What the Data Cannot Tell You

Vehicle data is precise about narrow things. The gap between precise and comprehensive is where the wrong conclusions live.

Whether a driver is careless

Harsh deceleration counts depend on route, load, traffic and other road users.

An emergency stop for a cyclist scores identically to habitual late braking.

Where to look instead: collisions and near misses, investigated individually.

Why fuel consumption rose

It shows that it did, per vehicle.

Not whether the cause is a fault, a tyre, a load change, a route change or the weather.

Where to look instead: the vehicle first — tyres, filters, codes — because it is a fault more often than anything else.

Whether a vehicle is needed

Utilisation shows how often it moved.

Not whether the work it does could be absorbed elsewhere, or whether it is a specialist asset used rarely and correctly.

Where to look instead: the operation, over a full year.

Whether a fault is urgent

A code is a signal, not a severity.

The same code can be transient or terminal depending on context and recurrence.

Where to look instead: the triage list built from your own history, and a workshop.

What a vehicle will cost next year

The trend is informative; the projection is not a forecast.

A single major component failure dominates a year and is not predictable from a curve.

Where to look instead: cohort averages for planning, individual curves for replacement decisions.

Whether the depot is well run

Cost per mile reflects geography and vehicle mix more than management.

Where to look instead: the four figures within a site's control.

What it does support

That a component reported a fault, at a time.

Actual usage, in distance and hours.

Consumption, where it is measured rather than estimated.

Condition indicators that predict a failure.

Four things, reliably, and defending that boundary is what keeps them trusted when a fifth is challenged.

Do not produce the weak figure

The discipline that keeps the rest credible.

When asked for something the data cannot support, saying so is the answer.

A number produced anyway circulates without its caveat and returns as an expectation.

In this subject it ends up in a replacement decision or a depot comparison, where being wrong costs real money.

Follow the difficult record

Use see the documented scenario to frame one representative case. The useful evidence is the record created when a value is corrected, approved and exported.