In RCM, various maintenance strategies are employed to manage and optimize the reliability of assets. The main maintenance strategies, in usual priority order, include:
RCM Maintenance Strategies
- Predictive Maintenance (PdM)
Predictive maintenance is performed on assets that age in a variable, but with measurements, predictable way. It involves measuring the "condition" of the asset and performing a corrective action if it falls outside condition limits. As such, it is often known as "on-condition" or "condition-based" maintenance. There are two main ways of measuring the asset's condition, so predictive maintenance is often split into two distinct strategies:
- Continuous Monitoring & Machine Learning:
Definition: Continuous Monitoring refers to the ongoing, real-time or near-real-time tracking, analysis, and assessment of asset-related data and performance. Many modern assets can be considered "Intelligent Assets", providing self-diagnostics and self-prognostics. Alternatively, data collection can be combined with machine learning algorithms, such as noise analysis or anomaly analysis algorithms. The use of self-diagnostics, self-prognostics or machine learning algorithms can define when maintenance should be conducted. This is an advanced use of predictive maintenance, where the inspection is automated.
Example: Examples include vibration analysis, thermography, or other condition monitoring techniques.
Applicability: This is only applicable where the asset provides self-diagnostics and self-prognostics, or your organization has set up a condition monitoring system.
- Inspection Tasks / Condition-based Maintenance (CbM):
Definition: The other major form of Predictive Maintenance (PdM) uses scheduled inspection tasks to measure condition. It involves the systematic examination of assets/systems to assess their condition, identify potential issues, and ensure that they meet specified standards. Inspections are a proactive maintenance activity designed to detect and address potential problems before they lead to failures. The inspection will identify poor condition (e.g. loss/degradation of function), allowing corrective actions to be performed.
Example: A regular inspection for rust will identify where components need to be replaced before functional failure is achieved.
Applicability: For an inspection task to be feasible, the failure mode must have a P-F interval large enough to allow for a viable inspection frequency.
- Continuous Monitoring & Machine Learning:
- Preventive Maintenance (PM):
Definition: Preventive maintenance uses scheduled maintenance tasks performed at predetermined intervals to prevent the occurrence of failures or degradation. The maintenance is scheduled via usage intervals (e.g. a 10,000 mile service) or via calendar intervals (e.g. once a year). The maintenance task is usually either a scheduled discard task (replace the component before it fails) or a scheduled restoration task (service the component before it fails).
Example: Replace your spark plugs every 50,000 miles.
Applicability: The probability of failure must increase as the asset ages. Assets that suffer from random failure should not use preventative maintenance.
- Failure Finding (FF):
Definition: Failure Finding (or Detective Maintenance) uses proactive tasks to check that a protective function is working. A protective function is one that protects another function or the operator. Examples include alerts & warnings, guards & shields and emergency shut-down systems. If your smoke alarm is defective, you will not know until you have a fire. A failure finding task will check that the protective device (the smoke alarm) is working as expected.
Example: A smoke alarm test.
Applicability: Any hidden failure mode that cannot be better maintained by the strategies stated above.
- Redesign / Design Out Maintenance (DOM) / One-time Changes:
Definition: Modifying or redesigning equipment or systems to eliminate or mitigate risk or to reduce the need for maintenance tasks.
Example: Upgrading components for improved durability or reliability.
Applicability: Any failure that produces an unacceptable risk and cannot be effectively managed by a proactive maintenance strategy.
- Run-To-Failure (RTF) / No Scheduled Maintenance (NSM):
Definition: Allowing equipment to operate until it fails, without intervening with preventive or predictive maintenance.
Example: A household lightbulb.
Applicability: Typically used for non-critical components where the consequences of failure are acceptable or when the cost of maintenance outweighs the benefits.
Complimentary Maintenance Strategies
The above RCM maintenance strategies can be complimented by the following:
- Walk-around Checks:
Definition: Perform regular basic checks (usually visual) of your site and equipment, looking for poor condition and defects. A form of ad hoc condition-based and remedial maintenance. Note: this is not predictive maintenance and is not dependent on a P-F interval.
Example: Daily supervisor checks.
Applicability: This is usually done in combination with the other strategies outlined here. It is not failure mode specific and will not be identified in the RCM process.
- Remedial (Corrective) Maintenance:
Definition: Repair or replace a component when it fails.
Example: Repair a puncture.
Applicability: This is not a maintenance strategy as such. It is what to do if the maintenance strategy fails and your asset suffers a failure.
- Default Strategies (Unchanged):
Definition: Accepting the default maintenance strategy that comes with the equipment, often based on the manufacturer's recommendations or regulatory compliance.
Applicability: Suitable for components where manufacturer-recommended maintenance practices align with RCM principles, the criticality of the equipment is low, or regulations must be complied with.
- Change Orders (CO):
Definition: Implementing organizational changes that will improve equipment or process. Change Orders are often used as the mechanism to manage Redesign actions (as seen above).
Example: Modifying equipment design, implementing process changes or providing training.
Applicability: If a change is identified that will cost-effectively improve the organization.
RCM SAE Standard
The standard mandates that an RCM process must consider Condition-based, Scheduled Overhaul (Restoration), Scheduled Replacement (Discard), Failure Finding, One-time Changes (Redesign) and Run-To-Failure maintenance strategies.
Above, we have distinguished between 'Continuous Monitoring' and 'Inspection Tasks'. SAE JA1011 does not provide this distinction, it just states condition-based maintenance. Continuous monitoring is mentioned, but is referred to as condition-based maintenance with a zero schedule interval.
Experience has shown us that 'Continuous Monitoring' and 'Inspection Tasks' are two distinct maintenance strategies, with different cost calculations, applicabilities and implementations. However, both are covered in SAE JA1011 as condition-based maintenance.
Selection
The selection of the most appropriate maintenance strategy depends on factors such as equipment criticality, failure consequences, cost considerations, ageing patterns and the nature of the failure mode. The maintenance strategy must be technically feasible and worth doing. RCM aims to optimize the balance between minimizing risk, maintaining equipment reliability and managing maintenance costs efficiently.