Wireless Condition Monitoring Systems Series

The following article is the first of a three-part series in which we explore and clarify the challenges facing maintenance and reliability professionals considering wireless condition monitoring systems for their operations.

In this part, we’ll review the PdM framework as a foundation for adoption of any new technology or services, discussing such subjects as the challenge of assessing program ROI, some reasons for disappointing results and integration of condition monitoring into an existing program.

Part One – Why Should We Consider Deploying a Wireless Continuous Monitoring System?

Judicious application of Condition monitoring is an essential element of a robust PdM program. One of the challenges is to quantify the benefits financially, to justify the significant investment. Many advocates, consultants, and vendors project significant cost savings as a result of increased condition monitoring.  But these assertions are questionable when projections lack sufficient information regarding the operations or maintenance history or are based on non-comparable industries.

Industry aspirations for PdM are not just about costs. When Plant Services conducted their annual survey of Predictive Maintenance (PdM) programs, they included a question regarding overall program satisfaction. The responses indicated that 51% of programs were assessed as not effective or needing improvement. The question is why our programs struggle with overall satisfaction. An understanding of the underlying reasons and most importantly resolving them to ensure our programs are built on a solid foundation, complete with stakeholder expectations being exceeded. This will clear the path for expansion and continued funding.

Why the Disconnect?

From our experience, we believe many organizations struggle to project the financial benefits of a PdM program including:

Managing failure effectively but not eliminating recurring patterns.

Not getting to the root cause likely means the problem will come back

Cost avoidance versus cost savings – leadership may expect hard dollar savings

The typical benefit from PdM is derived from cost avoidance rather than cost reduction

Alignment of program metrics and site business goals

Ensure your PdM goals have a “line of sight” (impact) to the same goals the leadership is measured against.

“Leadership track business outcome metrics. Maintenance managers’ report on terms such as labor cost, parts cost, headcounts and MTBF. When the metrics are combined there is a lack of alignment. When metrics are mismatched, the investment in systems, process improvements, and resources to support continuous improvement initiatives will typically be lessor then requested or inadequate. Without proper funding, we end up with poor results, which are then cited as justification for not increasing investment levels next cycle, a vicious circle...”  Corporate Reliability client, food processing.

Cultural transformation from reactive to proactive

PdM programs are typically doomed in a reactive organization. An organized strategy for transforming the culture is a key aspect of achieving results from the program.

Another factor contributing to program dissatisfaction may be that the site was not fully prepared to effectively implement a proactive workflow needed to support condition monitoring. The PdM program by itself it is not a “quick fix”. Each element of a PdM requires instrumentation, process changes and training. Proper staffing is not only required to implement the PdM but also the maintenance staff to address the issues found. At the macro level, Corporate Reliability clients tell us that a balanced investment in processes, systems and resources is a requirement for success.

Getting more productivity from headcount; not enough staff to manage the program or act on findings, analysis inaccuracies, stacking PdM on existing PM.

Reducing failure occurrence across the assets being monitored while shortening the meantime to repair (MTTR), increases maintenance availability and productivity.

Not “stacking” the new PdM activities on existing PM tasks is another area for productivity gain. In short, implementing Predictive Maintenance enables a shift from a fixed-schedule Preventive Maintenance approach to a more condition-based strategy, leading to a reduction in now unnecessary PM tasks

Next steps on the journey

Assuming the organization addresses the justification hurdle, which clears the way towards a Conditioned-based Maintenance (CbM) strategy, the next step is to decide which technology to adopt for the fault coverage needed. The choices of technology are as varied as the faults and require specific training to implement effectively. Infrared, Ultrasound, Motor Testing, Oil Analysis, and Vibration monitoring all have unique value propositions targeting different faults distributed along the P-F curve. Technology advancements and wireless sensing across some of these condition monitoring techniques have expanded and improved over the past two decades. Given the complexity of new technology, some education will be required.

Many established enterprises, considering the use of these technologies have sought advice to understand their own needs, educate themselves on the use and limitations of the technology, and compare options to engage vendors confidently. Before engaging vendors, it is advisable for reliability or maintenance teams to conduct a thorough self-assessment regarding the role new instrumentation will play in their overall program, how it will augment existing resources and competency levels. Questions such as what instrumentation is necessary, what to do with the additional data, where the data will reside, who will conduct the analysis and how the results will drive actions need to be considered.

