ANPR vs DSRC vs GNSS in tolling: how the three identification technologies compare

Removing the barrier changes one thing above all: the toll system must identify every vehicle while it moves at normal speed. Three families of technology do this job. They are often presented as competitors, but the schemes that work best usually combine them. This article compares them; the ANPR, DSRC, RFID and GNSS pages explain each one in detail.

ANPR: the plate as identifier

Automatic number plate recognition uses cameras to read registration plates and match them to an account or a registered keeper. Its strength is universality: every vehicle has a plate, so no equipment needs to be distributed to users. That makes it the natural fallback in almost every free-flow scheme. In Norway, users without an AutoPASS agreement are invoiced from a photo of their plate; on Dublin’s M50, unregistered drivers pay after their journey on the basis of plate reads.

Its weakness is collection, not detection. A plate identifies a vehicle, not a willing customer. Korea Expressway Corporation’s smart-tolling pilot, which charged vehicles by plate recognition at nine tollgates between May 2024 and May 2025, found that 78.3% of plate-based charges were still unpaid after 15 days. Foreign vehicles, privacy rules and the cost of chasing small debts all weigh on schemes that rely on plates alone.

DSRC and RFID: the tag as identifier

Short-range radio tags (DSRC in most of Europe and Australia, passive RFID stickers in much of the Americas and Asia) are read by antennas on gantries. A tag is linked to a paying account, so transactions are cheap to process and collection rates are high. Taiwan’s nationwide distance-based tolling, in service since 30 December 2013, relies on passive RFID eTag stickers; Melbourne’s CityLink has read DSRC transponders at freeway speed since 2000; Dubai’s Salik deducts tolls from prepaid accounts linked to windscreen tags.

The limitation is coverage: drivers without a tag still have to be identified some other way, which is why tag-based schemes almost always add cameras. Malaysia’s Touch ’n Go illustrates the hybrid approach, with lanes that combine dual RFID antennas, ANPR and machine learning, RFID remaining the primary identifier and ANPR the secondary check.

GNSS: the vehicle reports its own position

Satellite tolling moves the measurement into the vehicle. An on-board unit determines its position, matches it against the charged network and reports usage, so the road needs no gantries for charging. This suits large networks and distance-based charges for heavy vehicles. Germany’s truck toll covers more than 50,000 km this way, Belgium’s kilometre charge has run on GNSS on-board units since 2016, and the Netherlands introduced a satellite-based truck toll in July 2026. Singapore is moving its urban road pricing to GNSS with ERP 2 from 1 January 2027, removing its charging gantries.

GNSS does not remove the need for roadside equipment: it moves it to enforcement. Germany’s system relies on roadside monitoring columns; Belgium’s enforcement uses stationary and mobile units combining ANPR cameras, DSRC readers and laser detection; the Dutch truck toll is enforced with gantries and movable systems that read on-board units by DSRC and plates by ANPR.

How to choose

• Network size and vehicle type: GNSS is most cost-effective for large networks and heavy vehicles; tags and cameras suit motorway sections, bridges, tunnels and urban cordons.
• Share of occasional and foreign users: the higher it is, the more the scheme depends on ANPR and on cross-border debt collection or EETS providers.
• Collection risk: tag-based accounts collect best; plate-only charging needs strong enforcement and recovery processes.
• Enforcement: every model needs it, and in GNSS schemes it is where most of the roadside investment goes.

The practical answer is rarely one technology. Most mature schemes treat the tag or on-board unit as the primary channel and ANPR as both the fallback for non-equipped vehicles and the basis for enforcement.