What is an Axle Counter? The “Digital Eye” Replacing Track Circuits

An Axle Counter is a train detection mechanism used in railway signaling to determine if a section of track is clear or occupied. Instead of using electrical currents through the rails like a Track Circuit, it uses magnetic sensors to count the number of wheel axles entering and leaving a specific zone. If the count matches, the track is declared safe. It is the preferred technology for flooded areas and steel bridges.

December 8, 2025 10:52 am | Last Update: March 20, 2026 7:22 pm

⚡ In Brief

An axle counter detects train presence by counting the number of axles entering and leaving a defined track section using magnetic sensors mounted on the rail web — when the exit count equals the entry count, the section is declared clear.
Unlike track circuits, axle counters do not use the rail itself as an electrical conductor, making them immune to the weather conditions (flooding, leaf contamination, rail rust) and infrastructure constraints (steel bridges, stray traction currents) that degrade track circuit performance.
The critical safety limitation of axle counters is their inability to detect broken rails — a track circuit detects a fractured rail because the break opens the electrical circuit; an axle counter cannot see the rail condition between its sensor heads and will report a section as clear even if the rail has fractured, provided no train is present.
Axle counters require a reset procedure when axle counts are lost or mismatched — typically after power failures, maintenance operations, or count discrepancies — which requires manual confirmation that the section is clear and can cause operational delays on busy lines.
The leading manufacturers of railway axle counter systems are Frauscher (Austria), Siemens (Germany), Thales (France), and Voestalpine (Austria), with Frauscher holding the largest global market share for new installations.

On a busy August afternoon, a maintenance team working on a section of the Vienna S-Bahn lifts a small trolley off the track mid-section and carries it to a road vehicle — a routine operation, done without notifying the signal box. The axle counter at the section entrance recorded the trolley rolling in. The axle counter at the section exit recorded nothing going out. The section remains permanently “occupied” until a signaller manually verifies the section is clear, performs a reset procedure, and restores the signal.

This scenario — the “ghost train” left by a miscounted or missing axle — is the most common operational nuisance associated with axle counters, and it illustrates both the strength and the limitation of the technology. The strength: the axle counter correctly detected an anomaly and held the section at danger rather than falsely clearing it. The limitation: the system cannot distinguish between a train that is still in the section and a trolley that was removed from the section without being counted out. The signal box must intervene manually to resolve what the system alone cannot.

What Is an Axle Counter?

An axle counter is a train detection system that determines whether a defined section of track is occupied by counting the metallic axles that enter and exit the section. A counting head (sensor assembly) is installed at each end of the section. As a train enters, the entry sensor counts the axles passing. As the train exits, the exit sensor counts the departing axles. When exit count equals entry count — net axles in section = zero — the section is declared clear.

The fundamental difference from a track circuit is what the system measures. A track circuit monitors the electrical state of the entire section continuously — it knows whether any part of the section is occupied at every moment. An axle counter monitors two boundary points — it knows how many axles have entered and how many have left, and infers section occupancy from the difference. It does not have any information about what is happening between the two sensor heads.

How an Axle Counter Works: The Detection Physics

Step	What Happens	Engineering Detail
1. Magnetic field generation	The sensor head generates a permanent magnetic field across the rail head	Typically using permanent magnets or low-frequency AC coils at ~30 kHz; field extends ~60–80 mm above rail head
2. Wheel disturbance	Passing steel wheel flange and tread distort the magnetic field as they enter and leave the detection zone	Each wheel produces a characteristic two-peak signal signature — one peak as wheel enters field, second as wheel exits
3. Direction detection	Two coil elements in the sensor head are spaced ~50–70 mm apart; the sequence in which they detect each wheel determines travel direction	If element A detects before element B: direction is “in”; if B before A: direction is “out”; prevents double-counting at section boundaries
4. Axle count	Evaluator electronics classify each wheel-pair detection as one axle; increment the in-count or out-count accordingly	Two wheels per axle detected simultaneously by both sensor elements; count incremented once per axle, not per wheel
5. Section state decision	Evaluator continuously calculates (in-count − out-count); if result = 0 → Clear; if result > 0 → Occupied; if result < 0 → Fault	Negative count (more axles out than in) is a fault condition — physically impossible in normal operation — triggers fail-safe response

Sensor Head Design: How the Counting Head Mounts on the Rail

The counting head is a robust steel housing clamped or bolted to the web of the rail — the vertical section between the head (top, wheel-running surface) and the foot (base flange). The mounting on the rail web rather than between the rails is significant: it means the sensor does not require any modification to the rail running surface, does not need to penetrate the track structure, and is not exposed to the direct wheel contact zone.