Vibration monitoring is arguably the most common measurand for general asset condition monitoring, having the broadest range of fault coverage for rotating equipment. Wireless vibration sensors are a new technology that has introduced some subtle changes in measurement techniques, fidelity and features. For these reasons, this article will focus on the potential role wireless vibration sensors might serve in the maintenance toolkit.

Key promises of vibration monitoring for rotating equipment have long been:

• Determining overall asset health and lifespan: Controlling wear and tear allows for timely corrective actions, extending the lifespan of critical equipment.
• Increased maintenance safety: Identifying potential failures before they occur limits the potential safety hazards associated with sudden equipment breakdowns.
• Increased maintenance productivity/optimizing maintenance scheduling: Assists in managing or prioritizing resources, procuring parts, and scheduling staff actions by detecting specific maintenance needs rather than an exploratory approach.
• Identifying recurring faults to determine root causes: Historical data can be analyzed to show common failure patterns for failure elimination initiatives.
• Reduced maintenance costs: Early detection results in controlled repair actions needed to address the identified issue, saving on labor, parts, and material costs associated with major, unplanned breakdowns.
• Improved operational efficiency: Controlling unplanned downtime supports production assurance while maximizing output.

“My role involves using the results of Reliability Centered Maintenance (RCM) to determine what we can do to extend fault coverage, the instrumentation needed and how we will use it to predict failure. I’ll review PM’s, task lists and assess the right mix of skills and systems. Our smaller plants don’t have to resources to do any of this, so I set up the program, introduce the system to identify the faults at the appropriate place on the P-F curve, and train the resources in their use. I support implementation, but I’m not the reliability police” - Corporate Reliability client, manufacturing.

Quantifying the assertions made above, the chart in Figure 1 shows a breakdown of issues detected with vibration monitoring. Insights from industry suggest the following observations

• Periodic Vibration Monitoring – Detects ~70% of potential failures.
• Continuous Online Monitoring – Detects up to 90% of failures, especially critical assets.
• Reactive Maintenance (No Monitoring) – Typically results in unplanned downtime and catastrophic failures in over 50% of cases.

It is critically important to clearly define and articulate the problems, then set expectations regarding the solution that new systems represent.

• Setting realistic expectations among stakeholders regarding how an on-line vibration monitoring system will fit into and add value to the existing operation
• Defining clear objectives, goals, and metrics to help quantify improvements
• Perform an honest assessment of the current program, resources and capabilities, and identify where elements of the program are delivering value and where they fall short
• Identify which assets need enhanced or expanded condition monitoring, and how technology can strengthen the current condition monitoring model (potentially including faults requiring non-vibration sensors[1])
• A long-term plan for continuous improvement, incorporating both new tools and training to enhance both efficiency and effectiveness of existing resources

How do these systems help achieve more reliable operations?

Monitoring vibration of rotating assets provides early detection of potential mechanical, electrically induced, or performance related issues. The goals are to trend condition indicators, define normal ranges, alert when outside the norm, and identify the fault(s) through detailed analysis (human, AI, ML or combination). Detection, early in the failure cycle, allows the equipment owners to proactively maintain the asset with planned actions. Proactive maintenance controls unplanned downtime, extended equipment lifespan, reduces collateral damage and ultimately saves repair / replacement costs by preventing catastrophic failures before they occur.

If we’re convinced of the utility of these systems, now how will we use one most effectively?

To properly select the most appropriate Condition Monitoring and Predictive Maintenance techniques, it is important to understand the level of maturity of your Reliability and Maintenance program or in other words, “getting your organization ready”. A big-ticket item that should be on the organizational preparation checklist is determining how to use / integrate the information provided by the program effectively.

The figure above shows an example of the stages of maintenance and reliability maturity. Performing a self-assessment or having one done is a good benchmark for the organization and the areas to focus improvements or optimization.

