What Is an Inventory Integrity Check in Planning?

An inventory integrity check is a plant-wide audit of the inventory side of the planning data, looking for conditions that make the numbers untrustworthy. It compares each product's on-hand figure against the sum of its ledger entries, finds balances that have gone negative, flags make-to-stock products that are not marked stocked, validates the yield value's range, confirms that a job recorded as satisfied from stock actually drew that stock in the ledger, and finds released master schedule rows whose manufacturing order no longer exists.

The core idea is the same one that governs any set of books. The ledger of receipts and issues is what happened. Everything else, including the on-hand number every screen displays, is a summary computed from it. When a summary and its ledger disagree, the summary is the one that is wrong, and every decision built on it inherits the error.

This entry belongs to the EDGEBIC by User Solutions glossary. For the wider vocabulary, see the manufacturing glossary, and for the parent concept, the definition of a scheduling anomaly check.

How the Inventory Integrity Check Works

On-hand against the ledger

Every product carries an on-hand quantity for speed, and a ledger of receipts and issues that is the authoritative record. After each posting the on-hand figure is recomputed from the ledger. The check sums the ledger independently and compares. A gap is critical, because the fast number that drives netting, projections, and replenishment has drifted from the record that produced it.

Negative balances

Physical stock cannot go below zero, so a ledger summing negative for a stocked product means more has been issued than was ever received. This is a warning rather than a critical, because the arithmetic is legal even though the physical situation is not. It usually points at issues posted against receipts that never landed.

Make to stock without the stocked flag

A product's build method says whether it is made to order or made for the shelf. A separate flag says whether it holds inventory at all. With that flag off, netting, replenishment, and receipt and issue posting are all skipped. A product set to build for stock but not flagged as stocked therefore builds and never arrives, and nothing downstream can net against it. The check flags the contradiction.

Yield range

A product's yield is a fraction, so it must be greater than zero and no more than one. A value of zero or below is critical because it is used as a divisor. A value above one is a warning, because the field has been filled with a percentage or a count rather than a fraction.

Satisfied from stock with no draw-down

When a job's demand is met from stock rather than by building, the schedule records it as satisfied from stock and an issue is posted to the ledger. The check confirms both halves exist. A job claiming stock was consumed with no matching issue entry means the plan and the ledger disagree about whether anything left the shelf.

Firmed-forever orphans

A master production schedule row released into a manufacturing order carries a link to that order. If the order is later canceled or removed, the row still reads as released. The planner believes the quantity is covered; nothing is building it. The check finds those orphans so the row can be reset and released again.

Why These Are Plant-Wide

Every one of these tests audits master data or a ledger rather than a single plan. There is no job to open. Running them per job would be meaningless, so they scan the whole product list and the whole ledger, which is also why the natural cadence is periodic rather than every publish: run them after a data migration, after a cleanup, and on a regular rhythm.

A Concrete Example

A shop keeps 400 brackets in finished goods and a make-to-stock policy on the part.

A customer order for 250 arrives. The plan satisfies it from stock, so no machine time is scheduled and an issue for 250 is posted. On-hand drops to 150. Both stores agree.

Two weeks later the planner runs the inventory integrity check as part of a monthly audit. Two findings come back. The first is a warning on a different part: it is set to build for stock but is not flagged as stocked, so three completed builds never posted receipts and the shelf still reads zero. The planner turns on the stocked flag and reposts.

The second is critical: a third part's on-hand reads 340 while its ledger sums to 285. The planner audits that part's ledger, finds a reschedule that reposted issues without the on-hand being recomputed, and lets a posting cycle rebuild the total from the ledger. The 340 becomes 285, and the replenishment suggestion that had been suppressed by the inflated figure appears as it should have two weeks earlier.

How EDGEBIC Reports Inventory Integrity

In EDGEBIC, the inventory integrity tests run inside the Scheduler Anomalies report under the Reports menu. Because they are plant-wide, run a full scan rather than a single-job one. Severity does the triage: the on-hand drift, the bad yield value, and the malformed satisfied-from-stock row are critical; the negative balance, the unstocked make-to-stock product, the missing draw-down, and the firmed-forever orphan are warnings.

The neighboring definitions are the inventory ledger that holds the authoritative record, consume from stock netting which produces the satisfied-from-stock rows, and the is-stocked flag that governs whether a product participates at all. For the practical walkthrough, see how to audit the inventory ledger for a part, and for the symptom-first version, inventory on-hand looks wrong after a reschedule.

A planning system is only as honest as its ledger. These checks are how you find out whether yours still is.