About reliability-centered maintenance (RCM)
Reliability-centered maintenance (RCM) is a structured process for deciding which maintenance strategy fits each individual asset, based on how critical that asset is to operations and how severe the consequence of its failure would be. Rather than applying one maintenance schedule across an entire fleet of equipment, RCM asks a narrower question asset by asset: what could go wrong, how likely is it, how severe would the consequence be, and what maintenance approach, if any, is worth doing about it. The framework runs on seven structured questions, traces back to a formal aviation reliability study, and was later codified into the SAE JA1011 standard. This guide covers where RCM came from and the seven questions it answers. It also walks through a full worked example, where RCM falls short, and how it connects to an ongoing maintenance program.
Where does reliability-centered maintenance come from?
Reliability-centered maintenance originated from a reliability study led by F. Stanley Nowlan and Howard Heap in the commercial aviation industry in the late 1970s. Their 1978 report, developed out of a United Airlines-era aviation maintenance context, showed that failure rates for most components do not reliably increase with age the way a fixed maintenance calendar assumes. Instead, a maintenance strategy should be driven by the consequence of failure, not by a uniform time interval applied to everything with the same part number.
John Moubray later refined and commercialized that thinking into RCM2, the methodology most widely taught and applied today. SAE JA1011, "Evaluation Criteria for Reliability-Centered Maintenance (RCM) Processes," is the standard that formalizes what a process must include to legitimately call itself RCM. It codifies the seven-questions framework covered next. This aviation origin also explains why RCM's logic is function-first: it starts by asking what an asset is supposed to do, not by starting from a list of failures and working backward.
The seven questions of reliability-centered maintenance
SAE JA1011 codifies RCM as a sequence of seven questions applied to every significant asset function. This is the same framework that gets asked two different ways in search: "the 7 steps of RCM" and "the 7 questions of RCM" both refer to this one sequence, not two different methods.
What are the asset's functions and the performance standards associated with them?
In what ways can it fail to fulfill those functions?
What causes each functional failure?
What happens when each failure occurs?
In what way does each failure matter, meaning what is the actual consequence?
What can be done to predict or prevent each failure?
What should be done if a suitable proactive task cannot be found?
The answer to question five, the consequence, is what actually drives the answers to questions six and seven. A failure with a severe safety or operational consequence justifies proactive investigation even when that investigation is costly. A low-consequence failure may be left to run to failure on purpose, because RCM's goal is matching maintenance effort to consequence, not eliminating every possible failure regardless of cost.
How RCM matches maintenance strategy to asset criticality
Working through the seven questions resolves into one of a small number of maintenance strategy types per asset: scheduled restoration or replacement, condition-based monitoring, a failure-finding task, or a deliberate run-to-failure decision. A failure-finding task applies to a hidden failure that would otherwise go unnoticed until it matters. Run-to-failure applies when the consequence is minor and prevention costs more than the failure itself.
Asset criticality, how important a given asset is to safety, compliance, output, or cost if it fails, is the input that drives which of those four outcomes applies. A compressor with no backup on a line that cannot tolerate downtime gets a different strategy than a nearly identical compressor with a redundant spare, even though both might carry the same manufacturer-recommended interval on paper.
Predictive maintenance and preventive maintenance are two of the strategies RCM can select for a given asset. See those pages for how each works on its own; this piece focuses on how RCM decides which one fits.
A worked reliability-centered maintenance example
Consider a conveyor gearbox on a packaging line. Its function is to transmit drive torque so the line runs at rated speed. Working through the seven questions:
The functional failure is straightforward: the gearbox seizes and the line stops. The cause traces to bearing wear from an inadequate lubrication interval, something that develops gradually rather than striking without warning. The effect is an unplanned line stoppage, downstream product loss, and a multi-hour repair window, since a spare gearbox is not stocked on site.
The consequence assessment matters most here: this is not a safety-critical failure, but it is severe from an operational standpoint. A full line stop with no redundancy and no spare on hand is exactly the kind of consequence RCM is built to catch before it happens.
Given that consequence, a condition-based strategy, vibration or temperature monitoring to catch bearing degradation before it fails, wins out over the two alternatives. A fixed preventive interval would either over-maintain a healthy gearbox on a calendar schedule or miss a failure that develops faster than the schedule assumes. Run-to-failure is unacceptable given how severe the operational consequence is.
Contrast that with a second functional failure on the same line: a guard panel latch loosening over time. The consequence here is low. It is cheap and easy to repair after the fact, and nothing downstream depends on catching it early. Run-to-failure is the correct answer for this failure mode, even though it is technically "letting something fail," because the seven questions point to a different consequence than the gearbox does.
This is exactly the asset-by-asset differentiation RCM is built to produce, one strategy for the gearbox and a different one for the latch, on the same line. It is also exactly the differentiation that collapses back into a single blanket preventive schedule if the analysis is not maintained as a living, revisited process.
How RCM relates to FMEA, FMECA, and other maintenance strategies
Reliability-centered maintenance depends on failure mode information as an input: identifying how an asset can fail is exactly what Failure Mode and Effects Analysis (FMEA) and Failure Mode, Effects, and Criticality Analysis (FMECA) are built to produce, with FMECA adding a formal criticality score on top. RCM is the broader decision framework that takes that failure mode and criticality data, however it was produced, and uses it to select a maintenance strategy for each asset. An organization can run an FMEA or FMECA study without ever formalizing an RCM program around it, but a rigorous RCM program typically has that failure mode thinking built into how it answers its own questions.
For the fuller definition of FMEA, the RPN calculation, and how FMEA and FMECA differ from each other, see FMEA (Failure Mode and Effects Analysis). This piece does not re-derive that material.
