What a Working Programme Looks Like

Rather than a summary, a description of the arrangement these notes point toward.

The measurement

Speed read from the vehicle rather than derived from position.

Thresholds set per vehicle type from a month of your own silent logging.

Calibration verified and mounting checked at fitting, re-verified after any device change.

System-initiated events labelled separately from driver-initiated ones.

Minimum exposure below which no score is produced.

The score

Formula visible to the people it describes.

Exposure basis suited to the work — not per mile for urban rounds.

Compared against a driver's own history, never as a public ranking.

Not linked to pay.

And never used as an allegation, because a score is a construction rather than conduct.

The video

Forward-facing by default; cab-facing only with a documented, specific risk.

Event-triggered, no continuous retention, no live viewing.

Audio off.

Access on incident only, logged, with the driver told when their footage was viewed.

Retention in days, with an incident hold.

The response

Coaching within days, about a specific event, with the context.

Conducted by someone trained, not by whoever would run a disciplinary.

Serious threshold defined narrowly and published.

Structural causes acted on when coaching finds the same thing across several drivers.

The channels back

Drivers see their own events, score components, thresholds and viewing log.

A dispute route with a named reviewer and a response time.

Near-miss reporting in thirty seconds, with the associated score event removed where the manoeuvre was correct.

And a visible response: junctions fixed, rounds resized, vehicles changed.

The honest claims

Outlier identification, a coaching channel, and evidence after an incident.

Not a collision reduction attributed to scoring alone, which the evidence does not support and which discredits the programme when it fails to appear.

What it produces

A system drivers use rather than work around.

Findings about the operation, not only about people.

And a position that survives a consultation, a grievance and a tribunal, because every part of it can be explained to the person it describes.

Reviewing it annually

Six questions that describe whether it still works.

Did collisions and severity move, over a long enough period?

Is the false trigger rate falling?

Is time from event to coaching still days rather than weeks?

Did near-miss reporting hold?

Can the score still be explained to the person it describes?

Would you deploy it the same way again?

Reproduce the workflow

For another way to make this requirement testable, consult inspect the service connection. Reproduce the case with real roles, failures and recovery steps.

Independent reference

For a thematic point of reference, see TRL. Use the current material as an outside check before turning the principle into a system rule.