MRP, the Pull Boundary and Master Data
In one sentence
MRP makes future material need visible, while pull starts shop-floor replenishment from real consumption; each boundary has a single authorized order source.
By the end of this lesson
- Separate a part's long-term purchasing plan from shop-floor replenishment.
- Match demand, order, stock point and consumption record in a single relationship.
- Calculate the surplus created by a double order with the stock balance.
- Determine the effect of errors in the bill of materials, container quantity and lead time.
Why it matters
MRP and pull can serve the same material flow over different time spans. Purchasing with a long lead time wants to see future need from today. Repeated shop-floor replenishment, on the other hand, meets the gap that real consumption opens and prevents the intermediate stock from growing unnecessarily. It is not a problem for these two functions to exist together. The problem is when it stays unclear which signal gives production or transport authority at which boundary. If both a planning order and a pull order run independently to meet the same consumption, the material is prepared twice. Two well-calculated local plans together can create wrong system behavior.
The pull boundary shows where the material is held and which event starts the next movement. A supermarket is an intermediate stock point managed with defined upper and lower limits. A kanban signal can give transport or production authority following real withdrawal from this point; the type of the signal must be explained. A production order and a transport authority are not equal. Moving a part that was made earlier from one place to another does not mean the same quantity of new parts has been produced. When the calculation and the record preserve this physical distinction, the roles of MRP and pull become clear.
Master data are the permanent definitions the orders rest on: the bill of materials, part identity, unit of measure, container quantity, resource, lead time and stock policy are all included. If these records are wrong, the system produces the wrong quantity by the right rule. Tracking the same part as two separate identities, by package and by unit, can multiply the need. Not knowing that a part is managed both by pull and by an independent MRP release makes a double order routine. Master data are not only an IT preparation; they are production, planning and logistics describing the same physical flow with the same numbers. The person responsible for changing the records and their effective date must also be defined. When a new container definition takes effect, old containers may still be in circulation. In this transition it is not assumed that all containers carry the same quantity; old and new fill are clearly distinguished. Treating partially filled containers as empty can also produce a signal before real consumption. The partial-container behavior is therefore defined with a rule that is understandable on the shop floor. A correct definition is not enough; the legibility of the label, the material going to the right place and the record of the closed signal are also verified. That way, when the source of a stock difference is investigated, a bill-of-materials error is not confused with a handling practice.
The boundary decision is not a one-time label. A new product, a changed container, a longer supply time or a discontinued consumption policy can affect the policy. In such a change, open orders and signals in circulation are checked together. If new cards are added without removing the old ones, the defined stock limit grows. When the forecast rises, instead of opening a second release path at the pull boundary, the long-term supply plan and the replenishment capability are reviewed. With every change it must be possible to trace which physical operation meets the single demand, and future visibility must not be mixed with today's production authority. For the numerical scenarios in the lesson quizzes as well: This dataset was generated synthetically for training; it does not represent a real business, employee or customer study.
Formula
Long-term component requirement (units) = Planned product (units) × Usage (units/unit)
Number of replenishment signals (signals) = Quantity consumed (units) ÷ Container quantity (units/signal), for full-container consumption
Period-end stock (units) = Opening stock + Actual replenishment − Consumption
Double-order surplus (units) = Quantity made a second time for the same demand
Usage error (units) = Product quantity × Difference between actual and recorded usage
Unlock the rest of this lesson with an Academy account
Create a free Academy account with your work email address; it can be used once it is approved.
- Step-by-step instructions
- Worked example with real numbers
- Where the method misleads
- Common mistakes