Modern counting heads from leading manufacturers (Frauscher RSR180, Siemens ACS2000 series, Thales ELECTROCODE) are designed for zero-maintenance operation over the design life — typically 10–20 years in service — with no moving parts, sealed electronics rated to IP67 or better, and operating temperature ranges of −40°C to +70°C or beyond. The sensor head communicates with the evaluator (the signal-processing electronics, which may be located in a nearby lineside cabinet or at the signal box) via a cable or fibre connection.

Axle Counter vs Track Circuit: Full Comparison

Parameter	Axle Counter	Track Circuit
Detection method	Magnetic sensors at section boundaries; counts axles in and out	Electrical current through rails; train short-circuits relay
Broken rail detection	No — cannot detect rail fracture between sensor heads	Yes — fracture opens circuit → danger indication
Flooding / submersion	Unaffected — magnetic detection immune to water	Flooding causes false clear (water short-circuits detection)
Leaf / rail contamination	Unaffected	Leaf film can prevent wheel shunt → false clear (ghost train)
Insulated rail joints	Not required — no electrical dependency on rail continuity	Required (DC) or eliminated by AF coding
Steel bridges	Works normally — no insulation required from bridge steelwork	Very difficult — bridge steelwork creates low-resistance shunt paths
Traction return current	Unaffected by DC or AC traction return	DC return current interferes with DC track circuits; AF mitigates
Section length	Unlimited — no electrical resistance constraint	Limited by rail resistance (~1–4 km maximum)
Maintenance burden	Low — solid-state electronics; no rail joints	Higher — insulated joints, relay inspection, calibration
Reset after failure	Required after count loss — manual confirmation process needed	Self-resetting once section is clear and relay re-energises
Can transmit data to train	No — detection only	Yes — coded track circuits transmit speed data

Where Axle Counters Excel: The Preferred Technology Contexts

Steel Bridges and Viaducts

Steel bridge decks create a fundamental problem for track circuits: the bridge steelwork provides a low-resistance electrical path between the two rails, effectively short-circuiting the track circuit and preventing detection. Insulating the rails from the bridge structure — inserting insulating pads between the rail base plates and the bridge deck — is possible but expensive, technically difficult, and creates potential maintenance issues. Axle counters have no electrical dependency on rail-to-structure insulation and are the standard solution for train detection on steel bridges worldwide.

Tunnels Subject to Flooding

In tunnels where groundwater flooding is a regular operational condition, standing water between the rails creates a low-resistance shunt path that reduces track circuit voltage below the detection threshold — exactly the same effect as a train axle, but without a train present. A flooded track circuit indicates “occupied” even when the section is clear. Axle counters are entirely unaffected by water accumulation because their detection relies on the magnetic properties of passing steel wheels, not on electrical continuity between the rails.

Very Long Sections

Track circuit section length is constrained by rail resistance — the longer the section, the lower the circuit current, and the more difficult reliable detection becomes. On lightly trafficked freight lines with long block sections of 5–20 km, track circuits are impractical without intermediate repeater circuits. Axle counters can monitor sections of any length between the two sensor heads, with the cable connecting sensor to evaluator transmitting data over any required distance via fibre optic communication.

Lines with Intensive Leaf-Fall

Autumn leaf contamination (“black rail”) is a significant cause of track circuit failures on tree-lined routes — the compressed leaf film is electrically non-conductive and prevents the wheel shunt that track circuits rely on. Axle counters are unaffected by leaf contamination because the magnetic detection mechanism does not require electrical contact between wheel and rail.

