Expected demand, forecasting, reordering rules, the master production schedule, net requirements, the supply plan and capacity in one system — so the numbers become a purchasing and production plan you can approve, instead of a reaction after a material runs out.
Direct answers
Manufacturing is execution: a work order, material consumption, an operation, time and cost. Planning is the decision before it: what to make and what to buy, in which quantity, by which date, and whether capacity allows it. They run on the same database, which is why we usually configure both — planning without execution stays a spreadsheet, and execution without planning stays a reaction.
Both, and that is the point: confirmed demand — sales orders, manufacturing orders, minimum stock — is combined with a forecast entered by hand or calculated from consumption history, and both appear in the net requirement with an option to exclude forecast already covered by real orders. We set which source is trusted per product family instead of applying one rule to the whole plant.
It subtracts what is available, reserved and already allocated to other documents from the requirement, then generates component requirements through the bill-of-materials levels, and proposes either a purchase, a manufacturing order or subcontracting for each shortfall — dated from the lead time or production time rather than from today. We configure the reordering rules, minimum order quantities and purchase packaging so the proposals are actually executable.
Partly, and we say so plainly: the system calculates planned loads on work centres and exposes the bottleneck and the late incoming work, which is enough for most small and mid-size plants. Full finite-capacity scheduling — rescheduling every operation on every machine against live capacity — is not automatic; it needs rules, configuration and accurate operation times. We do not claim perfect scheduling, and the design states what will be calculated manually and what automatically.
Anatomy
Eight elements built on your own data: the forecast supplies future demand, the rules supply the trigger point, and net requirements supply the decision.
A forecast calculated from consumption history or entered by hand from the sales plan, with a clear horizon and an accountable owner per product family.
Future demand with a known source
A minimum, an order quantity and a maximum per item and location, based on consumption rate and lead time rather than a remembered constant.
A calculated trigger point
A plan at product and family level over a time horizon, balancing demand against capacity before it becomes orders.
A decision before the order
Subtracts what is available, reserved and allocated, then walks down the bill-of-materials levels to generate component requirements with their dates.
A shortfall with quantity and date
Calculating planned loads on work centres and exposing which centre will choke first — before a delivery date is promised.
The bottleneck seen early
Lead time, minimum order quantity, purchase packaging and preferred vendor are part of the calculation, so a proposal is never a quantity nobody will sell.
An executable proposal
What is produced off-site is planned as a vendor with its own lead time and cost, so its effect appears in the plan rather than after the delay.
Subcontracting inside the plan
Comparing what was planned against what was actually consumed per family, to adjust the method with data instead of changing numbers by feel.
Learning from the gap
The planning cycle
Six steps from gathering demand to generated orders — with human review before any commitment, not blind automation.
Confirmed sales orders, existing manufacturing orders, minimum stock and expected orders — combined into one demand picture over a time horizon.
Output: one consolidated demand picture
Building the forecast from consumption history or the sales plan, naming who owns the number and stating the horizon.
Output: an approved forecast with an owner
Subtracting available, reserved and allocated stock from gross demand, then calculating component requirements down the bill-of-materials levels.
Output: a shortfall list with quantities and dates
For each shortfall: purchase, manufacture or subcontract — in a quantity that respects minimums and packaging, dated from the lead time.
Output: a proposed supply plan
Loading the plan onto work centres, exposing the bottleneck, then adjusting dates or quantities before committing to them.
Output: a capacity-adjusted plan
Human approval of the plan, then generating the purchase and manufacturing orders, with a weekly review of plan against actual.
Output: generated orders and a weekly review
Inside the system
The expected set: the planning overview, demand forecasting, the master production schedule, net requirements, capacity load, and the supply plan.
The images shown now are temporary placeholders that mark each slot and what it should contain. They are replaced with real screens from the dashboard (/admin/mrp-screenshots) with no code change.
Controls
Controls that make the plan explainable and reviewable: who changed it, why the system proposed that number, and whether the plan was accurate.
Why did the system propose that quantity? The demand that triggered it and the rule that decided it are both visible — no number is accepted without an explanation.
Who may change the forecast and who approves the plan is set by role and permission, so the whole stock position does not move on an unrecorded personal edit.
Who changed a number, when and why — which is what makes measuring planning accuracy possible instead of blaming the numbers.
Lead time, minimum quantity and purchase packaging are part of the proposal, so a tidy plan nobody can supply never reaches the buyer.
Planned against actual per family, so the method is corrected with data rather than by changing numbers.
