Scheduling a maintenance team of five
Below a certain size, scheduling stops being an optimisation problem and becomes a search for any arrangement that satisfies the constraints.
Scheduling literature assumes a workforce large enough that assignment is a statistical exercise. With five technicians, one absence removes a fifth of capacity, and the number of valid arrangements — satisfying skills, tickets, rest and coverage — is often in single figures. For a product-side comparison related to annual labour capacity calculations, see how many work hours are in a year of full-time work before deciding which data the operation actually needs.
Skills are the binding constraint
Headcount is rarely what limits a small maintenance schedule. What limits it is how many people hold the specific qualification a given job requires: the electrical competence, the gas ticket, the confined space certification, the authorised person appointment.
Build the skills matrix before the schedule — people down the side, competencies across the top — and look for the sparse columns. Any capability held by exactly one person is both a scheduling constraint and a continuity risk, and small departments usually have more of these than anyone has counted. For a wider operational and compliance reference, consult the Department of Labor field operations handbook.
Each is a single point of failure with a long replacement time, because the substitute cannot be hired next week and may not exist locally. Building a second holder for each is worth more than most efficiency measures.
Place the least flexible things first
A tightly constrained schedule becomes tractable if it is built in order of rigidity, and impossible if it is built in order of convenience.
- Approved leave, training and known absence.
- Statutory work with a fixed deadline.
- Single-holder skills — whoever is the only qualified person goes on the board next.
- Contractor attendances already booked, which need an escort or an authoriser.
- The rest of the planned schedule.
- Reactive reserve.
- Deferrable work, into whatever remains.
Keeping the last category last makes the amount of remaining slack visible. If nothing fits in it, the week has no give, and it is better to know that on Monday than to discover it on Thursday.
Protect the reactive reserve
A schedule with every hour allocated to planned work has assumed nothing will break. The reserve exists precisely so that the first emergency does not displace a statutory inspection.
Size it from your own history rather than a rule of thumb, and treat it as unavailable rather than as a buffer. Reserves described as flexible are consumed within a fortnight, after which reactive work resumes displacing planned work and the cycle restarts.
Two competencies per person, not five
Small teams multi-skill by necessity, and the instinct is to make everyone capable of everything. Spread too thin, this produces a team that is passable at many things and depended upon for none, and it makes the schedule harder to read.
A primary discipline and one credible secondary gives every task two qualified people — the threshold at which a single absence stops being a crisis — while keeping the depth that complex plant actually requires.
Expect the week to be edited
A five-person schedule will change. That is a property of reactive work, not a planning failure, and treating each change as an exception generates unnecessary friction.
What helps is one authoritative version of the week, a defined way to change it, and a rule that a change is not real until it is recorded. Reallocations agreed verbally and never entered anywhere are the main reason a department's records disagree with what actually happened.
Publish it far enough ahead to be useful
Technicians who know their pattern two weeks out can arrange their lives. Those who find out on Friday cannot, and in a trade with a tight labour market that is a retention issue rather than an administrative one.
The trade-off is that a schedule published early is less accurate at the moment of publication. That trade is almost always worth taking: a plan that changes twice with notice is easier to live with than one that arrives the day before.