Quality Management
Which defects keep coming back from the same supplier
Inspection lots, defects and notifications, traced back to the supplier, the batch and the process step that produced them.
See it answer one of theseAsk it
Questions you can ask on day one
Asked in plain language. Each answer prints the SAP entity, field and period it was computed from, underneath the figure.
- “Which defects keep coming back from the same supplier?”
- “What is our first-pass yield by plant?”
- “Which inspections are overdue?”
- “How has our rejection rate moved quarter on quarter?”
- “Which materials generate the most quality notifications?”
- “How long are notifications staying open, by type?”
What it reads
The quality objects behind an answer
Inspection lots
What was inspected, against which plan, with the usage decision that closed it.
Defects and characteristics
What failed and by how much: the detail that turns a pass rate into a cause.
Quality notifications
Complaints and issues raised, with their type, priority and how long they stayed open.
Batches
The link from a defect back to the material, the batch and the delivery that brought it in.
Supplier and material master
The attributes that let a quality figure be grouped by who supplied it rather than by a code.
Why it is hard
What no standard report does for you
Recurrence is a comparison across windows
A defect that appears once is noise. The same defect from the same supplier three quarters running is a pattern, and nothing flags it, because each occurrence is individually unremarkable.
A rate needs both halves, and they are different reads
First-pass yield is what passed over what was inspected. The two live at different grains, and a rate computed from the wrong denominator is a plausible number that moves in the right direction for the wrong reason.
The cause is in another module
Tracing a defect back to the supplier who shipped it means crossing into procurement. The link exists, but it is not on the inspection lot, which is why the question usually ends in a manual lookup.
Quality questions are pattern questions
A single inspection result is a fact, and your system shows it perfectly well. What costs money is the pattern: the same defect, from the same supplier, on the same material, three quarters running.
Nothing raises that, because each occurrence on its own is unremarkable. Seeing it requires grouping across a period and comparing against the period before: a query that has to be written, by somebody, deliberately.
Why rates need care
First-pass yield is what passed divided by what was inspected, and the two halves are different reads at different grains. Compute it from the wrong denominator and you get a number that is plausible, moves in the right direction, and is wrong.
AnrilX computes ratios in code and refuses one where the denominator is not known for the whole set. A rate over a partial result looks exactly like a rate over a complete one, which is why the refusal matters more than the calculation.
Crossing into procurement
Tracing a defect to its source means leaving the quality module: batch to delivery to purchase order to supplier. That chain exists in SAP, and following it by hand is a lookup somebody does one document at a time.
AnrilX reads it as a chain and names every entity it passed through, so an answer about a supplier’s quality is checkable rather than asserted.
One of these questions, answered
Which defects keep coming back from the same supplier?
Three supplier and defect pairs have recurred in every month since April. Tabuk Fabrication with defect code 0042, surface finish, is the clearest: 47 notifications, and first-pass yield on their deliveries is 8.2 points below the plant average.
Source: A_InspectionLot and A_QualityNotification · notifications grouped by Supplier and DefectCodeGroup · first-pass yield from lot usage decisions · monthly, 2026
Assumed: keeps coming back means present in at least four consecutive months, not merely frequent. A supplier with fifty defects in one bad month is a different problem and is not in this list.
Where it stops
What it will not answer here
The limits particular to this module. Worth knowing before a pilot rather than during one.
-
An inspection is only as granular as the plan behind it. Where characteristics are recorded as one overall verdict, the platform cannot break a failure down to the characteristic that caused it.
-
Recurrence is a pattern the platform can count. Calling it a supplier problem is a conclusion you draw, not one it reaches for you.
-
Where a link between a defect and its source is not modelled in your system, the platform reports the gap rather than inferring it.
Questions people ask us
Can it trace a defect back to a supplier?
Where your system links the batch to the delivery and the delivery to the purchase order, yes, and the answer names each entity it read on the way, so the chain can be checked.
Can it compare quality across periods?
Yes. The same read runs against two windows and the difference is computed in code, so the comparison is reproducible rather than reconstructed by hand each time.
What about a rate over a partial result?
It is refused. A percentage whose denominator covers only part of the set is one of the easiest wrong numbers to produce and one of the hardest to spot.
Questions often cross into
Purchasing, read at the grain the money sits on
Purchase orders and their items, goods movements, supplier invoices and info records, with spend sitting on the item, where it actually lives.
Which orders are exposed to something that has not arrived
Production orders, operations and component demand: where a schedule is exposed to something that has not arrived yet.
Equipment that costs more to keep than to replace
Equipment, work orders and downtime: the cost of keeping a machine set against what it would cost to stop.
Bring the question your reports cannot answer
Thirty minutes against a live SAP system we provide: no access to yours, nothing to set up. If it cannot answer, you find that out in half an hour rather than three months into a pilot.