Consume From Stock: How EDGEBIC Nets Demand Against Inventory

Inventory netting in production scheduling is the decision, made per order and in sequence, of whether demand is covered by stock on the shelf or must be built on the shop floor. Get it wrong in one direction and you build parts you already own; get it wrong in the other and you promise stock to two customers at once. EDGEBIC by User Solutions makes this decision inside the scheduling run itself, decrementing availability as each order consumes it and posting every consumption to an auditable inventory ledger. This walkthrough traces three orders for one stocked product through the whole decision, with every quantity and timestamp shown.

If scheduling terminology like routing and work center is new territory, start with what production scheduling is and come back; this post goes straight to the numbers.

The Cast: One Product, 80 Units, Three Orders

The product is a Hydraulic Seal Kit, a stocked item with 80 units on hand. Its build policy is make-to-order, meaning orders for it normally get scheduled onto the floor, but individual orders can be flagged build-to-inventory to act as stock producers.

Three manufacturing orders arrive in the same scheduling run:

The producer's routing is one assembly step: 1 hour of setup plus 0.5 hours per unit, so 50 units cost 26 hours of shop time. The consumers have the same routing available, but whether they ever touch it is exactly what netting decides.

The inventory ledger opens with a single entry: opening balance, plus 80, running balance 80.

Producers Schedule First, and Never Consume

When a scheduling run contains a build-to-inventory producer, EDGEBIC sorts producers ahead of consumers so supply is on the timeline before demand asks for it. Job-001 goes first.

A producer never consumes stock, by rule; it exists to create it. So Job-001 skips the netting check entirely and schedules as ordinary shop work: 26 hours on Assembly-1 across four days, finishing Thursday June 19 at 10:00. That completion matters later, because with forward netting the engine registers the 50 finished units as a future receipt on the product's availability timeline, dated to the moment the producer completes. Stock on the shelf is available to anyone at any date; the producer's 50 units are available only to consumers whose need date falls at or after Thursday 10:00.

Job-002: Satisfied From Stock

Job-002 needs 60 units on Wednesday June 18 at 08:00. Before expanding its routing, the engine runs the netting check: how many units are available as of Wednesday 08:00?

The shelf holds 80. The producer's 50 do not count, because they arrive Thursday, after this need date. Available is 80, need is 60, and 80 covers 60, so Job-002 is satisfied from stock:

• No shop-floor work is scheduled. The engine emits a single instant fulfillment row, work center "Inventory", zero hours, start and end both Wednesday 08:00. On the Gantt it appears as a point with a stock icon rather than an operation bar.
• Availability decrements immediately. 80 minus 60 leaves 20, and that 20 is what every later order in the run will see. Two orders can never spend the same units.
• The order's status changes to Satisfied From Stock, and the order list shows an inventory badge instead of a schedule of operations.

After the run persists, the consumption becomes a ledger fact: an issue of minus 60 posts against Job-002 with the comment "Schedule Transaction", the running balance moves to 20, and the product's on-hand count updates to match. Sixty units of demand were met with zero machine hours, and there is a row in the ledger that says exactly when, for whom, and why.

Job-003: Twenty Is Not Thirty

Job-003 needs 30 units on Thursday June 19 at 08:00. The netting check runs again, against the decremented state: 20 units remain on the shelf. The producer's 50 arrive at 10:00 Thursday, two hours after this order's need date, so they still do not count.

Available is 20, need is 30, and 20 falls short. EDGEBIC does not partially net in this decision: the order fails the full-coverage check and is scheduled onto the shop floor for its whole quantity, 30 units of normal assembly work, queued behind the producer on Assembly-1 and starting once that work center frees up. Its status is Scheduled, its ledger footprint is zero, and its diagnostic record notes that 50 units of future receipts existed but arrived too late to use.

That two-hour miss is the sharpest lesson in this walkthrough. Netting is not just about quantity; it is about quantity at a date. Which leads to the most useful what-if in the whole scenario.

Move the Need Date Two Hours, Change the Answer

Suppose Job-003's need date were Thursday at 11:00 instead of 08:00. Now the availability check as of 11:00 sees both supplies: 20 remaining on the shelf plus the producer's 50 landing at 10:00, for 70 available against a need of 30. Job-003 would be satisfied from stock too, drawing on units that did not exist when the run started, and a second issue would post to the ledger.

This is what time-phased forward netting buys you: the scheduler can plan a consumer against a producer's output in the same run, as long as the dates genuinely line up. A simple point-in-time snapshot of on-hand stock can never see that receipt; it either overpromises stock that is not there yet or forces a build the shelf will shortly make unnecessary. EDGEBIC's netting reads supply as a timeline, which is the same reasoning a good planner does on paper, executed consistently across every order in the run. This is netting as a planning capability inside the scheduler; it is the scheduling half of the make-to-stock story, and it works with however demand reaches EDGEBIC, whether entered directly or brought in from your ERP through Excel, CSV, or database import masks.

The Ledger Keeps the Whole Story

Everything netting does lands in the inventory ledger as append-only transactions:

Two properties are worth calling out for anyone who has fought phantom inventory:

Reschedules reverse, they never delete. If you reschedule after Job-002 consumed its 60, the engine first posts an offsetting reversal of plus 60, restoring availability to 80, then re-runs the netting decision from scratch. Job-002 qualifies again and a fresh issue posts. The ledger ends with three rows instead of one, and that is the point: every consumption and every un-consumption is a dated, order-linked fact you can audit months later.

On-hand always reconciles to the ledger. The product's on-hand number is a cache of the ledger's running balance, and a built-in consistency check flags any drift between them. When the count says 20, you can trace precisely which orders took the other 60.

What You See in EDGEBIC

The final state of the three orders, as the planner sees it:

In the order list, Job-002 wears an inventory badge with the tooltip "Satisfied from stock, 60 units." The product's transaction history shows the opening balance and the issue, color-coded. The on-hand tile reads 20. And the scheduling diagnostic log carries one line per netting decision, recording need, availability, quantity used, and quantity still to build, so the reasoning behind every yes and every no is written down, not folded silently into the plan.

Compare that with what most shops actually do: someone checks the stock screen in the morning, remembers there were 80, and two order-entry clerks promise the same units to different customers by lunchtime. Netting inside the scheduler removes the race, because the decrement and the decision happen in one place, in one order, every run. Manufacturers running the predecessor products from User Solutions have leaned on this kind of single-source discipline for decades; Cummins coordinates planning across 33 locations on the strength of it, and EDGEBIC carries the approach forward with the netting arithmetic built into the engine.

Three Rules to Take Away

1. Producers first, and producers never consume. Build-to-inventory orders are sorted ahead of consumers so their output is on the timeline before demand nets against it.
2. Availability is checked at the need date, and decremented in run order. 80 covered Job-002; the remaining 20 could not cover Job-003; and a two-hour shift in a need date can flip the answer.
3. Every consumption is a ledger transaction, reversible and order-linked. No deletes, no phantom stock, no arguing about where the units went.

This walkthrough is part of the EDGEBIC worked examples series; if you are setting up products, routings, and work centers for the first time, the greenfield setup walkthrough builds the master data these runs depend on. To see the full platform picture, start at EDGEBIC.

Have a stocked product and a stack of orders of your own? Bring your on-hand counts and open orders to a demo and we will run the netting decision live on your numbers.