The Reset Problem: Occupancy Without a Train

The most operationally significant disadvantage of axle counters is what happens when the axle count becomes inconsistent — when the system believes a train is in the section but the section is actually clear. This situation, called a “false occupancy” or “uncertain state,” can arise from:

Power failure mid-passage: A power interruption while a train is in the section loses the count. On restoration, the system does not know how many axles were in transit and defaults to “occupied.”
Maintenance trolley or on-track machine: A manually propelled trolley or small on-track machine that enters the section and is lifted off the track without being counted out leaves the system believing a vehicle is still present.
Count discrepancy: If a sensor misreads one axle (either missing a count or double-counting), the entry and exit totals will not match when the train has fully left the section.
Train division in section: If a train divides (becomes separated) while in the section, the rear portion may exit via the entry sensor — creating an exit-direction count that confuses the system.

In all these cases, the axle counter holds the section at “occupied” — a fail-safe response that prevents conflicting train movements — but requires a reset procedure before normal service can resume. The reset procedure requires a signaller (or in some systems, an automated verification process) to confirm that the section is visually or operationally clear, then issue a reset command that returns the evaluator to the “zero” state. On a busy metro line, a reset procedure mid-peak may take 5–15 minutes and affect several train paths.

Types of Axle Counter Evaluator Systems

System Type	Architecture	Typical Application	Key Suppliers
Single-section evaluator	One evaluator per section; dedicated hardware per detection zone	Simple installations; isolated sections on mixed systems	Frauscher, Voestalpine
Multi-section central evaluator	One central processor evaluates multiple sections; sensor heads connected by cable/fibre	Station areas, junctions, large installations with many short sections	Siemens, Thales, Frauscher
Distributed network (ACS)	Intelligent sensor heads communicate over data network; central server manages all sections	Large-scale resignalling; integration with ETCS/CBTC	Siemens ACS2000, Frauscher FAdC

Axle Counters and ETCS Level 3: The Critical Gap

The same broken-rail detection limitation that affects axle counters in the context of track circuit replacement is the central barrier to ETCS Level 3 deployment. ETCS Level 3 proposes to eliminate all fixed-block train detection infrastructure — both track circuits and axle counters — in favour of continuous train-position reporting. This creates the broken-rail detection gap: without track circuits or axle counters at intermediate positions, no infrastructure-based mechanism detects a fractured rail in an unoccupied section.

Importantly, even if axle counters replaced all track circuits, this gap would remain — axle counters at section boundaries cannot detect a broken rail within the section any more than ETCS Level 3 can. The gap is not specific to moving block; it is a characteristic of any detection system that monitors only train presence at section boundaries rather than continuous rail integrity.

This is why the ETCS Level 3 problem and the axle-counter-vs-track-circuit debate are related: both point to the same unresolved question of how to detect broken rails without continuous electrical monitoring of the rail itself.

Editor’s Analysis

The axle counter versus track circuit debate is a genuine engineering trade-off, not a case where one technology is clearly superior. Axle counters offer real and significant advantages in the environments where track circuits struggle — steel bridges, flood-prone tunnels, leaf-contamination routes, and long sections where rail resistance would limit track circuit reliability. Their lower maintenance burden and immunity to weather conditions make them the preferred choice for new installations on most European networks. But the broken-rail detection gap is not a minor caveat — it is a safety-critical limitation that must be actively managed through compensating measures: increased physical inspection frequency, alternative detection technologies (DAS, in-track monitoring), or route-specific risk assessments that justify the reduced inspection interval on axle-counter sections. The network operators who have converted extensively to axle counters — DB Netz in Germany is the largest example, with over 40,000 axle counter sections — have done so with this gap explicitly in their safety cases, backed by evidence that the residual broken-rail risk is managed by inspection regimes and by the lower inherent broken-rail frequency on well-maintained track. What has changed in the past five years is the emergence of DAS sensing as a credible broken-rail detection alternative. If DAS can be certified to SIL 3 or SIL 4 standard for broken-rail detection on mainline track — and the research programmes are making progress — it would remove the last argument for retaining track circuits specifically for their broken-rail capability, and would clear the way for complete axle counter replacement of track circuits on all route types. That outcome is probably five to ten years away, but it would represent the most significant change to railway train detection infrastructure since the track circuit itself was standardised in the 1880s. — Railway News Editorial

