Infinite capacity scheduling assumes every machine can take whatever arrives. Finite capacity scheduling respects what each one can really do. Both break on a real floor, for opposite reasons. The approach that works schedules one point to its real capacity and protects everything else with buffers.
Infinite capacity scheduling
Most ERP and MRP systems plan this way by default. Each order is planned back from its due date using standard lead times, and every resource is assumed to have room for it. A capacity report may show overloads afterwards, but the dates themselves don’t move.
The result is familiar: the plan says every order is on time, the floor knows it isn’t, and priorities get set by whoever shouts loudest. More on that in why the ERP schedule gets ignored.
Finite capacity scheduling
Finite capacity scheduling, also called finite loading, fills each resource only up to the hours it has. When a machine is full, the next job moves later. Dates become realistic, at least on paper.
The catch is detail. Scheduling every operation on every machine to the minute needs accurate data everywhere, and real floors don’t stand still: a breakdown, a rework loop or late material changes the sequence within hours. A fully finite schedule is often out of date by mid-morning, and rerunning it reshuffles everything.
Side by side
| Infinite | Finite (every resource) | Finite at one point | |
|---|---|---|---|
| How it works | Plans back from due dates with standard lead times. Load is not checked. | Loads each resource up to its available hours; jobs move later when it is full. | Finite at one pacing point; everything else planned back from it with time buffers. |
| Due dates | Look fine on paper, then slip on the floor. | Realistic, if every input is right. | Realistic: set from the pacing point’s real capacity. |
| Data needed | Routings and standard lead times. | Accurate times for every operation on every resource. | Good data for the pacing point; rough times elsewhere. |
| When things go wrong | Overloads show up as late orders and expediting. | The detailed schedule breaks and needs rerunning. | Buffers absorb normal problems; the daily buffer meeting handles the rest. |
| Typical tool | ERP / MRP default. | Advanced planning and scheduling (APS) software. | Scheduling board or spreadsheet; works alongside the ERP. |
The middle way: finite at one point, buffers everywhere else
A plant’s output is set by very few points. Schedule the one the line is paced from (the pacemaker, or in Theory of Constraints terms the drum) to its real capacity. Plan every other step back from it, with a time buffer in front of it and before shipping. Release work only at the pace that point can take.
- Map the line and decide where to pace it, from a value stream map and the plant’s own data.
- Schedule that point finitely: due date first, then group shared setups where no due date is put at risk.
- Plan the rest back from it with time buffers instead of detailed timings.
- Release work at its pace, and only when the kit is complete.
- Run a daily buffer meeting: green, amber, red; every red gets an owner and a date.
This is the logic of drum-buffer-rope, and it fits on one sheet: see production scheduling in Excel or get the free scheduling board template.
Want it set up on your floor?
We set this up on your own data and the business system you already run, and coach your team until they run it. See our production scheduling consultants and Theory of Constraints implementation.
FAQ
What is infinite capacity scheduling?
Scheduling that works back from each due date using standard lead times and assumes every resource can take whatever load arrives. It is how most ERP and MRP systems plan by default.
What is finite capacity scheduling?
Scheduling that respects what each resource can actually do in the time available. When a resource is full, the next job moves later, so the schedule shows realistic start and finish dates.
What is finite loading?
Another name for finite capacity scheduling: loading each resource only up to its available hours, rather than piling on load and dealing with overloads afterwards.
Is finite capacity scheduling always better?
Not when it is applied to every machine to the minute. On a real floor, breakdowns, rework and late material change the plan within hours, and a detailed finite schedule for every resource is out of date almost as soon as it is published.
Do we need APS software for finite scheduling?
Not to start. Finite scheduling at one pacing point, with buffers elsewhere, fits in a spreadsheet for one plant. Our free Excel template does exactly that.
How does drum-buffer-rope relate to finite scheduling?
Drum-buffer-rope is finite where it matters and infinite elsewhere: the drum is scheduled to its real capacity, and every other step is planned back from the drum with time buffers.
Take the free scorecard — ten questions, instant result.
Take the scorecard