<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by The Maintenance Guru on Medium]]></title>
        <description><![CDATA[Stories by The Maintenance Guru on Medium]]></description>
        <link>https://medium.com/@maintenanceguru?source=rss-53d6f9846bf7------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*ivU_JDEbz_i1WxWnKCHYaQ.png</url>
            <title>Stories by The Maintenance Guru on Medium</title>
            <link>https://medium.com/@maintenanceguru?source=rss-53d6f9846bf7------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 22:04:02 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@maintenanceguru/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Defect Elimination]]></title>
            <link>https://maintenanceguru.medium.com/defect-elimination-867df7bc557c?source=rss-53d6f9846bf7------2</link>
            <guid isPermaLink="false">https://medium.com/p/867df7bc557c</guid>
            <category><![CDATA[reliability-engineering]]></category>
            <category><![CDATA[fault-tolerance]]></category>
            <category><![CDATA[reliability]]></category>
            <category><![CDATA[failure-to-success]]></category>
            <category><![CDATA[root-cause-analysis]]></category>
            <dc:creator><![CDATA[The Maintenance Guru]]></dc:creator>
            <pubDate>Mon, 19 Jul 2021 09:49:10 GMT</pubDate>
            <atom:updated>2021-07-19T09:49:10.535Z</atom:updated>
            <content:encoded><![CDATA[<p>What is it and how can we start using it?</p><p>Defect elimination is an approach for improving reliability by eliminating common causes of failure. In the next pair of blogs we will cover two methods that could be employed in reaching this goal.</p><p>In an earlier blog, w<a href="https://medium.com/ox-mountain/functional-failure-the-primary-causes-670ede36fa9d">hich you can find here</a>, we discussed what the main influences of failure are with this diagram. This provides an important context for defect elimination.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/668/1*kSTu45rhX7fIYpWngFBfQw.png" /><figcaption>Influences on failure</figcaption></figure><p>There is emerging empirical evidence that in some organisations up to 60% of the root causes of failure are derived from the Operating context, whether this be:</p><p>· Maintenance induced failure</p><p>· Operations induced failure (operating beyond the equipment limits)</p><p>· Problems with interfaces to machinery (for example, low quality fuel or noisy electrical supplies)</p><p>The vital point to grasp is that failure causes rooted in mal-operations, maintenance inputs or outputs, or environmental issues, are within the purview of the operating company. They are best placed to identify and eliminate these issues.</p><p>The causes due to design and manufacture are usually relatively small as manufacturers would lose business if they did not focus on producing reliable equipment. Failures due to supply problems can be addressed using Warranty. If the operating environment is harsh and highly variable it can cause higher rates of failure, so this influence may be more significant. This doesn’t necessarily mean the solution is complex. For example, in a mining situation the quality of road surfaces can be actively maintained to reduce shaking and shock conditions to mobile assets leading to improved reliability.</p><blockquote>Defect elimination can be approached with two approaches split between a top-down general approach focused on quality, and a bottom-up data-centric approach based on identifying critical machinery, quantifying their reliability. Then both methods apply Root-Cause Analysis (RCA).</blockquote><p>This blog will cover the first, quality-oriented approach.</p><p>The quality approach is a good place to start if the organisation does not have a developed and mature data set, or analysis processes and reliability systems. All we need is to use the expert knowledge in our experienced front line maintainers. With this method we recognise the lack of quality of maintenance and operations as the largest cause of premature failure. We could aim to adopt total productive maintenance (TPM) or total quality maintenance (TQM) where we rely on delegating and empowering the frontline workforce (both maintainers and operators) to become responsible for continuous improvement. This is where they determine which problems are causing issues and decide how to solve them. The management and leadership functions support their teams with guidance and work to remove barriers for improvements.</p><p>If TQM is not employed, it should be possible to work with the front-line maintainers and determine which maintenance processes could be improved. Improvement projects could then be implemented. The roles of the managers or maintenance leaders should be conducting an analysis of the maintenance system to determine which working practices can be improved.</p><p>Metrics can be set that will determine the data that should be collected in order to show progress. There may be a delay of some months before seeing improvements in machinery reliability as there is a time lag between an improved maintenance or operating procedures and seeing better reliability. This time lag must be recognised, and improvements should be regarded as investment for the future.