How to Build an Inspection Schedule That Holds

Most inspection schedules I am shown were not designed. They accrued. A monthly one because the last manager did it monthly, a quarterly one because a client asked for it in 2019, and an annual one nobody can trace at all. They look orderly on a wall planner and they are still being missed.

This post is about the schedule itself, not the app that runs the checklist. If what you are weighing up is the capture side, which photographs go where and how an action closes, that is How to Choose a Site Inspection App. Here the question is narrower and harder: how do you build a schedule that survives contact with a real roster.

Frequency Comes From Risk, Not From Habit.

Australian WHS law fixes very few inspection intervals. What it does instead is put the burden of choosing on you, and then ask you to justify the choice.

The rule worth writing on the wall is in Safe Work Australia’s code of practice on managing risks: priority for review should be based on the level of risk, and control measures for high risks should be reviewed more frequently.¹ Frequency is not a property of the asset. It is a property of what happens if the control has quietly stopped working.

A few things do carry a default. For plant, regulation 213 of the model WHS Regulations requires maintenance, inspection and, where necessary, testing by a competent person, carried out in line with the manufacturer’s recommendations, or a competent person’s recommendations where the manufacturer gives none, and where neither is reasonably practicable, annually.² That last fallback is a floor, not a target. If a machine runs three shifts in a coastal wash-down bay, annual is a decision you will have to defend. Daily checks sit outside this post entirely, because a vehicle pre-start is governed by its own law; that is What a Digital Vehicle Inspection Has to Record.

The practical version, for a schedule you are rebuilding: take your top 10 risks, and for each write down what the control is, what it would look like if it had failed, and how long you are willing to leave that undetected. That last number is your frequency. Most people find two or three lines where the honest answer is far shorter than what they do now, and several where it is longer.

Five Events That Bring a Review Forward.

A schedule is a floor, and the Regulations are explicit that certain events override it. Regulation 38 requires a duty holder to review and, as necessary, revise a control measure when the control is not controlling the risk so far as is reasonably practicable, before a change at the workplace likely to give rise to a new or different risk, when a new hazard or risk is identified, when consultation indicates a review is necessary, or when a health and safety representative asks for one.² The code of practice puts the same five in plain words and adds the line that matters most: do not wait until something goes wrong.¹

Build that into the schedule rather than leaving it to memory. A new subcontractor on site, a changed process, a near miss on the same asset, an item that has now failed three inspections running: each one should be able to pull the next inspection forward, and somebody should own the decision. A schedule that can only tick forward at a fixed interval is a calendar, not a control.

Schedule Drift, and Where It Comes From.

Drift is what happens when a monthly inspection is done on the 3rd, then the 19th, then the 28th, then not at all. It has four usual causes, all fixable.

• The recurrence restarts from completion rather than from the due date. Do one late and every future one moves with it. Six months of small slips become a quarter.
• The inspection is assigned to one person. They take leave and the month is simply skipped, because nothing in the system knows anyone else could do it.
• The frequency was never credible. A weekly inspection on 40 assets with one supervisor was always going to lose, and the schedule quietly taught everyone that overdue is normal.
• Nobody sees the overdue list except the person who is overdue. That is a private to-do list, not a schedule.

The first is a configuration question to put to any vendor in the demo. The other three are yours.

Assign the Role, Not the Person.

This is the single highest-value change most teams can make, and it costs nothing.

A recurring inspection assigned to a named individual is a schedule with a single point of failure in it. When that person changes role, goes on extended leave, or leaves the business, somebody has to notice and reassign every recurrence they held. Nobody does, and it is invisible until an audit.

ComplyFlow lets a scheduled inspection be assigned to named users, to a site, or to a staff category, so that anyone in the Supervisors category on that site can satisfy it, and the requirement is met when any one of them completes it. Its own guidance is blunt about which to choose: create a staff category to reduce the ongoing maintenance of moving people in and out of the schedule, and only assign to an individual where you genuinely have to, such as an annual workstation assessment.⁴ The same idea applies to the actions that come out of the inspection; the person responsible list can be limited to the staff categories that should ever own one.⁷

Pair it with accountability, or you swap one failure for another. The code of practice asks that accountability be clearly allocated so procedures are followed and maintained,¹ and “anyone on site” is not an allocation. Assign the recurrence to the category, and name one person who owns the fact that the category completed it.

What Happens When One Is Missed.

Decide this deliberately, because the default in most systems is that nothing happens loudly enough.

In ComplyFlow, a scheduled inspection that passes its date shows as overdue in the inspection list, and the assigned user gets a notification on their dashboard seven days after the due date, with dashboard notifications emailed out weekly.⁴ That is a reasonable rhythm for a monthly walk-around and much too slow for a high risk item, so the design question is which of your inspections deserve something sharper. Where an inspection action is high priority, an SMS alert can be configured to fire to a nominated number.⁵

Whatever tool you use, hold to one principle: escalation goes to somebody other than the person who missed it. An overdue notice that only ever reaches the person who is already not doing it is a reminder, not an escalation. The second notice belongs to their manager, and the operations manager should be looking at an overdue list by site every week, not an inbox.

The Evidence That the Schedule Was Followed.

This is what an auditor actually asks for, and it is not the thing most teams have ready.

They will not ask how many inspections you ran. They will pick one asset or one site, ask what your schedule says for it, and then ask to see the last four. The gap between those two lists is the finding. Safe Work Australia’s code of practice asks you to keep records of how and when control measures were implemented, monitored, and reviewed, and it is explicit that keeping the records is how you demonstrate what you did to meet the duty.¹ For plant, the register should record the allocated responsibilities, the standards inspected against, and the frequency of inspections, so the schedule itself is part of the record.³

In practice that means you want to be able to filter every checklist item you hold by site, by template, by due date and by completion date, and export it.⁸ If producing that takes a week, you do not have evidence; you have an archive. Inspections and audits is worth judging on this one report as much as on anything in the app.

Two Numbers to Put in Front of the Board.

Safe Work Australia’s 2017 report on measuring and reporting sets a lead and a lag indicator against each control. For a maintenance schedule it gives the lead as the percentage of major equipment assets covered by the schedule, and the lag as the percentage of assets with planned maintenance overdue. For inspections it pairs the count of inspections, or the share of sites inspected, with the proportion reviewed to schedule and the non-conformances found.⁶

Those two do a lot of work. Coverage tells you whether the schedule includes everything it should, which is the failure nobody sees, because an asset that is not on the schedule is never overdue. The overdue percentage tells you whether it is being honoured. Report the count of inspections completed on its own and you will be congratulated for a number that can rise while both of those get worse.

Put Three Columns Beside Your Current Schedule.

Export your current schedule and put three columns beside each line: what the control is, how long you are willing to leave a failure undetected, and who the recurrence is assigned to. Anything assigned to a person’s name rather than a role gets changed this week. Anything where the honest detection window is shorter than the interval goes up. Anything that has been overdue twice in a row is either the wrong frequency or the wrong owner, and it is almost always the wrong owner.

Then check the one thing nobody checks: the list of assets and sites that appear nowhere in the schedule at all. On every portfolio I have looked at, that list is longer than anyone expects, and it is where the next incident is. If your estate runs to dozens of sites, How Inspections Work Across a Property Portfolio picks up from there, and when you want to see a schedule chasing itself, book a demo.