Welding is the stage where quality and cost are both decided and both hidden. A joint that fails NDT is cut out and re-welded — double the wire, double the time, on a metre already counted. And the gap between arc-on time and shift time is where the real capacity goes, on a line the ERP never draws.
Your ERP logs the weld metres deposited.
It won’t tell you forty of them went back for repair, or that the arc was on for less than half the shift.
Why the gap existsWhere the losses hide
A repaired joint is one metre in the output and two in the labour and wire. The repair is invisible unless it is logged against the joint that needed it.
Porosity, undercut, lack of fusion and cracks each trace to a different cause — gas, technique, fit-up, procedure. Counted as a single repair rate, they average into a number that points nowhere; tagged by cause, each points at a fix.
Arc-on time is the honest measure of a welding station’s capacity. The difference between it and shift time is setup, positioning and waiting — real hours that never book as anything.
Wire, gas and electrode per metre move with technique and with procedure discipline. Per-metre cost is the number that shows a WPS being drifted from, and it is the one nobody keeps.
The join
Same run, same clock, same asset — three streams that only mean something together.
Weld metres or joints, consumables issued, and the NDT outcomes booked later.
Arc-on time from the welding set, the welder’s repair log tagged by cause, and NDT results tied to the joint — with consumption tied to the station.
A repair rate you can read by cause and by shift, arc utilisation per station, and consumable cost per metre — the three numbers welding usually hides.
What Munshi surfaces
Teal is normal, maroon is flagged. Illustrative data.
From NDT, tagged by cause at the joint — so gas, fit-up and technique separate out.
Arc-on as a share of shift time. The low bar is a shift that went to fit-up and waiting.
Normalised per metre. The spikes are the runs that carried repairs.
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 piece that came off the rolls to size — the second time.
Delays and mismatches that never book as downtime.
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.