</p><p>Other initiatives such as adopting the 5S regime to improve cleanliness and tidiness in the workplace may also be initiated at the same time. Some workers may be trained in 5S or other simple problem solving and improvement processes so they can champion the change in working practices. These changes are driven by asking questions such as: are tools kept on shadow boards to check that they are not only available, but that they have not been inadvertently left inside machinery?</p><p>Questions to ask yourself in an analysis of the maintenance system, that constitute ‘vulnerabilities’ where maintenance induced failure could originate include:</p><p>1. What are your critical machines, which machines have the greatest impact on loss of production or product quality, and which take the most time and effort to recover?</p><p>2. How much maintenance is invasive? What maintenance requires machinery to be broken down and what containment boundaries (gas and fluid) need to be breached for maintenance? Is maintenance hygiene such as keeping dirt of foreign objects out of gas or fluid systems an area that could be improved?</p><p>3. In electronic systems, how many times are PCBs unnecessarily removed and replaced. Is anti-static hygiene consistently applied. Are PCBs swapped to find faults (willy-nilly board changing is bad practice as this may propagate faults to healthy PCBs)</p><p>4. What machines are critical in terms of alignment — decoupling and coupling of motors and driven equipment is critical. A slightly misaligned train of equipment can severely reduce life.</p><p>5. Are there issues with bedding machinery down and proper securing to bedplates that will expose the machinery to cyclic stresses? Could vibration surveys be done to check vibration as part of recommissioning machinery to capture and fix any issues?</p><p>6. How exposed are machines to environmental stressors? Could some simple modifications that reduces environmental factors help reduce their severity?</p><p>7. Are there maintenance jobs where spatial or other constraints such as ‘hard to maintain’ issues exist? Does the lack of accessibility increase the risk of breaches of quality?</p><p>8. Does replacing a part risk damaging mating surfaces, seals, alignment or securing, potentially lowering quality?</p><p>9. Are the right tools being used? Simple practices such as using mole-grips instead of wrenches should be discouraged. Are tools that need to be calibrated in date for their maintenance? Are simple processes such as the right order for torquing up a set of nuts or bolts regularly practiced?</p><p>10. Are maintenance processes captured, used and adhered to? Are these tasks captured in task lists that are issued with work-orders? Are there organisational cultural issues that work against standardised processes?</p><p>11. What is the state of training and experience of the workforce? Would they benefit from further training and qualifications?</p><p>12. Do you have to strip down other machinery (which may be invasive) to get adequate access to the component you need to work on or change? (In the Navy we called this “work in wake”)</p><p>This list is by no means exhaustive, but if the management draws up such lists, and with their front line they can determine which should be improved first. If we set priorities to eliminate quality causes of failure that are most common, then it is a great start on the road to defect elimination.</p><p>The changes using the quality approach should also start to look at how the right data should be collected and what simple analysis could be used to derive metrics to measure the improvement. Improvement in reliability is always a primary function of the maintenance team and will always align with the needs of an asset-rich business.</p><p>If we start using data driven metrics to measure improvement using a quality improvement approach, then the other data centric approach to defect elimination becomes more feasible. We will cover this in the next blog.</p><p>Have any of you employed a quality improvement approach or defect elimination? What questions did you ask yourself that are not on the list above. How did you decide what to improve? We would love to hear about your experiences.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=867df7bc557c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Maintenance Jack-knife chart]]></title>
            <link>https://maintenanceguru.medium.com/the-maintenance-jack-knife-chart-c14e2ade97d1?source=rss-53d6f9846bf7------2</link>
            <guid isPermaLink="false">https://medium.com/p/c14e2ade97d1</guid>
            <category><![CDATA[maintenance]]></category>
            <category><![CDATA[situational-awareness]]></category>
            <category><![CDATA[engineering]]></category>
            <category><![CDATA[data-visualization]]></category>
            <category><![CDATA[scatter-plots]]></category>
            <dc:creator><![CDATA[The Maintenance Guru]]></dc:creator>
            <pubDate>Mon, 17 May 2021 10:50:12 GMT</pubDate>
            <atom:updated>2021-05-17T10:50:12.062Z</atom:updated>
