When machines keep failing between services, the instinctive response is to shorten the interval. Sometimes that is right. More often it is expensive and ineffective, because the interval was never the problem — the service scope did not cover the failing item, or the service was scheduled and not actually performed to specification, or the schedule is running on calendar dates for machines whose duty varies enormously. Shortening an interval that already misses the cause simply means missing it more often at greater cost. The useful first step is working out which of several quite different problems you have. You can go through the diagnosis in a free 30-minute session.
Breaking Down Between Services? Diagnose Before You Shorten
Six different causes of between-service failure, how to tell them apart, and why only one of them is fixed by servicing more often.
Six Diagnoses
How to tell: the same components keep failing, and when you read the actual PM task list, they are not on it.
Fix: add them to the scope. Shortening the interval repeats a service that was never going to catch this.
How to tell: service times are consistently shorter than the task list implies, or completion is recorded with no parts consumed.
Fix: a scope that is recorded item by item rather than signed off as one line, so what was actually done is visible.
How to tell: failures cluster on the hardest-worked machines while lightly used ones are serviced early and never fail.
Fix: move from calendar to meter hours, or to cycles for systems that cycle rather than run.
How to tell: failures are spread across the fleet, cluster near the end of each interval, and involve items that are in scope and being done.
Fix: shorten it. This is the one case where the instinctive answer is correct.
How to tell: the failures concentrate on a single asset rather than spreading across identical units.
Fix: investigate that machine. Changing the fleet schedule because of one bad unit is an expensive way to avoid a diagnosis.
How to tell: failure rates changed after a supplier or specification change, with nothing else different.
Fix: a supply question rather than a scheduling one, and worth checking before anyone rewrites the schedule.
The Data That Tells You Which
All of this comes from work order history you already generate, provided failures are recorded against the machine and the component.
Which components are failing, grouped across the fleet rather than read one job at a time.
Whether failures concentrate on one asset or spread across identical machines.
Where in the interval they occur — early, middle, or clustered near the end.
Whether the failing item appears anywhere in the PM task list.
Labour time and parts consumed per service, compared against what the scope requires.
Hours accumulated between services, and how much they vary across the fleet.
Get the History That Answers This
Failures recorded per machine and per component, with service scope captured item by item rather than as a single sign-off.
Why Shortening Blindly Is Expensive
Every additional service costs labour, parts, and availability. Halving an interval roughly doubles the servicing burden on a workshop that was probably already the constraint, which pushes other machines closer to their own failures.
It also hides the original problem. If the failing item was never in scope, more frequent services produce more cost and the same failures, and the conclusion drawn is usually that PM does not work rather than that the scope was wrong.
And it is hard to reverse. Once an interval is shortened after a run of failures, lengthening it again requires someone to argue for less maintenance, which is a conversation nobody wants to have even when the data supports it.
Frequently Asked Questions
Should we just follow the OEM interval?
It is the right starting point and it assumes typical conditions. Severe duty, heavy dust, or high cycle counts justify deviation, but the deviation should follow from your own failure data rather than from a general belief that Indian conditions are harder. You can work through the comparison on a call with our team.
How do we tell whether services are actually being done fully?
Compare recorded labour time and parts consumption against what the scope requires. A service recorded as complete in an hour with no filters consumed did not include the filter change.
Can different machines of the same model have different intervals?
Yes, and often they should, since duty varies by site and application. A machine on rock is not the same asset as an identical one on soft ground.
How much history do we need before changing intervals?
Enough to see a pattern rather than a run of bad luck. A handful of failures over a few months is not evidence; consistent failures across several service cycles are.
Where should we start?
By listing which components are actually failing and checking each one against the PM task list. That single comparison resolves the most common diagnosis. Start with a free trial.
Find the Cause, Then Change the Schedule
Check whether the failing components are in scope, whether the service is being done in full, whether the interval basis matches how the machine is worked, and whether the failures belong to the fleet or to one bad machine — before shortening anything.
