January 2025: From Possibility to the Ground

For months, we had been talking about a simple idea.

What if a vehicle did not need to slow down, stop at a barrier, wait for identification and then complete a transaction? What if technology could take care of everything while the vehicle simply continued its journey?

By January 2025, we did not want to keep discussing the idea.

We wanted to see it working on the ground.

This was the point where our ANPR journey started moving from research, meetings and possibilities towards actual infrastructure and implementation.

And that changed everything.

No Barrier. No Stop. Just Movement.

Our objective was becoming clearer.

We were not interested in creating another system where a vehicle approached a gate, stopped, waited for a camera to read its number and then moved ahead.

That would solve only part of the problem.

Our thought was much more ambitious.

The vehicle should continue moving naturally.

A car could pass.

A truck could pass.

Different categories of vehicles could pass.

The system should identify the registration number, process the required information and enable the relevant transaction without creating another physical bottleneck.

There should be no dependency on a traditional toll barrier merely to make the technology work.

That was the direction we wanted to pursue.

Then Came the Real Test: Speed

Reading a stationary number plate is one thing.

Reading it from a vehicle moving on an actual road is completely different.

And we wanted to push the system further.

What happens when a vehicle passes at more than 120 kmph?

Can the camera capture it?

Can the system identify the number correctly?

Can it process the image quickly enough?

What happens when multiple vehicles pass?

What happens with trucks where the plate position, size or condition may vary?

What happens during different lighting conditions?

What if one character is unclear?

Suddenly, every passing vehicle became a test.

Every incorrect reading became a learning.

Every successful reading gave us more confidence.

This was no longer innovation on a presentation.

Cars were passing. Trucks were passing. Vehicles were moving at highway speeds. And the system had to respond in real conditions.

A Number Plate Is Easy Until You Try Reading Thousands of Them

One of the biggest learnings came from something that looks extremely ordinary.

The number plate.

In theory, every registration number follows a structure.

On the road, reality can be very different.

Some plates are old.

Some are dirty.

Some are damaged.

Some have unusual spacing.

Some are partially hidden.

The angle can change.

Lighting can change.

The speed can change.

Trucks can create an entirely different reading environment from cars.

And a single wrongly identified character matters enormously when that registration number is going to become the basis for a financial transaction.

If the system reads the wrong vehicle number, this is no longer simply an ANPR accuracy problem.

It can become a wrong transaction.

That changed how seriously we looked at every weak point.

Find the Weak Number, Then Solve for It

Our approach started becoming very practical.

We did not want to celebrate only the registration numbers that the system could read correctly.

We wanted to find the ones it could not.

The difficult number plate was more valuable to us than the easy one.

Why was a particular plate not being recognised?

Was it the camera?

Was it the angle?

Was it speed?

Was it lighting?

Was it the plate itself?

Was a particular character combination creating difficulty?

Was the problem happening more frequently with a certain type of vehicle?

Could infrastructure placement improve the result?

Could software improve it?

Could another validation layer reduce the possibility of a wrong identification?

Find the weak point. Understand it. Correct it. Test again.

That became the process.

Getting the Nods Changed the Energy

Another important change was happening simultaneously.

For a long time, whenever we discussed this concept, one of the most common responses was that it would be extremely difficult to implement.

And those doubts were understandable.

ANPR itself was not the only challenge.

There were financial transactions involved.

FASTag integration had to be understood.

Infrastructure had to work.

Vehicle information had to match.

Operational exceptions had to be addressed.

The relevant permissions and stakeholder support mattered.

But slowly, conversations started becoming more positive.

We were getting the necessary nods to take things ahead.

The discussion was no longer limited to:

“Can this really happen?”

It was moving towards:

“How do we implement it?”

For us, that was a major shift.

Infrastructure Became Part of the Product

Being a technology company, it is natural to think primarily in terms of software.

But this project was teaching us that software alone would not solve the problem.

Camera positioning mattered.

Height mattered.

Distance mattered.

Vehicle speed mattered.

Lighting mattered.

Lane configuration mattered.

Connectivity mattered.

The physical environment mattered.

The software and infrastructure had to behave like one system.

That meant getting our hands dirty with things that might not traditionally look like the work of a digital logistics platform.

But that is what implementation demands.

You cannot say, “Our software works, the infrastructure is somebody else’s problem.”

If the complete system does not work, the customer does not care which component failed.

From ANPR to an Actual Transaction

There was another distinction that remained extremely important for us.

We were not trying to demonstrate that a camera could read a registration number.

That had already been done in different applications.

Our ambition was to explore the complete journey.

Vehicle passes -> Registration identified -> Vehicle validated -> Relevant system connected -> Transaction processed -> Vehicle continues moving.

That last part changed the seriousness of everything before it.

Once actual money enters the system, accuracy, reconciliation, security and exception management become critical.

A demo can tolerate an error.

A financial infrastructure cannot.

That was why we were spending so much time understanding the edge cases rather than only demonstrating the happy path.

Also Read: December 2024: The Year That Made Us Ask Bigger Questions

The Road Became Our Laboratory Again

There was something deeply familiar about this phase.

Years earlier, when we were building TruckSuvidha, we had gone to transport markets and roads to understand how the industry actually worked.

Now, more than a decade later, we were again standing close to roads, watching vehicles pass and trying to understand another transportation problem.

Only the questions had changed.

Earlier:

How do we connect this truck with a load?

Now:

How do we identify this moving vehicle accurately enough that infrastructure can interact with it automatically?

Technology had changed enormously.

But one principle of our journey had survived.

When you want to solve a problem for the road, go to the road.

January 2025 Was Different

This was why January felt positive.

Not because everything had suddenly been solved.

It had not.

There were still technical challenges.

There were still difficult number plates.

There were still operational questions.

There were still integrations and processes that needed to mature.

But something fundamental had changed.

We had started implementing.

Vehicles were no longer boxes moving across a presentation.

They were passing in front of actual infrastructure.

Cars.

Trucks.

Different vehicle categories.

Different speeds.

Different conditions.

And every vehicle was helping us understand what needed to be improved next.

After months of asking whether our vision could work, we had reached a much more exciting stage.

Build it. Put a vehicle through it. Find what breaks. Fix it. Run it again.

That was January 2025.

The dream had finally left the meeting room.

It was now standing beside the road.