Every construction platform promises accuracy. Models are imported, coordinates are established, control points are set, and crews head into the field expecting what they see on screen to match what is actually on the ground.

But how do you know it does?

Do you know how to validate that a control point is actually accurate? Can you compare the physical control point with the same point’s location inside the model? What happens when control points on one side of a project agree perfectly, but points on the other side don’t? And if everything you’re redlining suddenly appears too high or too low, how do you determine whether the problem is the GNSS equipment, the vertical datum, the model, the control or something else entirely?

These aren’t hypothetical questions. They are problems we’ve encountered on real construction projects.

Most of the time, the alignment between design and field works exactly as expected. But on high-stakes projects, where tolerances are tight and multiple design disciplines converge on the same physical space, small errors from different sources can stack up quickly. The result is often not an obvious error message. It’s something more dangerous: data that looks believable but isn’t quite right.

That’s why every vSite subscription includes access to trained geospatial specialists. They aren’t simply a help desk for software questions. They understand coordinate systems, model registration, GNSS, control point validation, vertical datums, and the chain of accuracy decisions that determines whether a construction team can trust what it sees in the field.

For straightforward projects, much of this work happens quietly during setup. On complex projects, it can be the difference between building with confidence and building on top of a problem nobody has identified yet.

When Small Errors Start Stacking

A recent critical-facility project shows how quickly these assumptions can unravel.

The project—a large-scale data center—required approximately 4in/10 cm tolerances for installation and data collection. That’s demanding, but achievable with the right controls.

The problem wasn’t one large mistake. It was several smaller ones.

Models supplied by different design disciplines did not align properly. Structural, mechanical, and electrical information that should have shared a common reference didn’t.

Then the structural model was rotated during the project. The change was intended to correct an earlier issue, but it introduced discrepancies of up to 2ft/60cm in the field.

Surveyors subsequently corrected the positioning to within approximately 1.5–2in/3-5cm. On its own, that would have been well within the project’s tolerance.

But it wasn’t the only source of error.

When the remaining survey uncertainty was combined with the discrepancies between design disciplines, the team was dealing with an effective positional uncertainty of approximately 10–12in/25-30cm—far outside the project’s 4in/10cm requirement.

Each individual issue could be explained. Together, they created a much bigger problem:

The field team no longer knew what to trust.

That’s when accuracy problems become expensive. Crews slow down. Supervisors start making additional calls. Decisions get deferred. Work may be installed and questioned later. And once concrete is poured or utilities are buried, discovering that the reference was wrong becomes a very different problem.

The Questions That Matter in the Field

This is where construction teams need to go beyond asking whether their GNSS receiver is showing an “accurate” position.

Suppose you’re standing over a known survey control point. Your GNSS says you’re within a few centimetres. Does that point also land where it should inside the model?

Now walk 500 metres across the project and check another control point. Does it still align?

If the first point matches but the second doesn’t, simply shifting the entire model won’t solve the problem. You may be dealing with rotation, scaling, an incorrect coordinate transformation, inconsistent survey control, or a problem in the source model itself.

Vertical discrepancies can be even harder to diagnose.

If redlined utilities or collected points consistently appear too high or too low, what do you correct? The antenna height? GNSS configuration? Geoid? Vertical datum? Model elevation? Survey control?

Changing the wrong one may make one part of the project look correct while making another part worse.

These are not questions most superintendents or project managers should be expected to answer themselves. But someone needs to answer them before the crew starts building against questionable information.

Getting the Project Back to a Trusted Reference

On the data center project, the vSite geospatial team worked directly with the project’s surveyor, general contractor, and design coordinators.

The objective wasn’t simply to “make the model line up.” It was to understand why it didn’t line up.

The team validated site control, investigated the discrepancies between models, and developed an alignment strategy that brought the information back within the project’s required tolerance. Just as importantly, the result could be checked against known references rather than accepted because it looked right on a screen.

That gave everyone something they had been missing: a common, validated reference.

The field team knew what it could build against. The surveyor had a clearer picture of the remaining accuracy constraints. The project team had evidence showing how the discrepancies had been identified and addressed before they became installation problems.

Without that intervention, the project could have spent weeks trying to isolate whether the issue belonged to survey, design, software, or field operations. Under schedule pressure, there’s also a more concerning possibility: the crew simply continues working while everyone assumes someone else has verified the data.

Accuracy Is a Chain, Not a Number

One of the biggest misconceptions about digital construction is that accuracy comes from the device.

A GNSS receiver may be capable of centimetre-level positioning, but that doesn’t mean the information displayed on top of that position is centimetre accurate.

The entire chain has to be right.

The survey control has to be right. The coordinate system has to be right. The vertical datum has to be right. The model has to be registered correctly. Different discipline models have to agree. Equipment needs to be configured correctly. And when something changes during construction, the impact of that change needs to be understood.

A highly accurate GNSS receiver positioned against an incorrectly registered model simply gives you a very accurate position relative to the wrong information.

That’s why field validation matters.

Why Geospatial Support Comes with vSite

Most construction organizations don’t have geospatial specialists sitting beside the superintendent.

Surveyors manage control and positioning. Designers manage models. Project managers coordinate the work. IT manages systems. Field crews build.

Problems often hide in the gaps between those responsibilities.

When something doesn’t line up, the superintendent doesn’t necessarily care whether the root cause is EPSG configuration, a vertical datum transformation, model rotation, or bad control. They need an answer to a much simpler question:

Can I trust this enough to build from it?

That’s the gap our geospatial team is there to help close.

We work with surveyors, design teams, and field crews to establish a reliable data foundation during setup—and help diagnose what changed when something suddenly doesn’t look right in the field.

It’s included with vSite because we’ve learned something important after thousands of projects: software can display information perfectly and still show the wrong thing if the underlying spatial data isn’t right.

And on a construction project, discovering that after you’ve built it is far more expensive than asking the right questions before you start.