Generating Value Through Predictive Maintenance – Part One: The Maintenance Maturity Journey


The improvement of maintenance processes is a continual journey. In this first in a series, we provide a maturity model and framework for progressing on that journey, from reactive maintenance to preventative, condition-based, predictive, prescriptive, and ultimately autonomous maintenance.


This article is an excerpt from the reportThe Value of Foresight: Generating Value Through Integrated Predictive Maintenance. A copy of the full report can bedownloaded here.

The Maintenance Maturity Journey

For asset-intensive organizations, the maturity of their maintenance practices is a key determinant of their ability to operate reliably, without interruption, profitably. Investments in improving maintenance practices, processes, and systems can have a large return. The improvement of maintenance practices is a journey, not a destination, as outlined in the maintenance maturity model below.

Figure 1 – Maintenance Maturity Model

As a company progresses and adopts each next higher level of maturity, it does not necessarily abandon lower-level, less sophisticated maintenance strategies. Companies that implement Reliability Centered Maintenance (RCM)1 recognize that a one-size-fits-all maintenance strategy wastes scarce maintenance resources on less critical assets while underserving more critical assets. For example, a company may continue a reactive approach to very-low-criticality items (e.g. light bulbs, pencil sharpeners, etc.) ignoring them until they fail. Preventative maintenance may be appropriate for low-medium criticality assets requiring periodic inspection, replacement of lubricants, and so forth. However, with preventative maintenance, some equipment will be serviced before necessary, whereas other equipment will fail before being serviced.

Condition-based maintenance starts to address uptime and maintenance costs by monitoring one or more key measurements, such as temperature, vibration, pressure, or other indicators of an out-of-spec condition.
Thus, maintenance tasks are more likely to be performed when they are actually needed. However, condition-monitoring typically involves monitoring only a few key measurements in isolation, lacking a more comprehensive view of overall asset health and more subtle indicators of deteriorating operation.

Predictive maintenance (PdM) typically involves a broader set of input data and more sophisticated analysis (e.g. motor current analysis, oil analysis, infrared thermography, ultrasonic analysis, etc.). More importantly, predictive maintenance analyzes these multiple variables together to provide a more reliable indicator of the overall health and condition of the asset and a more accurate prediction of when a piece of equipment is going to fail and what should be done about it.

The Advantages of Predictive Maintenance (PdM)

With predictive maintenance, equipment is serviced more in line with actual wear and tear and need for service, while reducing unexpected outages. It brings the multiple advantages of making fewer scheduled maintenance repairs or replacements, using fewer maintenance resources (including spare parts and supplies), while simultaneously reducing failures. A well-run predictive maintenance program can have a dramatic impact on OEE (Overall Equipment Effectiveness). It impacts each of the three components of OEE: higher availability (uptime), increased performance (% of full design speed/output being achieved), and higher quality (first pass yield). According to the DOE,2 implementing a functional predictive maintenance program can reduce equipment breakdowns by 70%-75%, reduce maintenance costs by 25%-30%, reduce downtime by 35%-45%, and increase production by 20%-25%.

Predictive Maintenance also provides the prerequisite foundation for Prescriptive and Autonomous Maintenance. Prescriptive maintenance builds on the infrastructure and data collected for predictive maintenance, observing the various corrective actions taken by maintenance personnel and the outcomes that resulted. Using machine learning, prescriptive maintenance learns and recommends the best timing and course of action for a given set of conditions. Autonomous maintenance takes it a step further by executing those actions automatically, without human intervention.3 Currently, autonomous maintenance is largely a vision for the future, rather than a current reality for most organizations. Even prescriptive maintenance is in early stages at only the most advanced companies. In contrast, many companies are ripe for implementing predictive maintenance, to gain substantial benefits and significant improvements over their current approach.

In Part Two of this series, we look at what it means to take a holistic, integrated approach to PdM.


1 RCM is a systematic approach to optimizing the mix of maintenance strategies by prioritizing the failure modes and impact of failure for various assets. It’s roughly analogous to the risk-based approach used by some organizations to optimize the use of scarce compliance resources. — Return to article text above
2 See Operations & Maintenance Best Practices: A Guide to Achieving Operational EfficiencyReturn to article text above
3 There are still humans involved in the physical steps such as gathering spare parts and performing the actual maintenance tasks. However, the system automatically makes decisions about what tasks to execute and when, optimal replenishment and location of spare parts inventory, assignment and scheduling of maintenance resources, and so forth. — Return to article text above

To view other articles from this issue of the brief, click here.

Scroll to Top