Why open road tolling fails at the back office, not the gantry

Open road tolling has transformed what happens at the roadside, allowing vehicles to move through tolling points without stopping while the infrastructure above them captures the journey.

‍

But removing the toll booth also changed something less visible. The booth created a natural pause in the operation, giving the systems behind it time to confirm payment, reconcile accounts and deal with exceptions before the vehicle continued.

‍

Open road tolling removes that pause entirely.

‍

Vehicles continue at road speed while cameras and roadside infrastructure capture the event. For the driver, the improvement is obvious. There’s no booth to approach and no queue to wait in. But the operational complexity that used to sit at that physical checkpoint hasn’t disappeared.

It has moved into the back office. And that’s where many open road tolling systems still struggle.

‍

Operators have invested heavily in roadside infrastructure capable of identifying vehicles accurately at speed, while the systems receiving those events are often built around an operating model where processing could happen later.

‍

That mismatch is behind many of the problems still associated with open road tolling.

The gantry gets blamed because it’s the part everyone can see

When something goes wrong, the first question is usually whether the vehicle was read correctly.

‍

That makes sense. Cameras and gantries are visible parts of the operation, and ANPR accuracy has historically been an obvious place to look when a charge is disputed or a vehicle appears to have passed through without being recognised.

‍

But modern roadside hardware is increasingly reliable. Reads can still fail, particularly in difficult conditions, but repeatedly upgrading cameras doesn't address what happens when the image and registration were correct in the first place.

‍

The more important question is what happened next:

• Was that registration connected to the right account? 
• Had payment been confirmed when the charging decision was made? 
• Was an exemption visible to the system at the point it mattered? 
• If enforcement followed, was it based on a complete view of the journey or simply the information available at that moment?

‍

This is the same problem we've seen in parking operations, where ANPR is frequently blamed for failures that actually happen much further downstream.

‍

The camera can capture exactly what passed beneath it and the operation can still reach the wrong outcome. At that point, replacing the camera only gives the back office better information to mishandle.

Open road tolling changes the speed at which decisions have to happen

The fundamental difference in an open road environment is that every vehicle movement becomes a live transaction.

‍

There is no physical checkpoint holding the vehicle while the operation catches up. The vehicle passes beneath the gantry and continues its journey, which means the systems behind it need to establish what that event means without relying on a later manual process to complete the picture.

‍

That creates a very different requirement from the tolling environments many legacy platforms were originally designed to support.

‍

Account status needs to be understood alongside the vehicle event. Payments and exemptions have to be recognised early enough to influence the decision being made, while enforcement needs an accurate picture of what happened before it acts.

‍

Legacy back offices built around batch processing and later reconciliation assume there will be time between an event taking place and its operational consequence.

‍

Open road tolling removes that assumption.

‍

This is why simply removing toll booths from an existing operation doesn't necessarily create a modern tolling platform. The roadside experience changes immediately, but unless the architecture behind it changes as well, the operation is still processing live journeys using systems designed for a slower environment.

‍

That gap is where problems begin to surface.

The failures happen after the vehicle has already gone

A driver with a valid account passes beneath a gantry and the registration is captured correctly. The account confirmation doesn't reach the enforcement process in time, so a penalty is raised anyway.

‍

Elsewhere, an exempt vehicle moves through the network. The exemption exists, but it isn't available to the charging system when the event is processed. The charge is created and someone has to correct it afterwards.

‍

Neither journey failed at the roadside.

‍

The information existed. The problem was whether the right systems had access to it at the moment the decision needed to be made.

‍

Once that happens at volume, the operational impact extends well beyond the original error. Disputes need investigating, penalties get cancelled, account histories have to be reconstructed and teams spend time reconciling outcomes that should have been correct without their involvement.

‍

As we explored in this blog post, much of the cost in legacy tolling operations sits in this work around the transaction rather than the transaction itself.

‍

And because every individual case can eventually be resolved, the architecture underneath rarely looks completely broken.

‍

The operation continues. It just requires more human intervention than it should.

Foreign vehicles expose the problem particularly quickly

Foreign-registered vehicles are one of the clearest examples of why roadside accuracy alone isn't enough.

‍

Reading the plate is only the beginning. Once that registration has been captured, the operation still needs a viable way to understand who is responsible for the journey, determine whether payment has been made and decide what happens when it hasn't.

‍

Domestic vehicles generally operate within established account, payment and enforcement structures. Cross-border journeys introduce additional complexity because identification and enforcement may depend on information held elsewhere or processes that operate differently between jurisdictions.

‍

That creates a disproportionate challenge in open road environments.

‍

A high-quality image of a foreign registration doesn't protect revenue by itself. If the back office cannot connect that event to an appropriate payment or enforcement route, the vehicle can leave the network with the roadside infrastructure having performed exactly as intended.

‍

This is why foreign vehicle handling cannot be treated as an exception process bolted onto a domestic tolling platform. 

‍

In an open road operation, it has to be part of the operating model from the beginning. Otherwise, the journeys that are hardest to process become the ones most likely to require manual intervention or disappear into revenue leakage entirely.

A back office built for open road tolling behaves differently

The defining feature of a modern tolling back office isn't simply that it processes transactions faster. It's that it treats vehicle movements as events that need to be understood when they happen.

‍

When a registration is captured, the platform needs enough context to determine what should happen next. Existing accounts and exemptions have to be recognised as part of that decision, while payment status and enforcement rules need to contribute to the same operational picture rather than being reconciled later across separate systems.

Just as importantly, the decision needs to remain explainable afterwards.

‍

If a driver disputes a charge, the operation should be able to establish which vehicle event triggered it, which rule was applied and what information the system used to reach that outcome without someone manually rebuilding the journey across multiple platforms.

‍

That audit trail becomes increasingly important as physical control disappears from the roadside.

‍

The toll booth made the operating model visible. Drivers understood that they stopped, paid and continued. But open road tolling asks the back office to achieve the same level of certainty without the vehicle ever stopping.

‍

That requires architecture designed around the journey as it happens, rather than a collection of records waiting to be reconciled later.

The question isn't whether the camera got the read

Open road tolling has changed the roadside experience dramatically, but the back office hasn't always changed with it.

‍

That's why operators can invest in better cameras and more capable roadside infrastructure while continuing to experience the same disputes, reconciliation workload and revenue leakage behind the scenes.

‍

The vehicle read matters, but it's only the beginning of the transaction. What determines whether the operation actually works is everything that happens next: whether that event can be connected to the right context quickly enough for the system to reach the correct decision without waiting for someone to fix it afterwards.

‍

The industry has spent years asking whether cameras are accurate enough. For open road tolling, the more useful question is whether the back office can act accurately on what those cameras see, at the moment it matters.

‍

If it can't, the gantry isn't the problem; the architecture behind it is.

‍