            <content:encoded><![CDATA[<p>A way to visualise asset criticality and help prioritise improvement</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*yPZHPn6C3PYf9P92" /><figcaption>Photo by <a href="https://unsplash.com/@kctinman?utm_source=medium&amp;utm_medium=referral">Shane McLendon</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Have you ever wanted a single chart that provided a quick way of seeing the main reliability problems for your fleet? Perhaps a Jack-knife chart will help.</p><p>Before presenting the Jack-Knife, it is worth stepping back and thinking about some fundamental principles of how information is shared and absorbed.</p><p>I am an ex-military maintainer. I believe you could walk into any military organisation’s technical office that uses equipment and see a whiteboard showing asset status. The whiteboard enables any person in the team to grasp situational awareness at-a-glance. Gaining rapid situational awareness is part of the military’s DNA, especially in combat situations. You can also observe similar ideas being used in hospital wards and workplaces with Kaizen or Kanban boards. The hospital shows all the beds, occupancies and essential information. whiteboards are updated in real-time by the staff. They have the attributes of being instantly accessible, easy to change and read, resulting in a very simple but very effective means for imparting information.</p><p>We can discuss other ways situational awareness can be enhanced in the maintenance domain in a future blog.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/850/1*ScY7ik5gw6Ij5v6TMlFe2A.png" /><figcaption>A hospital ward white-board</figcaption></figure><blockquote>How can this idea of “grasping situational awareness at a glance” be extended in the reliability domain?</blockquote><p>The jack-knife chart may be a candidate. This is a type of scatter chart where the x and y-axis represent two important dimensions of criticality for an asset’s components. The x-axis shows the likelihood or frequency of failure, and the y-axis shows the impact or consequences of failure. Each component’s criticality is represented as a plotted data point on the chart. Usually, the impact axis is the ‘time to repair’ or downtime of the failed assets. The diagonal lines represent high or low criticality borders. Any component plotted in the upper right quadrant is bad. Borders may show limits of acceptability for an organisation, or for standards or legislation. The horizontal lines show borders between high or low acuteness or frequency. The axis’ are also logarithmic enabling a wider set of data to be fitted. Here is a real example:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/728/1*B921brAqWRG-G_hTUvqEJg.jpeg" /><figcaption>An industry example of Jack-Knife</figcaption></figure><p>We can build on the baseline Jack Knife chart to include some other display options. The chart shown below is the equivalent of the chart above. With most modern charting tools it is possible to label data points with other data, or if the charts are interactive labels can be seen using mouseover. A couple of labels are showing part descriptions as an example of what may be possible.</p><p>An asset may have hundreds or even thousands of functionally significant components. We should be able to plot everything and use logarithmic scales to fit everything in a single chart. This detailed chart provides us with the overall picture of the asset. The clustering of the data may be compared between similar assets. However, we should also be able to filter down to the topmost critical components to focus where we want improvement efforts to be concentrated. The example we have used only shows a limited set of parts for illustrative purposes.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/900/1*0r_zdD0IMjSsZVXkeDiMsw.png" /><figcaption>Basic Jack-knife showing component labels</figcaption></figure><p>The conventional Jack Knife shown above could be extended with other dimensions or attributes of criticality. For example, in a maintenance Failure Modes and Effects Analysis (FMEA), the detectability of a failure is often included as a dimension of the RPN Risk Profile Number (RPN) to help quantise criticality. Detectability refers to how evident a component failure is.</p><p>If a failure occurs suddenly with no symptoms or other indications, the detectability is low. Low detectability is more severe than a failure that is gradual with many symptoms and indications. This is even better when symptoms present in a predictable sequence over the degradation time.</p><p>In a scatter chart we can show the third dimension of detectability using colour coding or the size of the individual markers in the chart. This could use a simple classification of detectability into three (or more) sections, to keep the visualisation simple. In the example below, we show a simple example with size designating detectability criticality. The chart legend shows the three classes of detectability using the same components as the basic chart above.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*SUjpjQxHZkMb_nULkBXXJQ.png" /><figcaption>Jack-knife showing detectability criticality using plotted point size</figcaption></figure><p>Other attributes of criticality can be mixed with either of the frequency or impact dimensions.</p><ul><li>We could use the B-20 Weibull measure using the inverse of B-20 Age for the frequency of failure. In this previous blog, we discussed how B-20 may be a more informative reliability metric than MTBF or an averaged failure rate.</li><li>We might include attributes such as Safety or operational factors, material, labour and logistics costs, as well as factors for logistic delay or lead time to the impact score.</li></ul><p>Other indicators that may be toggled on and off using colour coding maybe</p><ul><li>Colour coding based on Weibull Shape. Parts with failure patterns that are premature, random or aged. It is probably that using B-20 as the major frequency dimension score the premature failure pattern will score highest. Premature failures are predominantly caused by low-quality issues in operations and maintenance. These causes should be preventable by the operating company.