These are operating controls we configure in the system — not a certificate or an accreditation from any authority. Forecast accuracy depends on the quality of your history and data, and approving the stock policy and the plan is the client’s decision.
Reporting
What purchasing, production and management need: demand against supply, shortages, the proposed plan, capacity, and forecast accuracy.
A time-phased view of where demand exceeds supply and where it does not, before it shows up as a sudden shortage.
What has reached its minimum and what will run short in the coming weeks given current demand.
What to buy, in which quantity, from which vendor and by which date — ready for review and approval.
Planned against available hours per centre, to find the bottleneck before promising a date.
Planned against actual by family and period, to adjust the method on evidence rather than impression.
Items whose lead time exceeds the planning horizon — the ones that need an early decision rather than daily chasing.
Integration
One plan becomes a purchase order for the vendor, a manufacturing order for the line, a stock movement and a journal entry — from the same number.
A purchase proposal becomes an order with the right vendor, lead time and quantity, not a list copied by hand.
The makeable side of the plan becomes manufacturing orders dated from capacity, not only from demand.
Available, reserved and allocated stock are read from inventory in real time, so the calculation never rests on an old count.
Confirmed sales orders enter demand immediately, and delivery promises rest on a plan rather than on hope.
Purchase commitments and the inventory value the plan creates appear in accounting, so finance sees the plan’s effect before it is executed.
Scope
Described scope with no published prices: cost depends on the number of items, families, BoM levels and integrations, and is quoted in a proposal.
Reordering rules for every moving item, net requirements and purchase proposals, and the shortage and below-minimum reports.
The right start to control shortages
Everything in Foundation plus demand forecasting, the master production schedule, multi-level net requirements, and capacity and bottleneck analysis.
The common single-plant scope
Everything above with several sites and warehouses, planning across them, product families with different rules, and additional customisation.
Quoted after the needs analysis
We publish no price before the needs analysis: cost follows item count, product families, BoM levels and integrations.
FAQ
Direct answers on scope, responsibility and what the system does not do — without inflation.
Manufacturing is execution: a work order, material consumption, an operation, time and cost. Planning is the decision before it: what to make and what to buy, in which quantity, by which date, and whether capacity allows it. They run on the same database, which is why we usually configure both — planning without execution stays a spreadsheet, and execution without planning stays a reaction.
Both, and that is the point: confirmed demand — sales orders, manufacturing orders, minimum stock — is combined with a forecast entered by hand or calculated from consumption history, and both appear in the net requirement with an option to exclude forecast already covered by real orders. We set which source is trusted per product family instead of applying one rule to the whole plant.
It subtracts what is available, reserved and already allocated to other documents from the requirement, then generates component requirements through the bill-of-materials levels, and proposes either a purchase, a manufacturing order or subcontracting for each shortfall — dated from the lead time or production time rather than from today. We configure the reordering rules, minimum order quantities and purchase packaging so the proposals are actually executable.
Partly, and we say so plainly: the system calculates planned loads on work centres and exposes the bottleneck and the late incoming work, which is enough for most small and mid-size plants. Full finite-capacity scheduling — rescheduling every operation on every machine against live capacity — is not automatic; it needs rules, configuration and accurate operation times. We do not claim perfect scheduling, and the design states what will be calculated manually and what automatically.
No. The system generates proposals and draft orders and needs human approval before they become a commitment for the business. Approval can be scoped by value or item type, but sending a purchase order to a vendor unreviewed is not the default — and we do not configure it that way unless management explicitly decides to.
Cleaning consumption history when it is noisy or incomplete, deciding the stock policy on management’s behalf (what we make to stock and what we make to order), supply-chain consulting or vendor negotiation, and configuring unrealistic lead times. We configure the system and train the team, but the commercial decision and the stock policy belong to the client.
We start with reordering rules on moving items, because they need a recent consumption rate rather than a long history, and enter the forecast by hand from the sales plan or existing contracts. After about three months of running, there is real history to build a calculated forecast on, and we move over gradually instead of waiting a full year before starting.
Four details are enough: item count, whether you have usable consumption history, number of production lines or work centres, and your current planning method.
Tell us your item count, product families and production lines, and we will come back with a delivery plan and a clear scope.
Request a proposalWe are a systems implementation and integration company, not a certification body. We configure the planning, forecasting and requirement rules and test them on your data; forecast accuracy depends on the quality of your history, and the stock policy and the plan are the client’s responsibility and decision. The system generates proposals and draft orders rather than automatically approved purchase orders, and we announce no accuracy or service-level percentage before measuring it on your data.