Preventive Maintenance and EDGEBIC: Where the Service Window Meets the Plan

EDGEBIC by User Solutions does not track service intervals and does not warn operators that a machine is due. There is no maintenance due calculation and no maintenance banner at the shop floor terminal. Preventive maintenance meets the plan at exactly one point: once you have decided when a machine comes out of production, EDGEBIC is where you turn that window into hours the scheduler will not book.

That is a narrower role than maintenance modules elsewhere claim, and the narrowness is the point. A reminder nobody acts on changes nothing. A reminder the scheduler cannot see changes nothing either. The step that moves work off a machine is the one that edits its available hours, and that is the step this system owns.

What Lives Where

Keeping the two concerns apart is what makes each of them honest.

If you were expecting the scheduling system to nag you, this will read as a missing feature. In practice the shops that run maintenance well already have a cadence they trust, and what they lack is a plan that respects it. That is the gap worth closing.

The Mechanism: A Per-Day Capacity Override

A work center's day is normally computed from its shift: the shift's hours, scaled by the number of machine instances and the utilization setting, minus anything that makes the day shorter. A per-day capacity override replaces that computation for the days it covers with a number you state.

That makes it the right tool for a service window. You open the per-day capacity dialog on the work center, select the shift, find the date, and enter the hours the machine can genuinely offer. Four hours of service on an eight-hour shift becomes 4. A full day on the bench becomes 0. Add a reason so the entry is self-explanatory later.

Two details decide whether this works in practice. Overrides are entered per day and per shift, so a machine running days and nights needs the entry on both or the plan will quietly book night work onto a machine that is in pieces. And an override wins over the shift formula on any date it covers, so it holds no matter how many times you reschedule. The step-by-step version is how to block a machine for a maintenance day, and the precedence rules are in daily capacity overrides explained.

What Not To Use

A plant holiday closes every work center on every shift. Reaching for one to take a single press down stops the entire factory, which is a large and very visible mistake. Plant holidays belong to national holidays and full shutdowns. A partial holiday is the plant-wide version of a reduced-hours day and is right when the whole site is affected, such as a utility outage or an all-hands meeting.

Shortening the shift is the other tempting shortcut, and it misrepresents the day: the start or end time has to move, which says the machine is unavailable at a time it is actually running. Enter the number, not a new shape for the day.

The Honest Gap

Recurring downtime windows, such as two hours every Wednesday morning, and closures scoped to a single shift are supported by the scheduling engine but have no configuration screen in the current release. For a standing weekly slot that means per-day overrides on the affected dates, which is more entries than a single recurring rule would be.

It is worth stating plainly rather than implying a feature that is not reachable. The compensation is that the routes that do have screens, partial holidays and per-day overrides, are read by every scheduling run with the same precedence as any other capacity input, so nothing about the resulting plan is second-class.

Nothing Moves Until You Run the Schedule

Entering an override changes master data. It does not replan anything by itself. Work sitting on those hours stays there until a scheduling run reads the new calendar and places it elsewhere.

So the sequence is always the same:

1. Maintenance and planning agree the window.
2. The planner enters the per-day capacity override on that work center, on every affected shift and date, with a reason.
3. The next scheduling run plans around the gap instead of into it.
4. The service is performed.
5. The planner resets the override so the normal shift formula applies again.

Step five is the one that gets skipped, and skipping it is expensive in the quiet way: a zero left in place keeps a healthy machine out of the plan, and nobody notices until the load looks strange a month later.

Work Already Logged Is Untouched

A capacity change entered today does not rewrite yesterday. Operations that are completed or in progress keep their actual dates and their logged hours, and a reschedule replans only what has not started. This is the same rule that governs why actuals are immutable everywhere else in the system, and it means you can enter a maintenance window without worrying about disturbing the record of what the plant already did.

Why the Small Regular Losses Matter

Two hours a week does not sound like a scheduling problem until you multiply it. Across a year on a constraint machine, a standing maintenance slot nobody entered is roughly a hundred hours the plan promised and the plant never had, and each of those hours was quietly borrowed from a job that then ran late.

That is the real argument for entering maintenance windows even when they feel too small to bother with. The plan is only as honest as the calendar underneath it, and a calendar that ignores predictable losses produces promises the floor cannot keep. The complete EDGEBIC guide maps how the capacity model feeds the rest of the system, downtime events explained covers the wider picture of hours lost inside a working day, and /edgebic covers the platform as a whole.