What is maintenance scheduling?

What maintenance scheduling is

Maintenance scheduling is deciding which maintenance jobs get done, by whom, and on which day — and committing to it. It takes a list of work that needs doing and turns it into a week that a specific set of people can actually complete, given the shift pattern, the spares on hand and the hours the production plan will release the machine for.

It is not the same as deciding what maintenance to do, or how often. That decision comes earlier. Scheduling is the part where intention meets the calendar, and it is where most maintenance programmes quietly fall apart.

Planning and scheduling are two different jobs

This distinction is worth getting right, because a plant that confuses them ends up with a beautiful PM plan that nobody executes.

Planning answers what and how. What work is needed on this machine, what the procedure is, which spares and tools it needs, how many people, how many hours, what safety permits apply. A planner's output is a job that is ready to be done, with nothing missing.

Scheduling answers when and who. Given a stack of ready jobs and a finite number of technician-hours next week, which jobs go on which day, on which shift, in which order, and who does them. A scheduler's output is a committed week.

The failure mode when they are merged into one person's head is always the same: jobs get scheduled that are not ready. The technician arrives, the spare is not there, the machine is running production, the permit was not raised. The job slips, and after enough slips the schedule stops being believed, which is the real damage.

The five things a schedule needs before it can exist

1. A list of jobs that are actually ready. Spares identified and available, procedure written, permits possible.
2. Honest labour hours. Not the headcount on the roster. The hours actually available after leave, absenteeism, training and the breakdown work that will inevitably arrive.
3. Machine availability from production. The hours the plan will genuinely release each machine for, agreed rather than assumed.
4. Priority. When next week cannot hold everything, what gets dropped first, decided by asset criticality rather than by whoever asked most recently.
5. A backlog you can see. Every ready job not yet scheduled, with its age. A backlog that is invisible always grows.

What it looks like week to week

The rhythm that works in most plants is simple and boring, which is why it survives.

Midweek, the planner closes next week's job list: everything ready, hours estimated, spares confirmed. Thursday or Friday, the maintenance head and the production planner sit for thirty minutes and agree the week — which machines are released, when, and for how long. That meeting is the entire method. Without it the schedule is a wish.

The schedule is published before the week starts, visible to every shift. Daily, at the shift meeting, what was completed is marked and what slipped is re-scheduled with a reason. At week end, one number is reviewed: what fraction of the committed schedule was actually done.

What breaks a schedule in an Indian plant

The textbook version of this assumes a stable week. Plants here have specific pressures that a generic guide will not mention.

• The production plan changes on Monday. Machine hours agreed on Friday evaporate. The answer is not to abandon scheduling but to schedule less: commit to what will survive a plan change, and hold the rest as a ready backlog.
• Festival season and harvest absenteeism. Available hours can fall by a third in some weeks and everybody in the plant knows which weeks those are. Plan the light weeks deliberately instead of discovering them.
• Monsoon shutdown. For many plants this is the one long window for heavy jobs. Work backwards from it: the jobs that need it should be identified months ahead, with spares ordered against long lead times.
• Contract labour turnover. A schedule that depends on a specific person's knowledge fails when he leaves. Job procedures have to be written for somebody doing it the first time.
• Single spare, long lead time. A job needing an imported part cannot be scheduled honestly until the part is in the store. Scheduling it anyway is how backlogs become fiction.
• The shift boundary. A four-hour job started at the end of a shift becomes two half-done jobs and a handover nobody wrote down. Schedule jobs inside a shift or plan the handover explicitly.

Building your first schedule in a week

If nothing exists today, this sequence works and takes about a week of part-time effort.

Day one — list the machines that matter. Not every asset. The ones where a stoppage stops output or costs real money. Usually thirty to sixty in a mid-size plant.

Day two — one PM per machine, no more. The manufacturer's basic interval service, or the thing your most experienced fitter says should be done monthly. Resist the urge to build a perfect task library.

Day three — put them on a calendar with a basis. Monthly, quarterly, or every so many running hours. Spread them so no week carries more than the hours you have. Use a format with the columns you will actually need — machine, task, frequency, basis, owner, last done, next due. We publish a ready-made PM schedule format with exactly those columns, in Excel.

Day four — agree machine hours with production. Thirty minutes. Write down what was agreed.

Day five — publish it where the shift can see it, and start marking done or not done, with a reason when it slips. The reasons are the most valuable thing you will collect in the first month.

Do not add condition monitoring, criticality scoring or a software tool in week one. Get a schedule that survives four weeks first.

Knowing whether it is working

Four numbers, and only one of them is the famous one.

• Schedule compliance — of the jobs committed for the week, how many were completed in that week. Below about seventy percent the schedule is not being believed; chase why before chasing the number.
• PM compliance — of the preventive jobs due, how many were done. High PM compliance with high breakdowns means the PMs are the wrong PMs.
• Backlog age — not the size of the backlog but how old the oldest ready job is. A growing age says you are scheduling less than arrives.
• Emergency work ratio — what share of hours went to unplanned work. This is the number that actually tells you whether maintenance is in control of its week.

Watch them together. Any one of them alone can be made to look good by someone under pressure.

Where software helps, and where it does not

A schedule on a spreadsheet works, and plenty of good plants run one. Software earns its place at a specific point: when due dates have to be generated from running hours rather than dates, when more than one person updates the same schedule, when you want compliance measured rather than estimated, and when the evidence has to survive an audit or a warranty claim.

What software cannot do is agree machine hours with production, decide priority, or make a technician stamp a job complete at the moment he finished it. Those are the parts that decide whether scheduling works, and they are all human.

If you want to see scheduling running against your own machine list, our maintenance scheduling module does the due-date generation, the weekly planner and the compliance tracking, and book a demo puts it on your data rather than ours.