Planning Furniture Shop Capacity Season by Season

Furniture demand is not flat, and neither is a furniture shop's capacity. EDGEBIC by User Solutions resolves how many hours a work center actually has on any given date through a fixed chain: a daily override, then a date-range override, then the shift formula. Understanding that chain is the difference between a seasonal plan you can quote from and one that quietly assumes every week is average.

The Pain: A Calendar That Does Not Match the Year

Case goods and upholstery shops live on a rhythm. There is a market show to build for. There is a pre-holiday retail push. There is a summer stretch when half the finish room is on vacation. There is a shutdown week. There is the retooling period after a new collection lands.

Most scheduling setups model none of it. A shift pattern is entered once, and after that the plan believes every Tuesday in the year is identical. So the plan is optimistic in July, pessimistic in October, and confidently wrong about the shutdown week until somebody notices jobs scheduled into a closed plant.

The workaround is usually a planner mentally discounting the system's dates, which works right up until they are on holiday themselves. The general practice for this industry is covered in the furniture and woodworking scheduling overview.

The Resolution Chain

Every capacity question resolves the same way, first match wins.

Two things follow from that ordering, and both catch people out.

A daily override beats a date-range override on any date both cover. If you have set a reduced-capacity month and then approve one specific Saturday, the Saturday value wins for that cell.

And a daily override is used exactly as entered. The engine does not multiply it by instances or by the utilization percentage. The planner owns the final number, which is right, because a planner authorising four hours of Saturday overtime means four hours of output, not four hours multiplied by two booths and then discounted.

Full mechanism detail is in shifts and calendars explained.

The Formula, Worked

Take a finish room on a Wednesday.

The engine computes:

gross          = 16:30 minus 08:00        = 8.5 h
deductions     = 0.5 break + 1.0 downtime = 1.5 h
net            = 8.5 minus 1.5            = 7.0 h
raw capacity   = 7.0 x 2 booths           = 14.0 h
effective      = 14.0 x 0.85              = 11.9 h

A job needing 10 hours fits. A job needing 13 spills into Thursday.

Whenever a day looks wrong, walk those four numbers in order. In practice the surprise is almost always the utilization percentage, which somebody set to 85 two years ago and nobody has revisited, or an instance count that no longer matches the machines on the floor.

Three Seasonal Tools, Three Different Jobs

The chain is only useful if you use the right tool for each situation. Three come up constantly in furniture.

Plant shutdown: a holiday

The shutdown week is a closure, and closures belong in the holiday layer. There are three scopes:

• Plant-wide holidays close every work center on every shift. National holidays and the shutdown week live here.
• Shift holidays close one shift, which is how you model a night-shift-only maintenance week without touching days.
• Work center holidays close one machine, which is how you model a teardown day on the wide-belt sander while everything else runs.

Plant-wide holidays can also be partial. A negotiated 45-minute assembly on the first Monday of the month is a partial-day holiday: the hours come out of that day's capacity but the day stays open. Notably, a partial holiday is subtracted before the instance and utilization multipliers, exactly like downtime, so 45 minutes off a two-booth room at 85 percent removes more than 45 minutes of planned capacity. That is intentional and it matches reality: both booths lose the time.

Recurring holidays repeat on the same month and day each year and need a correct reference date to anchor from. That is the field most often entered wrong on first setup.

Preventive maintenance: a downtime event

A holiday deletes a day. Downtime deletes hours inside a day. A weekly two-hour PM window on the moulder is downtime, and modeling it as a holiday throws away six schedulable hours every week.

Downtime can be one-off or recurring by weekday, and it subtracts from the shift before the multipliers. It also must not exceed the shift's own hours, which sounds obvious and is worth checking when somebody enters a full-day PM as a downtime event rather than a work center holiday.

Approved overtime: a daily override

The Saturday case. Without an override, Saturday is not a working day and no shift exists, so the formula produces zero and the day is skipped.

