Finite capacity scheduling fits work into the available time of the resources required to run it. If a machine has eight staffed hours and two operations require nine, at least one operation must move, find another eligible resource, or get more capacity. Put that choice on the schedule before giving the dates to the plant.
Microsoft’s finite capacity planning documentation distinguishes finite capacity from a plan that can load more work than a resource can perform. A finite schedule still depends on the time assumptions and constraints you enter. Verify the calendar and routing behind a slot before calling it feasible.
Measure the hours the operation can use
Start at the constrained work center, such as a shared oven or inspection station. Build its calendar from staffed shifts, planned maintenance, and operating restrictions. If an operation also needs a certified technician, intersect the two calendars. Eight available machine hours do not give a job eight feasible hours when that technician is free for only six.
Record setup and run time separately. An oven may take an hour to change temperature between products, then process a batch at a fixed cycle length. A rate per unit alone cannot explain the first changeover or the batch boundary. Preserve the incoming machine state or previous product when the next setup depends on it.
Confirm that material and the prior operation will finish before the scheduled start. Use the component’s usable time, including receiving or inspection when required. Manufacturing capacity planning can reveal aggregate overload before you place individual jobs. The detailed schedule still needs a slot that respects the actual resource calendars.
When comparing a schedule with a capacity chart, account for nonproductive time once. If maintenance already removes a half hour from the oven’s availability, do not also charge that same half hour to the scheduled setup. If the actual oven cycle includes normal load and unload time, keep that time in the run standard where the planner can see it.
Compare the effect of two job sequences
Assume one oven is staffed from 8 a.m. to 4 p.m., with no other reservations. Job A needs four hours and is due at noon. Job B needs three hours and is due at 4 p.m. The oven is already set for A. Switching from A to B takes one hour, while setting up for B first and then changing back to A takes two hours total.
With A first, run A from 8 a.m. to noon, switch over until 1 p.m., then run B until 4 p.m. Both jobs fit. With B first, its initial setup and run take four hours. A starts after a second setup and ends at 5 p.m. under a hypothetical extended shift. In the normal eight-hour shift, that sequence leaves an hour of work outside available capacity and misses A’s noon due time.
If A’s material is unavailable until 10 a.m., neither original sequence meets both deadlines within the shift. The planner can check another approved oven, seek a staffed extension, or ask which date may move. Make the missing material visible on the job record so sales and production can discuss the same constraint.
| Sequence | Result under stated assumptions | Decision |
|---|---|---|
| A then B | A ends noon, B ends 4 p.m. | Fits staffed hours and both due times |
| B then A | A would end 5 p.m. with an extension | Normal shift lacks one hour and A misses noon |
| A materials available at 10 a.m. | A needs four hours from 10 a.m. at earliest | Escalate a date or capacity decision |
Inspect planned and actual time for one work center
Bring in a routing and shift export and check how jobs and observed durations join.
Protect near-term work and update explicit exceptions
Choose a short window in which changes need a supervisor’s approval. Otherwise every new urgent order can reorder work whose material or tooling is already staged. Keep the prior schedule version and the reason for moving a job so you can check whether the schedule changed or the plant missed the plan.
For one review, move a material receipt later and recalculate. Check that the job and its downstream operations move, the alternate resource is approved for that routing, and a downstream due date still reflects shipping time. Then return the receipt to its original date and verify that the scheduler does not move unrelated work across the protected window.
Review planned against observed start, completion, and changeover time before changing the standards. Use production schedule adherence with the exact version crews were given. A later schedule that already moved the job cannot show how the earlier plan performed.
An alternate machine only helps if the job can use it. Have quality approve a routing with the alternate resource, and include any extra setup or testing it needs. A second machine that requires a technician already assigned to the first can still leave the plant with one feasible job at a time.
Separate firm work from proposals in the published schedule. A planner can place a proposed order for a quote, then remove that load if sales declines it. Ask who is allowed to turn a tentative job into a firm slot and whether that decision can displace work already released to the floor. Use the same priority policy when you test a rush order.
How Permute connects the evidence
Permute is the context layer we sell between business data and the AI tools people already use. Connect supported records for orders, shifts, and work centers, then match the identifiers that tie planned time to observed operations.
Define which schedule version and delay reasons count in the plant’s reports. A planner can ask why an order moved and inspect the related schedule and material records within their access.
What stays in the source systems
An ERP or scheduling system still owns the schedule and applies shop constraints. Permute can read and reconcile records for review. It does not run a finite scheduling solver, set work-center capacity, or write a new plan into the source system.
Keep the constraint with the next schedule
Finite capacity scheduling turns an overload into a choice about job sequence, resources, or dates. In the oven example, setup order determined whether both jobs fit before any material changed. A late receipt then made the original schedule infeasible. Keep the constraint and schedule version alongside the new promise so people can act on it.
Review a schedule that could not fit the shift
Bring a shared resource calendar, a late order, and the rules used to choose the next slot.
Questions about finite scheduling
Is finite scheduling the same as capacity planning?
Capacity planning can test aggregate load against available hours. Finite scheduling places specific operations in resource windows and respects their order and availability.
What happens when no slot can meet the due date?
Flag the shortfall and identify the constraint. A planner can review another eligible resource, an approved extension, or a revised date with the order owner.
Should setup time be included in capacity?
Yes, when setup occupies the resource. Keep it separate from run time so a change in job sequence can change the correct amount of time.