The limits of reliability-centered maintenance
RCM is a rigorous framework, not a shortcut, and it carries real costs worth naming plainly rather than treating as a footnote.
A full RCM analysis is time- and labor-intensive to run properly, since it requires working through all seven questions for every significant asset function, not just the obvious failure modes. It depends on accurate failure history and consequence data, and a program built on incomplete or wrong failure data produces a confidently wrong strategy, one that looks rigorous on paper while missing the failure that actually matters. It also requires cross-functional buy-in: maintenance, operations, and safety all have to agree on how severe a given consequence is, and that is a judgment call, not a measurement everyone will read the same way on the first try.
The limitation competitors on this topic tend to skip entirely: an RCM analysis is a snapshot of the operation at the moment it was built. Operations change, equipment ages, and criticality shifts as production volumes, staffing, or redundancy change. A program treated as a one-time study rather than a maintained, revisited data set drifts out of alignment with the operation it was built to protect. None of this makes RCM the wrong choice for critical assets. It means the effort has to be sized to where criticality actually justifies it, applied deliberately rather than uniformly, and kept current rather than filed away.
Why an RCM analysis fails if it stays on paper
A completed RCM analysis produces a maintenance strategy and a task list per asset. On its own, it does not schedule a PM interval, trigger a condition-monitoring alert, or reprioritize a work order. That gap between analysis and execution is where the value of RCM is won or lost.
The value of RCM is only realized once its asset-by-asset strategy decisions get translated into scheduled tasks, condition-monitoring thresholds, and prioritized work, ahead of the failures the analysis already flagged as consequential. An analysis that stays a spreadsheet or a filed report, however rigorously it was built, delivers none of that value. It answers the seven questions correctly and then changes nothing about what actually happens on the floor.
That translation from analysis to executed work is what the next section addresses.
Reliability-centered maintenance in Ultimo
Reliability and maintenance teams running an RCM program use Ultimo to keep its strategy decisions connected to the maintenance program they are meant to inform. This replaces leaving those decisions in a separate document. Proactive Maintenance offers AI-embedded suggestions for proactive maintenance planning and failure prevention, supporting reliability-centered and risk-based maintenance strategies. That is the mechanism that turns an RCM strategy decision, a condition-based monitoring point, a scheduled task, a failure-finding check, into planned work rather than a static conclusion sitting in a report.
Reporting and Dashboards, built on Power BI integration, gives reliability engineers and maintenance directors cross-departmental visibility into which asset strategies are actually reducing downtime over time, closing a loop an RCM study alone cannot close on its own.
One illustrative outcome, cited by name rather than generalized: Dr. Schär, a food manufacturer, reduced downtime by 30% across its manufacturing operations after connecting its maintenance program to Ultimo's EAM software. This is offered as one customer's result, not a universal RCM guarantee. Ultimo is Intelligent Asset Management software, the category that brings enterprise asset management, CMMS functionality, analytics, and AI-native experiences together for asset-intensive organizations running a reliability-centered maintenance program or building toward one.
Frequently Asked Questions
Frequently asked questions
Does RCM always mean reliability-centered maintenance?
No. "RCM" is also a common abbreviation for Revenue Cycle Management, an unrelated healthcare billing and finance discipline, and it is in fact the more common meaning of the bare acronym in general search results. This article covers reliability-centered maintenance, the asset-management and reliability-engineering discipline that decides which maintenance strategy fits each asset based on criticality and failure consequence. The confusion is common because search engines and AI assistants often default to the healthcare meaning when the acronym appears without qualification, especially in a business or finance context.
What are the seven questions of RCM?
The seven questions, codified in SAE JA1011, are: what are the asset's functions, how can it fail to perform them, what causes each failure, what happens when it occurs, how much does each failure matter, what can predict or prevent it, and what should be done if no proactive task exists. This is the same sequence referred to as "the 7 steps of RCM" in some searches; both phrasings describe one framework, not two different methods. Working through them in order is what produces a defensible maintenance strategy decision for each asset, not a checklist to complete casually.
What are the disadvantages of RCM?
A full RCM analysis is time- and labor-intensive to run properly across every significant asset function, and it depends on accurate failure history and consequence data; incomplete or wrong data produces a confidently wrong strategy. It also requires cross-functional agreement on how severe a given consequence is, which is a judgment call rather than a measurement, and it can go stale if treated as a one-time study instead of a maintained, revisited process as operations change. None of these limitations make RCM the wrong choice for critical assets, but they explain why the effort has to be sized deliberately rather than applied everywhere at once.
How is RCM different from preventive maintenance?
Preventive maintenance is maintenance performed on a fixed schedule regardless of an asset's actual condition or criticality. It is one of the strategies RCM might select for a given asset, not a synonym for RCM itself. RCM is the decision framework that determines whether a fixed preventive interval is the right approach for a given asset, or whether condition-based monitoring, a failure-finding task, or run-to-failure would serve better. See preventive maintenance for the fuller definition. Choosing between them is exactly the decision RCM's seven questions are designed to make for each individual asset.
Does RCM reduce downtime?
A rigorous, well-maintained RCM program is built to reduce downtime by directing maintenance effort toward the failures with the most severe consequences instead of spreading it evenly across every asset. Dr. Schär, a food manufacturer, reduced downtime by 30% across its manufacturing operations after connecting its maintenance program to Ultimo's EAM software, offered here as one customer's outcome from a connected maintenance program, not a guarantee tied specifically to running an RCM analysis. Any downtime reduction still depends on translating an RCM analysis into scheduled tasks and monitoring, not on the analysis alone.