A daily override for that work center, shift and date, carrying the hours and a reason, makes it real:

Work center   : Finish room
Shift         : Day
Date          : Saturday
Hours         : 4.0
Reason        : Overtime approved, rush order for the market show

The engine returns 4.0 for that cell without computing the formula at all, and the scheduler can now place work on that Saturday. The capacity view shows it in a distinct color so nobody mistakes an authorised Saturday for a normal one.

One trap worth knowing: setting an override of zero on a weekend does not make the shift appear and then empty it. Zero on a day that was already closed changes nothing.

Reduced-capacity periods: a date-range override

Retooling, a staffing dip, a partial line rebuild. You know the total hours available across a window but not how they land day by day.

Work center : Assembly
Start       : Monday July 14
End         : Friday July 18
Capacity    : 60.0 total hours
Reason      : Retooling, partial availability

The engine counts the available days in the window, five weekdays here, giving 12 hours per day. That work center runs two shifts a day, so each shift slot gets 6 hours instead of its usual figure. Every run that week respects the budget automatically, and it lifts when the window ends.

The distribution is across available days only. If a plant holiday falls inside the window, the 60 hours spread across the days that remain rather than evaporating with the closed day. That is usually what you want, and it is worth knowing so the numbers do not surprise you.

Building the Seasonal Calendar

A workable annual routine, once a year plus small adjustments:

1. Enter the fixed closures as plant-wide holidays: national holidays, the shutdown week, the inventory day.
2. Enter recurring maintenance as downtime events on the machines that carry it, by weekday.
3. Enter known reduced periods as date-range overrides: the retooling window, the vacation stretch, the ramp after a collection launch.
4. Leave overtime out of the annual plan. Daily overrides are a weekly decision made against a real backlog, not a forecast.
5. Check the utilization percentages while you are in there. They are the quietest source of wrong capacity in most installations.

Then read the capacity view before each season rather than after it. A week that goes red before the market show is a staffing conversation in August; the same week discovered in October is an expedite.

The Dashboard and the Scheduler Must Agree

A capacity number is only useful if everyone is reading the same one.

The resource calendar and the scheduling engine resolve capacity through the same chain and the same override data. That is deliberate and it is worth verifying after any calendar change: open the resource calendar for a work center and a date, note the hours, then check what the schedule actually placed there. If the two disagree, one of them is reading a stale override snapshot and the discrepancy will not resolve itself.

The calendar also has a backlog view that shows what is loaded against each cell rather than only what is available. That is the view to read before a season rather than the plain availability figure, because a work center with 11.9 hours of capacity and 14 hours of load is a problem you can still solve in August and cannot solve in October.

When a cell shows zero and should not, there is a defined order to check rather than a guess: plant holidays, then shift holidays, then work center holidays, then whether the shift has an entry for that weekday at all, then downtime, then overrides. Working the list top to bottom finds the cause in a minute or two. Guessing at it can take an afternoon, and the answer is usually the weekday entry, because a shift with no Saturday row produces no Saturday capacity whatever its nominal hours suggest.

Where Capacity Meets the Promise

Seasonal capacity and promised dates are the same question asked twice.

If you schedule make-to-order work backward from delivery dates, the calendar decides whether backward is even feasible: a shutdown week inside a job's window removes the shifts the backward pass needs, and the job rolls forward instead. That interaction is covered in scheduling custom furniture to a promised date and in backward scheduling explained.

If you quote lead times, the calendar is what makes the quote seasonal rather than average. A quote simulation run against a July calendar and one run against an October calendar should not return the same date, and with the calendar modeled they do not. See quote simulation explained.

And when a work center shows zero on a day it should not, there is a defined order to check: plant holidays first, then shift holidays, then work center holidays, then whether the shift runs on that weekday at all, then downtime, then overrides. Working the list beats guessing. See why a work center shows zero capacity.

Further up the same supply chain, capacity is set by drying rather than machining: scheduling sawmills and lumber processing around kilns and planer lines covers that end. For the industry picture, see the furniture and woodworking scheduling overview and the furniture manufacturing software page. For the rest of this series, see EDGEBIC by industry, and for the product, EDGEBIC.