Request Demo

Evidence before claims

How we turn technical information into operational guidance.

Our method separates manufacturer specifications, standards, field observations and customer results so readers can see what is known, what must be tested and what depends on a particular deployment.

Last reviewed: 17 August 2026 · Owner: GT AFRIK Technical Team

The core principle

A telemetry value is not automatically an operational conclusion. Position, fuel level, engine state and controller alarms all require context. Our guides explain the measurement chain, the assumptions and the checks needed before a team acts.

Evidence hierarchy

  1. Primary technical sources: standards, official manufacturer manuals, product specifications and operator documentation.
  2. Direct operational evidence: installation records, calibration tables, device logs, platform events and support findings that can be traced to a deployment.
  3. Customer outcomes: only published when the baseline, period, calculation method and customer permission are available.
  4. Secondary sources: used for context, never to replace a more authoritative primary source.

Publication workflow

1. Define the question

We begin with a specific operational decision, such as how to investigate a fuel drop or choose telemetry inputs for a generator controller.

2. Map the evidence

We identify the device, sensor, controller, communications path and software involved, then link the relevant primary documentation.

3. Review limitations

A reviewer checks coverage assumptions, calibration requirements, hardware variants, power conditions and any point that could create a false conclusion.

4. Publish and revisit

The page carries an author, publication date, review date and sources. We update it when specifications, standards or field evidence materially change.

How claims are labelled

Claim typeRequired support
Product capabilityOfficial specification or manual, plus the exact model or interface.
Recommended practiceTechnical rationale, operational assumptions and a clear limitation.
Performance resultDefined baseline, measurement period, formula, supporting records and permission.
EstimateStated assumptions and a label that distinguishes it from a measured result.
What we will not do

We will not invent customer names, deployment totals, savings, credentials or quotations. An unpublished result remains internal until it can be verified and approved.

Corrections and review cycle

Operational guides are reviewed at least annually and sooner when a cited specification or standard changes. Material errors are corrected promptly under our editorial policy. Readers can report a technical error by email.