Inductive detector loops sit beneath the road surface and form the sensing layer for a significant portion of Australia's signalised intersections. When they work correctly, signal controllers receive accurate vehicle presence data and timing plans respond accordingly. When they don't, the consequences range from wasted green time to phantom calls and, in worst cases, undetected queues that back up into adjacent intersections. The faults aren't always dramatic. Some develop slowly over months, others appear on day one of commissioning.
How a detector loop actually works
A detector loop is a coil of wire cut into the pavement, typically in a square or rectangular saw-cut slot. The loop forms part of a tuned LC circuit in the detector card inside the signal cabinet. When a vehicle passes over or stops on the loop, its steel mass changes the loop's inductance. The detector card senses that shift and outputs a presence or pulse signal to the controller. The inductance change is small, typically between 0.5% and 3%, which is why the installation quality of every centimetre of wire matters.
Loop geometry affects sensitivity. A single 2-metre square loop detects most passenger vehicles reliably, but motorcycles and bicycles present far less ferrous mass. Some practitioners cut diagonal loops or use multiple loops per lane specifically to improve two-wheel detection. The choice of geometry should be specified in the design, not improvised on site.
The most common installation faults
Saw-cut depth is the first point of failure. Australian standards specify a minimum cut depth to ensure the wire is below surface stresses caused by traffic loading. Cuts that are too shallow expose the wire to pavement flex and rapid abrasion. Cuts that are too deep weaken the pavement structure. Both errors are hard to reverse once the loop is installed and the road is reopened.
Sealant application causes more long-term failures than almost any other factor. The sealant fills the cut, protects the wire, and provides the pavement with structural continuity. Sealant applied in wet or contaminated cuts doesn't bond properly. Sealant that's underpoured leaves voids where water infiltrates, freezes in cold weather (particularly relevant in alpine and tablelands regions), and forces the wire against the cut walls. Sealant applied over wire that's been laid with excessive tension creates stress concentrations at each corner of the loop.
Lead-in cable joints are another persistent fault point. The lead-in connects the in-pavement loop to the detector card in the cabinet, typically running through conduit in the kerb and footpath. Any joint in that run that isn't properly waterproofed becomes a path for moisture to enter the circuit. A single corroded joint increases loop resistance and reduces detector sensitivity without triggering an obvious alarm.
Corner cutting, literally and figuratively, matters. The wire must be formed around a corner insert (usually a triangular rubber block) at each corner of the loop. Without a corner insert, the wire is bent at a tight radius, and repeated pavement flex fatigues the insulation at that point. The failure may take 12 months to appear, but it's predictable from day one.
Testing protocols that catch faults before handover
Every detector loop should be tested at three stages: after wire laying (before sealant), after sealant cure, and after connection to the detector card. The minimum measurements are loop inductance, lead-in resistance, and insulation resistance to earth. Typical loop inductance for a 2-metre square, 3-turn loop runs between 100 and 300 microhenries. Insulation resistance should read above 100 megohms at 500 volts DC. Any reading below that threshold before the road is reopened to traffic is a fault that must be repaired.
Detector card sensitivity settings must be matched to the measured loop inductance. A card configured for high sensitivity on a high-inductance loop will generate false calls. A card set for low sensitivity on a short-inductance loop will miss motorcycles entirely. This matching step is frequently skipped or defaulted to a standard setting, which is not the same thing as a correct setting.
Bob Panich Traffic Signals designs detector loop specifications into signal projects from the drawing stage, not as a post-installation adjustment. Getting the loop geometry, inductance target, and sensitivity setting documented before construction reduces the scope for on-site improvisation that creates these faults. That discipline connects directly to the commissioning process: for a detailed look at what happens when the controller is energised and tested against design intent, the article on commissioning traffic signal systems covers the full sequence from hardware verification to sign-off.
Fault diagnosis in the field
A failed or degraded loop presents in several ways. An open-circuit fault causes a constant call (the detector card sees no inductance change and defaults to presence). A short-circuit fault causes no calls at all. A high-resistance fault causes intermittent calls that correlate with temperature, often appearing in cold mornings and clearing by midday as the pavement warms. That thermal signature is a reliable indicator of a corroded joint or damaged insulation rather than a faulty detector card.
Field diagnosis starts with disconnecting the loop at the detector card and measuring insulation resistance and loop inductance with a portable LCR meter. If both readings are within spec, the fault is in the detector card or its configuration, not the loop. If insulation resistance is low, the fault is in the pavement or lead-in. Tracing that fault requires a time-domain reflectometer, which locates the fault distance along the cable by measuring the round-trip time of a reflected signal. This is standard kit for a competent maintenance team.
Selective replacement of a single turn of a multi-turn loop is rarely a clean repair. The geometry of the replacement wire won't match the original exactly, and the new section introduces a joint in the pavement that will behave differently from the original. Full loop replacement is the correct approach in most cases, even though it means re-cutting the pavement.
When detector loops aren't the right technology
Detector loops have practical limits. They can't be installed in continuously reinforced concrete pavement without cutting through the reinforcement. Rapid pavement resurfacing programmes can bury loops at incorrect depths or cover lead-in junction boxes. In high-resurfacing-frequency corridors, video or radar vehicle detection may offer a better total cost of ownership, even if the per-unit cost is higher.
The choice of detection technology should sit in the design phase, not the procurement phase. Bob Panich Traffic Signals evaluates detection options against pavement type, maintenance regimes, and the controller's input requirements before specifying hardware. That upstream decision directly affects how well the signal system responds to live conditions. For a broader look at how detection data feeds into signal timing and adaptive control, the article on signal phase and timing plans explains how that data is used once it reaches the controller.
Poorly installed loops don't just reduce detector performance. They introduce noise into the data that adaptive systems depend on, and they compromise the fault records that maintenance teams use to prioritise repairs. Getting the installation right is a precondition for every layer of signal intelligence that sits above it.