Why finite capacity changes everything
A supply plan without a capacity constraint is just a wishlist. The finite capacity engine takes that infinite plan, constrains it period by period and propagates the decision through the entire chain: production, intermediates, purchases and transfers.
When capacity does not meet everyone, someone leaves the queue. The question is not whether prioritization will happen, but whether it will be driven by business criteria or by the loudest shout of the day.
Constraints beyond the machine
Finite capacity is not only machine hours. In real factories, the constraint that breaks the plan is usually somewhere else.
Machine capacity
Classic constraint: available hours per workcenter, considering shifts, efficiency and technical availability.
Tools and molds
The same mold may fit several machines but run on only one at a time. Without tool control, the plan becomes fiction.
Multi-skilled labour
An operator skilled for 3 stations is a shared constraint. Multi-skilling has to enter the capacity model.
Storage and WIP
Buffer space between operations and finished-good shelves also cap factory throughput.
Setups in the master plan
Even without sequencing, setup time must be accounted for. Ignoring setup creates optimistic plans the floor cannot meet.
Preventive and corrective maintenance
Scheduled PMs and average corrective time consume real capacity. They must show up in the calendar.
Dynamic priority
If everyone does not fit, someone has to come first. A dynamic priority queue adapts the scenario to the challenges and priorities of the moment, combining business criteria into a single score.
Typical queue criteria
Calendar and maintenance
Real capacity only becomes visible when the calendar accounts for shifts, holidays, technical stops and maintenance plans (preventive and corrective). Without it, any load calculation silently overflows.
Simulating alternative calendars (extra weekend shift, pulling PM in, reshuffling holidays) is a powerful scenario analysis tool. Each calendar becomes a plan copy to compare cost, lead time and risk side by side.
Order synchronization
A plan without orders is theory. Order generation and synchronization turn the engine output into executable production, purchase and transfer orders, respecting batch, multiple and consolidation rules.
Multi Level Sync
Finished-good demand propagates to intermediates, packaging and raw materials in a single coherent pass with no time gaps.
Pegging
Every production, purchase and transfer order is linked to the original demand. Full visibility from customer to RM.
Order Generation
Automatic order generation respecting min batch, multiple, component consolidation and rounding with no leftovers.
Shelf life and expiry
Pharma, food and cosmetics require expiry as a constraint: old lots cannot cover future demand.
Multi-level pegging
Pegging is the mesh that links every order back to the original demand. When the customer order changes, the system knows which production, intermediate and purchase orders are connected. Without this mesh, any change becomes a full recompute and loss of visibility.
A pegging logic guarantees correct net requirements calculation, gives traceability from customer to RM, and shows the real financial impact of each planning decision.
Consolidation and shelf life
In pharma, food and cosmetics, several presentations share the same intermediate (same bulk, different packaging). Order generation must consolidate presentation demand and round it to generate a formulation multiple lot, with no leftover discardable bulk.
These same sectors require shelf-life control: lots close to expiration should not cover distant future demand. The engine must treat expiry as a constraint, not a cosmetic attribute.
How we solve it with NPLAN
The NPLAN Capacity and Orders modules combine the finite capacity engine, the dynamic priority queue, calendar simulation and synchronized order generation with multi-level pegging.
Finite capacity engine
Infinite plan is constrained period by period: needs above capacity are pulled in, pushed out, or allocated to an alternative resource.
Tools, labour and storage
Secondary constraints enter the same engine: shared molds, multi-skill per shift and buffer capacity between operations.
Dynamic priority queue
When capacity is short, a queue weighted by margin, revenue, ABC, coverage and manual priority decides who runs first.
Calendar simulation
Extra-shift, planned downtime, holiday and PM scenarios become plan copies to compare cost, lead time and risk before committing.
End-to-end pegging
Customer, sales order, production order, intermediates and purchase orders are chained. A sales change shows impact down to RM.
Lot consolidation and multiples
Several presentations of the same component are consolidated and rounded to generate an intermediate multiple lot with no leftovers.
Test your Knowledge
When capacity is short for all demand, which approach is best?
Next in the series: Distribution Planning
With production synchronized, the next challenge is distributing the right product, in the right quantity, to the right consumption point while respecting logistics capacity.
Read the next article