Fabrication · Inspection & QC

You know the NCR count. You don’t know which NCR it keeps being.

Inspection is where fabrication finds out what cutting, welding and coating did. A pass/fail count and an NCR total tell you how much you lost; they do not tell you that this month’s hydrotest failures are the same weld procedure that failed last month. On pipe the stakes rise — a hydrotest fail is a cut-out and re-weld, and the repair loop is the cost.

Your ERP records the result: passed, rejected, NCR raised.

It won’t tell you this is the third batch running that failed on the same joint detail, or how many repair loops it took to clear.

Why the gap exists

Where the losses hide

The NCRs that are the same NCR.

An NCR count is a total. A recurring defect is a cause — and causes are what you can act on.

NCRs that repeat

Weld defects, dimensional faults and coating failures each trace to a different stage. Counted together they average into a number; grouped by family and traced back, they point at the procedure or the shift that keeps producing them.

Hydrotest failure clustering

A hydrotest fail is expensive — a cut-out, a re-weld and a re-test. The failures cluster by joint detail and procedure, and reading them that way is the difference between fixing a WPS and re-testing forever.

Repair loops nobody counts

A joint that fails, is repaired and fails again has been through three cycles booked as one. The loop is invisible unless each pass is logged against the joint.

The join

What the record holds, and what Munshi captures beside it.

Same run, same clock, same asset — three streams that only mean something together.

Your ERP records

Inspection outcome, NCRs raised, hydrotest pass/fail, disposition.

Munshi captures

The defect and hydrotest result tagged by cause against the joint, weld procedure and batch — and NDT results tied to the same record.

What that makes possible

NCRs and hydrotest failures grouped by cause and traced to the procedure that produced them, with the repair loop counted — instead of a total that resets every month.

What Munshi surfaces

Which defect, how often, and from which joint.

Teal is normal, maroon is flagged. Illustrative data.

NCRs by cause

Grouped by cause at the bench, not lumped into a single NCR count.

Hydrotest repair loops, batch over batch

Average repair passes per failed joint. A rising line is a procedure that has not been fixed.

Normal / in-spec Flagged by Munshi -- threshold / average

When an outcome moves and a signal moves with it, Munshi shows both and says they moved together. It does not tell you one caused the other. The operator confirms the cause, and the confirmed cause is written back to that asset’s register — so the pattern is already named the next time it appears.

Related stages

The stages either side of this one.

Surface prep & coating

The coat that went on, and the part of it that went on twice.

Other assets

Cranes, hydraulic packs, compressors, extraction and utilities.

Get started

See this on your own line.

Bring a month you already understand. The useful test is whether Munshi finds what you already know — and then what you didn’t.