For example, a Reactive organization may not have the ability to utilize the early warning fault information effectively. They may be more apt to respond with “How much longer can I run”, rather than proactively scheduling the work earlier in the failure cycle.  Some programs even hesitate to identify a developing fault in rolling-element bearings out of concern that, upon disassembly and inspection, there will be resistance or disagreement regarding the condition of the bearing.

It is also important to understand that increasing levels of vibration are doing damage to the machine and/or its components. Although predicting an incipient fault has a proven track record, predicting time to functional failure is less precise (PF curve). The severity of the damage over time depends on factors such as utilization, speed, load, and the type of fault. Repairs that are scheduled too close to the point of functional failure typically increase costs due to collateral damage.

Life at the top of the PF curve is a typical goal of most organizations. However, the reality is that some organizations prefer not to address PdM results that early. The more mature maintenance teams can manage failure data and the subsequent repairs later in the failure cycle, with actual timing of any repair taking into consideration production demands and the broader use of A-B (primary – spare) asset pairs.

To effectively manage PdM identified failures, progressive maintenance teams implement a thorough Work Management process (Planning and Scheduling). The process is designed to apply the right resources, at the right time, with the prescribed corrective action proportional to the asset’s severity. Managing the workflow in this manner is better informed by advanced vibration monitoring system and analytics. Assets with less criticality may be satisfied with more basic monitoring functionality and a limited set of analysis tools.

Below is a table that attempts to differentiate Advanced and Basic vibration monitoring programs based on the key decision-making factors:

Category	Basic Vibration Monitoring	Advanced Vibration Monitoring
What is the primary fault?	Determines if the machine is in good or bad condition.	Identifies specific failure modes (e.g., unbalance, misalignment, bearing wear, gear damage, resonance, electrical issues).
How severe is the fault?	Basic pass/fail assessment - is it running acceptably?	Provides severity assessment (minor, moderate, severe, critical).
What action should be taken?	Decides whether the machine can keep running or needs shutdown.	Recommends corrective or preventice action (alignment, lubrication, part replacement balancing).
Consequences of action?	Risk of unplanned failure but with minimal predictive details.	Predicts equipment failure risk, production downtime, safety hazards, and increased maintenance costs.
Special tools, skills, parts?	Usually does not specify; relies on basic maintenance decisions.	Determines if specialized tools (e.g., laser alignment, vibration analyzers), skilled personnel, or replacement parts are needed.
When is action needed?	Determines if immediate shutdown or scheduled overhaul is required.	Provides a timeline based on failure progression (immediate, scheduled maintenance, future observation).

Which assets to monitor

Primarily, there have been two methods for determining vibration PdM coverage. First, the more traditional approach would look at rotating assets in general or those who historically been problematic from a maintenance perspective. Second, considered a best practice approach, would be an asset criticality assessment, essentially ranking the most important assets to the least important in the facility.  Once a ranked list of assets is determined, a basic understanding of the typical failure modes for each and the resulting time to failure can be determined or estimated. Failure modes and time to failure can best be determined from experience or work order history. However, if a good failure history is not available, an estimate can be determined based upon the asset class, type of bearings, operating speed, and load.

With the ranked asset list in hand, the team can confidently determine which assets to prioritize for inclusion in the site’s PdM program. From there, a determination will be made for each asset whether vibration monitoring provides a cost-effective means to monitor the primary failure modes. At this point, the vibration monitoring asset list and scope has been developed.  

What techniques are best for effective monitoring

Vibration monitoring can be implemented with either hand-held manual collection or permanently mounted automated collection. The typical data collection frequency for hand-held manual collection is bi-weekly, monthly, bi-monthly or quarterly. For the non-critical asset, a monthly health check by a qualified professional may be the lowest cost option per measurement. For automated collection, the data collection can be continuous, with configurable intervals typically ranging from daily down to every few seconds.

In the past, installation cost that involved pulling wires, conduit and power to data acquisition devices which meant that continuous monitoring systems were used mainly to protect critical assets and turbomachinery. However, with the development of high data rate IoT protocols and radio technology transmitting data measured by low cost, low power accelerometers continuous monitoring has been ‘democratized’ and is no longer the domain of expensive safety systems. With the focus on managing costs of head count required to run a manual collection program, some program managers have opted for a mix of continuous monitoring with wireless sensors to help increase the productivity of existing resources.

