Forming is a game of tolerance against material that fights back. Springback, a worn die, an out-of-spec plate — each produces a piece that reads fine until it is checked, then goes back for another pass. On a pipe mill the same story is ovality and diameter drift, and the mill downtime around it rarely books as downtime.
Your ERP logs the pieces formed.
It won’t tell you a third of last shift’s sections needed a second hit, or that the mill kept stopping for a reason nobody recorded.
Why the gap existsWhere the losses hide
A part re-formed is one good part in the count and two passes in the time. The second pass is invisible unless it is logged as one.
Material springs back by an amount that moves with grade, thickness and tool wear. The piece that comes back for a second hit is real time on a part already counted, charged to nobody in particular.
Out-of-tolerance parts cluster by tool and by setup far more than they spread evenly. Averaged, the cluster vanishes; read against the tool, it points at the fix.
On a pipe mill, ovality and diameter drift with tooling and setup, and the stops to correct them come up as fewer pipes with no reason attached.
The join
Same run, same clock, same asset — three streams that only mean something together.
Pieces and pipes formed, and the dimensional rejects booked later.
Machine and mill signals, and the operator’s rework and downtime reasons tagged when they happen — tied to the tool and setup running.
Rework and dimensional NCs read by tool and setup, and mill downtime read by cause, instead of a count that treats every piece as first-time-right.
What Munshi surfaces
Teal is normal, maroon is flagged. Illustrative data.
Pieces drifting toward and past the tolerance band as a tool wears through a run.
Tagged at the machine, so the second pass lands against the tool or the setup.
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
Yield lost to nesting and rework, and consumables nobody tracks per plate.
The weld metres you logged, and the ones that quietly went back.
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.