Technical Lab · Synthetic demonstration

Forecast to inventory

The interactive demo lets you change the demand pattern and the model. This page holds those fixed and walks the decision from forecast to order.

All technical examples

Synthetic demonstration · Independent educational implementation. Same calculation as the interactive demo. Not a business result.

Turn a multi-week forecast and a coverage target into a quantity someone could order.

The forecast is not the decision. Stock already on hand, stock already coming, the lead time and the pack size all sit between the two.

Order-up-to level S = expected demand over (lead time + review) + safety stock

In a continuous-review policy the reorder point has the same shape, without the review interval. This demonstration uses the demo’s periodic-review level. The recommended order is the gap to S, rounded up to the pack size, and zero when stock already covers S.

Change the protection period

The order is the gap between the order-up-to level and the inventory position, rounded up to whole packs. Expected demand over the protection period plus the safety stock sets S. White ticks inside the order mark each pack.

Forecast: 63.7, 66.7, 70.4 units per week
Lead-time demand: 200.8
Safety stock: 5.3
Order-up-to level: 206.2
Inventory position: 100 on hand and on order
Recommended order: 110

What happens to stock

Nothing ordered today arrives before the lead time has passed, so the stock path until then is already fixed. Expected stock through the protection period. The shaded band is the lead time and the dotted line is when the new order arrives. Zero is the stockout line on this expected path, not a service guarantee.

The protection period is 3 weeks: lead time 2 plus a 1-week review. Expected demand over that period is 200.8. Safety stock, the gap between the 95% coverage target and that expectation, is 5.3. The order-up-to level is 206.2. Inventory position is 100. The gap is 106.2, which rounds up to an order of 110 in packs of 10. Expected stock goes negative in week 2, before the new order arrives. Ordering more today does not bring that delivery forward.

Open the interactive Forecast → Inventory demo to change the demand pattern, the model and the stock position.

About this example

Method: The periodic-review order-up-to rule used by the Forecast → Inventory demo. Safety stock is the gap between expected demand and a demand-coverage target, not a normal zσ formula.
What the code does: This page calls that demo’s decision function. It does not reimplement it. Lead time and the coverage target are the only controls.
Assumptions: Smooth synthetic demand, the demo’s recommended model, 80 units on hand, 20 units due at the start of week 2, a review every week and packs of 10. There is no expediting.
Result: A longer lead time raises the order-up-to level and the recommended order. It does not move the arrival earlier. If expected stock runs out in week 2, today’s larger order cannot prevent that.
Limitations: Coverage is a planning target, not a measured fill rate. The numbers are the demo’s synthetic series. They are not a Priory Direct order.

Implementation

The decision itself is the demo module, not a second copy.

src/lib/forecast-inventory/decision.ts View source on GitHub: src/lib/forecast-inventory/decision.ts

Protection period, coverage target, inventory position and pack rounding.src/lib/forecast-inventory/decision.ts

const protectionPeriod = input.leadTime + REVIEW_WEEKS;
  const weekly = model.pathMeans.slice(0, protectionPeriod);
  const expectedDemand = weekly.reduce((sum, value) => sum + value, 0);
  const target = model.targets[String(input.coverage)][protectionPeriod - 1];
  const position = inventoryPosition(input.onHand, input.onOrder);
  const requirement = Math.max(0, target - position);
  const quantity = roundToPack(requirement);