How a Schedule Accounts for Holidays and Downtime

A schedule accounts for holidays and downtime by turning them into the effective hours a work center actually has on a given day. EDGEBIC by User Solutions computes those hours from one formula per work center, shift, and date, subtracting downtime and partial holidays, applying instances and utilization, and honoring any capacity override. Before it allocates a single hour of work, the engine asks this calculation how much the machine can do, so every Gantt bar, every utilization percentage, and every due-date promise rests on the same numbers.

Get this configuration right and the schedule matches what the plant can deliver. Get it wrong and everything downstream is wrong too, because a job cannot be placed correctly against capacity that does not reflect reality. The good news is that the whole thing reduces to one resolution chain and one formula.

The one formula behind every capacity number

For any combination of work center, shift, and date, EDGEBIC resolves effective hours in three steps and takes the first that applies:

1. Daily capacity override. If a planner has entered a specific number of hours for this exact work center, shift, and date, that number is used verbatim.
2. Date-range override. If a range-based total covers this date, the engine spreads it across the available days and shifts in the window.
3. The shift formula. Otherwise:

gross = shift end - shift start
net   = gross - break - downtime - partial holiday hours
capacity = max(0, net) x number of instances x (utilization % / 100)

A machining center on an 08:00 to 16:30 shift, 30 minutes of it lost to a break, an hour of Wednesday maintenance, two instances, and 85 percent utilization comes out at 11.9 hours on a Wednesday: 8.5 gross, minus 0.5 and 1.0 for 7.0 net, times 2 instances times 0.85. That is the exact figure the engine schedules against and the exact figure the capacity dashboard shows, because both run the same calculation. How those first two subtractions get told to the system is the practical question, and the answer is the shift hours themselves or a per-day override, not a downtime record.

Before any of this, a date gate decides whether the work center is even open. A whole-day plant holiday, the work center's own availability flag or effective date, or a non-working weekend all return "closed" and the date is skipped entirely, with capacity of zero. Which calendar that gate reads depends on whether the work center inherits the plant's shifts and holidays or overrides them with its own.

Three levels of holidays that combine cleanly

Not every closure is plant-wide, so EDGEBIC models holidays at three scopes:

• Plant-wide holiday. Closes every work center and every shift. A public holiday, a company shutdown. It can be recurring (same month and day every year) or one-off, and it can be whole-day or a partial window. This is the only one with a screen you can enter records on.
• Shift holiday. Closes a single shift on the machines that use it. Honored by the engine, but not creatable in the current release: express it as per-day capacity overrides of zero on that shift.
• Work-center holiday. Closes one machine on a specific date. Also honored but not creatable: an annual teardown on CNC-3 is a per-day capacity override of zero on that work center.

They combine without double-counting because they are resolved in priority. If a plant-wide whole-day holiday exists for a date, every work center reports zero and no allocations are created, regardless of what shift or machine holidays also exist. If there is no plant closure, a shift holiday narrows the block to its shift, and a work-center holiday narrows it to its machine. A partial-day holiday (say a 45-minute monthly union assembly from 10:00 to 10:45) does not close anything; it subtracts only the overlap between the holiday window and each shift's window from that shift's hours.

Downtime: subtracting hours, not days

Downtime is different from a holiday. A holiday removes a day or a window; downtime subtracts hours from within a shift while leaving the shift alive. A recurring two-hour lubrication every Monday takes two hours off Monday's capacity on that machine, but the rest of Monday still schedules. Downtime can be recurring (fires every week on its chosen day) or one-off (fires only on its specific date), and its hours come off in the net line of the formula, before instances and utilization multiply. So a downtime window that exceeds the shift length simply drives capacity to zero rather than negative, which is exactly what you want.

Overrides: when the planner owns the number

The two override paths at the top of the resolution chain exist for the cases the standing calendar cannot express.

A daily override authorizes or restricts one cell. To open a normally non-working Saturday for a rush order, you enter the approved hours for that work center, shift, and date, say 4.0, with a reason. The engine uses 4.0 exactly, without re-multiplying by instances or utilization, so you enter the total you actually want the machine to have. This is the one place the planner, not the formula, owns the final number, and it wins over everything else for that cell.

A date-range override models a period-wide budget: a start and end date and a total capacity, which the engine divides across the available days in the window and then again by the shifts active each day. It sits in the middle of the resolution chain and a per-day override beats it on any date both cover. It is worth knowing the rung exists, because it explains the ordering, but the current release exposes no entry point for creating one. In practice you do the division yourself and enter per-day overrides: retooling a welding cell on 60 hours across five weekdays and two shifts is 6 hours per shift slot, typed into ten cells with a shared reason, and any plant holiday inside the range should be left out of your day count.

Why one chokepoint matters

The reason to insist on a single resolution chain is consistency. The scheduling engine and the capacity dashboard could, in a lesser design, compute available hours two different ways and disagree, leaving a planner unsure which number to trust. EDGEBIC computes capacity through the same logic on both sides, pre-loading the same override records and applying the same formula, so the dashboard's bar for Wednesday and the engine's Wednesday allocation always agree. When they ever appear not to, the cause is a stale snapshot (a change made after the run started) rather than two competing calculations.

This capacity calculation is the foundation the rest of the engine builds on. The multi-shift allocator that fills capacity consumes exactly these per-shift buckets, and the operator skill constraint stacks on top of them: a window has to be an open, capacity-bearing machine slot before the engine even asks whether a qualified person can staff it. Understanding holidays and downtime is therefore the prerequisite for reading almost any placement the schedule makes.

For the wider picture of how these hours flow into a plan, see the complete scheduling engine guide, and for the general distinction between a schedule that respects capacity and one that ignores it, finite versus infinite capacity scheduling covers why the bucket matters at all.

To configure your own plant calendar and watch a holiday zero out a day the way it should, bring your data to a demo.