Cyber Essentials certified ISO 27001 aligned Runs in a browser Unlimited shop-floor users Built in Sheffield

Resource

What MRP actually does, and what it does not

What material requirements planning calculates, what it needs from your data, and where implementations usually struggle.

This is a first draft awaiting review by somebody who runs a factory. It is published as a working document rather than as advice.

Material requirements planning answers two questions: what has to be made, and what has to be bought, in time for the orders you have accepted. It does that by taking demand, subtracting what you already have or have already ordered, and exploding what is left through the bills of material until it reaches things you buy.

That is the whole idea. Everything else is detail about how well it is done.

What it is not

MRP is not stock control. Stock control tells you what is on the rack now. MRP tells you what will be short in five weeks given the orders you have taken, which is a different and more useful question.

It is not scheduling either. MRP works out what and when; finite scheduling works out whether there is capacity to do it. A system can be good at one and poor at the other, and many are.

And it is not ERP. ERP is a category that usually includes MRP alongside finance, sales and sometimes payroll. A manufacturer can need MRP without needing the rest.

What it needs from you

MRP is unforgiving about data, and this is where implementations struggle rather than in the software.

  • Bills of material that reflect what is actually built, including the things nobody bothered to add.
  • Routings with operation times that are roughly right. Wrong is workable; missing is not.
  • Lead times that reflect what suppliers actually do, not what they said once.
  • Stock figures somebody believes. If the ledger and the rack disagree, MRP amplifies the disagreement.
  • Scrap and yield where they are material, otherwise every run under-orders.

Time-phasing, and why it matters

A naive calculation nets demand against supply as two totals. That will tell you that you have enough, right up to the week you do not.

Time-phased netting puts both on a timeline. Demand in week six is netted against stock now plus a purchase order landing in week four, not against a purchase order landing in week nine. The shortage appears in the week it happens rather than as an average across the horizon.

Pegging

A recommendation is only actionable if you can see what caused it. Pegging keeps the link between a suggested purchase and the demand that generated it.

Without pegging, MRP hands you a list and you take it on trust. With it, when a customer pushes an order back three weeks you can see which purchase suggestions should move with it.

Where to start

Pick your most repeated part and check its bill of material against what is actually built. Then check its routing against how it is actually made. Most MRP problems visible after go-live were present in that data before it.

What this does not cover

This is a general explanation, not advice about your business. Whether MRP is the right next step depends on your order pattern, how much you carry in stock and how good your part data is.

Material requirements planning in Keystone · Moving from spreadsheets to MRP

Next step

Show us how your factory runs

A walkthrough starts with your orders, your production flow and your material problems. We follow one of your jobs through the system rather than presenting a feature list.

Book a walkthroughSee pricing