</li><li>Part obsolescence risks may be included, where the supply may become restricted as parts may no longer be available.</li></ul><p>Themchart below shows a combination of detectability as the size of the plotted point combined with the colour coding for the pattern of failure, derived from the Weibull shape parameter.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UcO9yV7_UFuN07Aq4jPSFA.png" /><figcaption>Jack-knife combining size and colour for criticality and patterns of failure</figcaption></figure><p>Organisations may have other attributes of equipment criticality they may want to include, with weighting factors if they consider some attributes are more important in their context than others.</p><p>The Jack-Knife chart may also have a strategy value; it can be used help maintenance align with business strategy. This is illustrated on the chart below with the thick black arrows for improving cost or productivity.</p><ul><li>If there is strong market demand from the organisation’s output and / or commodity prices are high, then the emphasis may be put on improving reliability and availability of the assets downwards on the Y axis. This means primarily focusing on decreasing the likelihood or frequency of failure of the most critical parts. Eliminating defects and premature failure will pay the highest beneficial returns.</li><li>If the market takes a down-turn and profits are reduced then cost cutting may be the primary focus, whilst maintaining quality. This means focusing at reducing the highest scoring components on the impact dimension along the x axis.</li></ul><p>Improving detectability may be constrained by the physics of the failure modes, or the ability to monitor the equipment (either with people or sensing systems). For components that fail hidden, then doing the right maintenance involves implementing failure-finding tasks. The timeliness of these tasks is essential when they are applied to safety and protection systems. Failures may become more apparent with improving operator and maintainer awareness, and predictive maintenance may play a part where existing sensors may be able to provide symptoms of some of the failures with low detectability. Staff awareness and knowledge of what should be reported in notifications or work-order reporting is a subject we will cover in a future blog.</p><p>Within the previous three blogs looking at maintenance metrics we have looked at timeliness of scheduled maintenance and ensuring that we not only do the right maintenance, but we do maintenance right.</p><p>Given we have these other metrics, it should be possible to calculate what the improvement opportunities are. Specifically, these include eliminating premature failure, doing the right maintenance and improving the planned to corrective maintenance ratio. We also saw measures such as scheduling timeliness and adherence to estimates. These can be quantified, and improvement opportunity lines can be plotted on the chart and ranked in data tables. For senior management an aggregated top line improvement opportunity could be shown on a dashboard.</p><p>The chart below shows an improvement opportunity for a component with the green arrowed line, which would represent the savings made by eliminating premature failure from this component. The head of the arrow shows where the current plotted position could be shifted if premature failure could be prevented. The chart could be arranged to show all or a number of the component improvement arrows.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*GULW1_H4wNGRmP3FVNvirw.png" /><figcaption>Jack-knife showing support for business strategy and component improvement opportunities</figcaption></figure><p>It would be powerful if the Jack-Knife chart could be interactive, adding or subtracting different attributes from the three dimensions, and toggling flags on and off.</p><p>If the chart was associated with a data grid or table that was able to rank the parts for criticality that was linked to the Jack-Knife or showed the parts that were sorted by improvement opportunity, it would provide the future improvement worklist for the maintenance and reliability department.</p><p>We have discussed the concept of situational awareness and taking the examples of whiteboards as examples of how staff can use them to ‘grasp situational awareness at a glance’. The jack-knife may be a candidate for exploring asset components and their criticality visually. If the charts are made interactive allowing different factors to be toggled, they enable a powerful means of prioritising maintenance improvement work. Have you ever used a Jack Knife or similar plot to help situational awareness or identify priorities for improvement work? We would love to hear your experiences in the comments.</p><p>In the next blog we will explore defect elimination, which is sometimes called proactive maintenance. This is where common causes of failure for many components may be prevented by improving process quality. This approach to maintenance can also be linked back to the Jack-Knife diagram.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c14e2ade97d1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Maintenance Metrics]]></title>
            <link>https://medium.com/ox-mountain/maintenance-metrics-5c0f012c02e4?source=rss-53d6f9846bf7------2</link>
