Building a planned schedule you can actually staff
Most planned schedules are arithmetic problems that were never done. The hours required exceed the hours available, and the difference is absorbed silently.
A planned preventive maintenance schedule is usually assembled by taking each asset, attaching a frequency, and letting the system generate jobs. It looks complete. What is missing is the step that turns it into a plan: totalling the hours and comparing them to the hours you have. For a separate product-side perspective on turning recurring work into planned action, this practical guide provides additional detail.
When that sum is done for the first time, the answer is frequently that the schedule requires substantially more labour than the department contains. Nothing about the schedule changes as a result. It simply generates more work than can be done, and the surplus becomes an overdue backlog that everyone learns to ignore.
Do the sum in four steps
- Task time — how long each planned task actually takes, including access, isolation, and paperwork.
- Annual frequency — how many times a year each task runs.
- Annual hours per asset — the two multiplied, then summed across the register.
- Available hours — technician headcount times productive hours, which is materially lower than contracted hours.
The fourth number is where most estimates go wrong. A technician contracted for 37 hours does not supply 37 hours of task time: travel, briefings, breaks, parts collection, reactive interruptions and administration all come out of it first. Planning against contracted hours rather than productive hours overstates capacity by a third or more. For a wider operational and compliance reference, consult the Better Buildings O&M guide.
Measure the ratio before you use it, but expect available task time to be well below contracted time. Every schedule built on the contracted figure is over-committed from the day it is published.
Leave room for reactive work
Reactive work is not an exception to the plan; it is a permanent feature of it. A schedule that consumes all available hours has implicitly assumed nothing will break. For a wider operational and compliance reference, consult OSHA hazard prevention guidance.
Reserve a share of capacity for reactive work based on your own history rather than a rule of thumb — count last year's reactive jobs and hours if you have them. Whatever the figure, treat the reserve as unavailable for planned work rather than as a buffer to be raided, otherwise it will be consumed by the schedule within a month and reactive work will start displacing planned work again.
When the sum does not balance
If required hours exceed available hours, there are exactly four honest responses, and choosing none of them is not one of them.
Extend intervals where the evidence supports it. Many inherited frequencies are conservative manufacturer defaults applied regardless of duty and environment. Reduce scope on low-criticality assets, accepting run-to-failure explicitly for items where that is the sensible economic answer. Move work to contractors, which converts a labour constraint into a budget one. Or increase headcount.
What organisations usually do instead is publish the unachievable schedule and let the shortfall appear as overdue jobs. That is the same decision as reducing scope, made invisibly and without choosing which work gets dropped.
Level the load across the year
Annual and quarterly tasks cluster. Left alone, a schedule generated from commissioning dates produces months with double the workload of others, usually with several annual services landing in the same fortnight.
Spread them deliberately. Most annual tasks tolerate being moved by a few weeks, and levelling the load is the difference between a department that is consistently busy and one that alternates between idle and overwhelmed. Seasonal constraints — heating before winter, cooling before summer — are the fixed points to level around.
Review it against what happened
A schedule is a hypothesis about how much work the building needs. Once a year, test it: which planned tasks found nothing, repeatedly, over several cycles, and which assets failed between planned visits.
The first group are candidates for interval extension. The second are candidates for shortening, condition monitoring, or replacement. A schedule that is never revised against outcomes is a schedule that will still be wrong in the same ways in five years.