Using automated vibration measurement collection and handheld manual tools strategically.

For most facilities, leveraging a combination of handheld vibration monitoring tools and machine-mounted vibration sensors is a cost-effective way to capture machine health and performance data.

For some assets, a handheld vibration monitoring tool and route-based inspections may be sufficient, if the resources are available. If a change or abnormality is detected with a handheld tool, further inspection and action can be planned. Handheld vibration measurement tools can also be used effectively in conjunction with wireless vibration sensors used for initial problem screening as a ‘check engine light’, triggering an alarm to notify maintenance of a potential issue, at which point a technician can take a portable data collector or other instrument to the asset to capture more data.

In a recent survey of maintenance and reliability professionals, respondents were asked questions regarding wireless sensors, such as the driving interest in these systems, what problems they expected to alleviate, who the influencers were and what features would be considered priorities.

For example, respondents see value in deploying wireless sensors where safety is a concern (assets difficult or unsafe to reach), to enhance existing resource productivity or to selectively monitor critical assets or bad actors more closely with more frequent measurements.

The focus on resource productivity is an industry wide concern as experienced practitioners approach retirement and cost pressures constrain investment.

The upside to this trend is that at least some diagnostic experience is being captured for posterity as the CAT 3/4's hired by wireless vendors train the AI algorithms embedded as part of the wireless systems diagnostics, in a quasi-supervised machine learning approach. This partnership of man and machine is assumed to improve diagnostic accuracy for many assets. Perhaps more importantly, automated diagnostics can enable both vendors and end users to scale their asset monitoring coverage with wireless sensors, freeing up skilled resources to focus on big problems and process improvement. This capability may become a necessity as wireless vibration monitoring can generate a large volume of data, making it counterproductive to gather data at frequent intervals for all assets, overloading fixed resources with work orders. Not every asset in the facility requires constant condition monitoring data to be effective, nor do all assets require monitoring on short time intervals during all phases of operation. Features of wireless systems such as measurement intervals that can be reconfigured remotely enhance the utility of these systems. That said, some vendors who champion this automated diagnostic capability seem to feel the need to reassure the user that manual expertise still touches every maintenance or anomaly report generated.

To better understand the role of technology in the maintenance landscape, perhaps it might be instructive to consider the analogy of big city hospitals. Should an IoT system serve as triage on a busy weekend, resulting in faster and more accurate disposition of an urgent case? Should the system expand the diagnostic capacity for over-worked residents? How does the system improve the metrics of existing operations with the current hospital staff headcount and roles, such as nurses who serve both as operations and maintenance? In an environment of fixed costs, what is the optimum ratio of systems to people?

Certainly, the important role of doctors and nurses does not go away, but judicious use of technology may reshape their priorities, expanding the ability to focus on the bigger picture of improving hospital operations while delivering value in terms of overall patient well-being.  Returning to industrial applications, a maintenance team struggling to put out fires daily may not be interested in having the system generate work orders unless they are prioritized and scheduled for the very last minute. Operations may be more interested in an estimate of how much longer they can run the asset, putting a premium on predictive capabilities. Both priorities will depend on the accuracy of the system diagnostics, given the costs to operations if the system provides a misleading answer. Reliability Engineering has the unenviable task of organizing all these factors into a PdM program that delivers better results with constrained investment in either people or systems.

Are wireless condition monitoring systems a game changer for factory operation that vendors sometimes claim? For now, we can conclude that wireless sensors systems offer considerable flexibility to maintenance operations. Judicious deployment of permanently mounted sensors with features such as measurement time intervals that can be adjusted with a mouse click will save the time of skilled resources. Effective and efficient data analysis that can be accomplished by any analyst with access to the dashboard means less travel time or even outsourcing to expertise in other locations.  Judging from the interest we see in the Reliability eco-system; it seems that these systems are destined to play an important role in today’s PdM programs.

We’ll dive a bit deeper into the specifications and features specific to wireless sensors in Part two of this series, highlighting differences to be aware of when compared to legacy sensors and data collection systems long used for route-based measurements.