Frequently Asked Questions

Q: Why can’t an axle counter detect a broken rail?: An axle counter’s sensor heads are located only at the two ends of a section. The system has no information about the condition of the track between those two points — it only knows how many axles have entered and how many have left. If a rail fractures between the sensor heads when no train is present, the axle counts at both ends remain zero (or balanced), and the system correctly reports the section as “clear” — which is accurate in terms of train occupancy, but does not reflect the fractured rail that makes the section unsafe. A track circuit detects a broken rail because the fracture creates an open circuit that interrupts the detection current, causing the relay to drop and the signal to show danger. This continuous monitoring of rail electrical continuity — absent in axle counters — is the property that provides broken-rail detection.
Q: How is the axle counter reset procedure performed safely?: The reset procedure varies by system and by the cause of the uncertain state, but the core requirement is always the same: someone with authority must confirm that the section is actually clear of trains and obstructions before the evaluator’s count is zeroed. On a simple rural line, this might involve the signaller walking (or driving) the section visually, then issuing a reset command at the signal box. On a complex metro or mainline, the procedure may be more elaborate: a trained track worker performs a “walking inspection” of the section, reports to the controller, and a formal reset authorisation is issued. Some modern axle counter systems incorporate “cooperative reset” — where the passing of a known train through the section can be used to re-establish a consistent count without manual inspection. The safety principle is invariant: the reset may only be performed when the responsible person has confirmed, by direct observation or by operational procedure, that the section is genuinely clear.
Q: Can an axle counter detect a vehicle that is not a standard railway wagon?: Axle counters detect any ferrous (steel) wheel that passes through their magnetic detection zone. Standard railway vehicles — locomotives, coaches, wagons, DMUs, EMUs — all have steel wheels and are reliably detected. Maintenance vehicles with steel wheels are also detected. However, some maintenance trolleys and small on-track machines have very small wheels, or are lifted off the track rather than rolled off, which can cause the count anomalies described in the “reset problem” section. Rubber-tyred maintenance vehicles (which some airports and industrial railways use) would not be detected by axle counters at all — their rubber tyres do not distort the magnetic field as steel wheels do. For this reason, rubber-tyred vehicles working on axle-counter sections require specific operational procedures (possession, speed restrictions, direct communication with the signal box) rather than relying on the axle counter system to protect them.
Q: Are there situations where track circuits are preferred over axle counters for new installations?: Yes — the primary case for preferring track circuits in new installations is on routes where broken-rail risk is elevated and compensating measures would be onerous. This includes heavily loaded freight routes (higher tonnage accelerates rail fatigue and cracking), routes in cold climates with high thermal-gradient broken-rail frequency, and routes where the inspection interval required to manage the broken-rail gap safely would be operationally disruptive. Track circuits are also still preferred where their data transmission capability — carrying speed codes to ETCS Level 1 or cab-signalling trains — is required for the signalling architecture. For all other circumstances — new urban metro systems, modern high-speed lines (which have intensive maintenance regimes that manage broken-rail risk independently), mixed-traction environments, steel-bridge sections, flood-prone areas — axle counters are the preferred or mandatory choice for new installations on most major European networks.
Q: How many axles do common railway vehicles have?: Axle count varies significantly by vehicle type, which is why the axle counter’s entry and exit totals can vary from 4 to over 100 for a single train movement. A standard two-axle freight wagon has 2 axles. A four-wheeled bogie freight wagon (two bogies, two axles per bogie) has 4 axles. A typical passenger coach on bogies has 4 axles. An EMU vehicle (motor car on two bogies) has 4 axles. A standard 8-car EMU (Siemens Desiro, Bombardier Aventra, etc.) has approximately 32 axles. A 16-car TGV Duplex has 26 axles. The Eurostar e320 (18 cars) has 32 axles. When an axle counter registers a discrepancy between entry and exit count, the magnitude of the discrepancy can sometimes help identify whether the missing count represents a single axle (likely a count error) or a much larger number (suggesting a partial train still in section) — though in all cases, the safe response is to maintain the section as occupied until the situation is resolved.