked. The kerbside lane — where buses stop and left-turn movements occur — becomes the only option. The result: both lanes are degraded simultaneously. Effective directional capacity drops by 35–55%.
The fix: Paint a 3.2-metre central lane between the two directions of traffic. Turning vehicles pull into this lane to wait. Through traffic continues unobstructed in both directions. No building is demolished. No kerb is moved. The carriageway already has the width — it just needs to be re-striped.
Primary candidate: Scarborough Beach Road, Main Street to King Edward Road (2.67 km through Osborne Park). The lane re-striping reclaims space from underutilised concrete verges.
Effect: Directional saturation flow rises from approximately 1,950 veh/h to 3,400 veh/h (+74%). Peak space-mean speed through Osborne Park rises from 21 km/h to 39 km/h. Travel time through the segment: -3.5 minutes.
Other prime candidates: Albany Highway (Bentley/Cannington), Guildford Road (Maylands/Bayswater), Canning Highway (Applecross/Como), Stirling Highway (Nedlands to Fremantle), Fitzgerald Street/Alexander Drive (North Perth).
A single parked delivery van reduces a two-lane arterial to one effective lane for 150 metres upstream. Perth's clearways currently run 07:30–09:00 AM and 16:15–18:00 PM. Peak spreading means commuters are on the road from 06:30 AM and until 18:30 PM. The kerb lane is lost at exactly the shoulders when congestion is most sensitive to added friction.
Western Australia's clearway fine is $150 — the second-lowest in Australia. NSW charges $362. The deterrent effect is negligible.
The fix: Extend clearway hours by 90 minutes each peak. Raise the fine to $350. Deploy ANPR cameras with pre-positioned tow units on inner-ring arterials. No policy change required for Main Roads WA to act on the operating hours — this is an administrative directive.
The Swan River forces massive east-west flows across a small number of bridge crossings. In the AM peak, four lanes of the Causeway would ideally flow inbound. At PM peak, four lanes outbound. Currently the split is fixed at 3:3.
The fix: Overhead Lane Use Management Signs (LUMS) and movable zipper barriers flip the Causeway and Mounts Bay Road to a 4:2 inbound/outbound split at 06:30–09:30, reversing to 2:4 at 15:30–18:30. No new bridge is poured. 33–50% additional directional capacity is created with hardware.
At Albany Highway/Liege Street — the single worst bottleneck in the City of Canning — a Median U-Turn configuration replaces the current four-phase signal with a two-phase signal. Four conflicting phases become two. Intersection throughput rises 20–30%.
The QEII Medical Centre in Nedlands implemented a structured workplace travel plan. Over four years, private car trips to the precinct fell by 30% — removing 3,840 employee car trips and 3,000 patient and visitor car trips from the network each day.
Extending mandatory travel plans to 1,940 major employment and activity sites across Perth removes up to 194,000 peak-hour vehicle trips.
A 30% peak-versus-off-peak Transperth fare discount shifts discretionary trips out of the 08:00 AM and 17:00 PM peaks, where the traffic engineering equivalent of a fourth-power BPR function means each additional vehicle has a disproportionately large impact on everyone else's travel time.
Kwinana Smart Freeway (Perth, August 2020): 36% post-COVID volume surge on the Narrows Bridge corridor. Under static control, this would have caused catastrophic gridlock. With coordinated ramp metering across five on-ramps and a converted emergency shoulder lane, the freeway saved up to 10 minutes per AM trip and absorbed the full demand surge.
Mitchell Freeway Smart Freeway Southbound (Perth, December 2024): $209.6 million, 1,400+ ITS devices, 16 coordinated ramp signals. Saves up to 7 minutes in the AM peak. Northbound outbound PM: unprotected.
Scarborough Beach Road planning documents (City of Stirling / Main Roads WA): State planning documentation explicitly identifies the Main Street to King Edward Road segment as carrying severe transport, access, and parking friction with a high recorded crash rate. The flush median solution requires no property acquisition and no new planning approval.
Los Angeles: Systematic conversion of four-lane commercial boulevards to five-lane TWLTL cross-sections across the LA Basin — the same intervention now proposed for Scarborough Beach Road, Osborne Park.