Skip to content
LeakSonic
Approach

How we earn a skeptical engineer’s trust.

Sentrix is built on tested claims, not assumed capability. This page is how we work - with practising engineers, with evidence attached to every finding, and with honesty about what we do and don’t do. It is a genuine differentiator, and we publish it with confidence.

A capability you assume is a risk you’ve hidden. A capability you test is a claim you can defend.

Everything below is framed as a claim with a defined test, not a finished result. As each test concludes, we update its status here - and we report outcomes whether they confirm the claim or break it.

See how it works

Flight to decision, in one interactive walkthrough

The same flow described throughout this site - drone evidence, standardised by Sentrix, reviewed by an engineer, resolved into a decision - shown here as a small interactive 3D scene. Drag to orbit, or step through it below.

Loading simulation…
Elapsed
0.0 hrs
Covered
0.00 km
Crew
1, on foot
Step 1 of 6 · Manual patrol

Today, most rights-of-way are still walked on foot with a handheld gas detector. It works - but it is slow, it covers a small stretch at a time, and it depends on who is walking and what they happen to notice.

Illustrative, time-lapsed simulation - drag to orbit, tap a step to jump to it. Coordinates, timings, and the findings shown are for demonstration only, not a real network or engagement.

How we work

Four principles behind every claim we make

These are not slogans. They are the standard we hold our own product to before we ask an operator to trust it.

We validate with practising engineers

Every claim is tested against how integrity engineers actually work, not against a lab ideal. If it does not hold up in a real workflow, it does not ship.

We attach evidence to every finding

Nothing Sentrix outputs is a bare score. Each finding carries the evidence behind it, so the engineer stays in control of the judgement and can defend it later.

We say plainly what we don’t do

Sentrix does not measure pipe-wall thickness and does not sense underground. Naming a tool’s limits is the clearest signal that its capabilities are real.

We publish outcomes honestly

We state what we are proving and report the result either way - including when it forces us to change course. Confirmation and disconfirmation are both worth publishing.

Claims under test

What we prove before we claim it

If any of these turns out to be false, Sentrix has to change. That is exactly why we name them explicitly rather than assume them.

C1
Prioritisation actually changes what an engineer would do.
C2
Comparing one cycle to the last is reliable enough to trust.
C3
A finding holds up against real ground truth.
C4
Engineers will act on an evidence-backed finding.
C5
The platform gives a working engineer real time back per cycle.
All five assumptions are currently in progress. Status updates here as each test concludes.

Active validationAll five claims below are currently being tested with practising engineers - status updates here as each concludes.

C1
Claim

Prioritisation actually changes what an engineer would do.

How we test it

We work directly with practising integrity engineers to test whether a ranked target list changes their decisions - or whether existing judgement already suffices. If it does not help, it is not worth shipping.

C2
Claim

Comparing one cycle to the last is reliable enough to trust.

How we test it

We check our cycle-over-cycle comparison against known, ground-referenced change, so a difference we surface reflects something real rather than noise - and we characterise where it is dependable and where it is not.

C3
Claim

A finding holds up against real ground truth.

How we test it

We test our findings against independently confirmed conditions to understand where they are dependable and where they are not - because an honest statement of limits is worth more than an accuracy headline.

C4
Claim

Engineers will act on an evidence-backed finding.

How we test it

We put findings in front of domain experts, with and without their supporting evidence, and measure whether the evidence changes their willingness to act - because a finding no one trusts is a finding no one uses.

C5
Claim

The platform gives a working engineer real time back per cycle.

How we test it

We compare the current review-and-report workflow against a Sentrix-assisted one on the same data, isolating exactly where time is saved and where it is not, rather than asserting a savings figure.

Working together

Where we invite collaboration

We work with academic and industry partners on the harder questions in pipeline integrity and inspection decision-making - from validating findings against real ground truth to understanding what makes an engineering team genuinely trust an automated finding. If that overlaps with your work, we’d like to hear from you - including if you think one of our claims is wrong.

Work on these with us.

If you’re a researcher, lab, or integrity team working on pipeline inspection, decision support, or validation methodology, we’d like to hear from you - including if you think one of our claims is wrong.