            <guid isPermaLink="false">https://medium.com/p/5c0f012c02e4</guid>
            <category><![CDATA[maintenance]]></category>
            <category><![CDATA[classification]]></category>
            <category><![CDATA[engineering]]></category>
            <category><![CDATA[metrics-and-analytics]]></category>
            <category><![CDATA[planning]]></category>
            <dc:creator><![CDATA[The Maintenance Guru]]></dc:creator>
            <pubDate>Sat, 10 Apr 2021 09:50:08 GMT</pubDate>
            <atom:updated>2021-05-04T07:32:41.003Z</atom:updated>
            <content:encoded><![CDATA[<h3>Part 3, Tying the task metrics to a higher level</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*N50NLohtngX03qhl" /><figcaption>Photo by <a href="https://unsplash.com/@lukechesser?utm_source=medium&amp;utm_medium=referral">Luke Chesser</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>In the previous two blogs, we have defined low-level metrics for maintenance tasks. One of the critical success factors in deploying metrics is having a manageable number to set as organisational goals, to avoid being unfocused or overwhelmed. The metrics we have defined need to be part of a system that feeds higher-level metrics or performance indicators. This blog will discuss higher-level metrics and show how the task metrics can be linked.</p><p>There are several commonly used high-level maintenance metrics used in industry</p><ul><li>Safety Compliance: the number of accidents, incidents or near misses, number of days off work injured.</li><li>The percentage difference between planned maintenance and unplanned corrective maintenance</li><li>Maintenance compliance: The timeliness of planned maintenance, which may include some measure of deferred maintenance</li><li>A measure of reliability and maintainability: Often these may include MTBF and MTTR, but there are disadvantages to these measurements. We will question the use of mean values.</li><li>Measures of OEE Operational Equipment Effectiveness: OEE is the product between availability, equipment performance and product quality. Product quality is easy to define in a manufacturing situation, but even in mining mixing different grades of ore, or crushing to a required specification is often done.</li><li>Measures of maintenance resource utilisation and conformance to budgets</li></ul><p>What is missing from this list are measures for:</p><ul><li>Is the right maintenance is being done (as opposed to whether the maintenance is done right)?</li><li>Any team metrics, for example, does the team have the right mix of skills and experience, and how is the team developed?</li><li>How complete is the master data for maintenance? This has a direct impact on whether the right maintenance is being done.</li></ul><p>We can show how we should approach calculating these higher-level metrics</p><h3>Safety or SHEL Compliance</h3><p>SHEL stands for Safety, Health, Environmental &amp; Legislative compliance. These are all factors that an organisation cannot afford to make mistakes because the ultimate consequences are existential.</p><p>Safety is mandatory and non-negotiable and should be considered from two broad perspectives:</p><ul><li><strong>Operational safety</strong> where day to day working safety is considered. Common metrics include recording and trending the number of reported accidents, incidents or near misses, lost time due to accidents and injury incidence rates</li><li><strong>Functional safety</strong>, where there needs to be an understanding of what assumptions and possibly maintenance are linked to the intrinsic safety of an asset or a machine. For example with rotating machinery, how Overspeed checks must be done, for pressure vessels how relief safety systems must be periodically tested, fire detection and suppression systems may have statutory tests. The maintenance master data must contain all of this data, and the associated equipment tagged to prevent things like changing test periodicities, which may be mandated by law, standards or organisational requirements. Maintenance managers must also be aware of safety implicated equipment Planned Maintenance must not be deferred and needs to be completed on time.</li></ul><p>Besides health and safety, the environment and legislative compliance should also be included. Pollution and emissions have growing statutory requirements that must be complied with, and some emissions reporting is mandatory in some countries. Maintenance has a huge</p><p>Safety culture is based on a willingness to accept mistakes as learning opportunities where any person can be confident in reporting their own without being denigrated and instantly punished. Obviously, gross negligence or deliberate acts must have disciplinary consequences, but the ethos needs to be on learning from mistakes, instead of energy being expended in covering them up.</p><p>Safety training and awareness for staff should be measured and tracked as well as attendance at safety briefs where previous accidents incidents and near misses are analysed and discussed, so lessons can be drawn.</p><h3>How complete is the master data?</h3><p>We hinted in a previous blog that most CMMS systems are not designed to contain much of the data we require to manage maintenance in a principled and scientific manner. The CMMS holds</p><ul><li>Equipment registers to identify equipment and their physical position</li><li>maintenance tasks and their scheduling data</li><li>Plans for maintenance</li><li>Estimates for levels of effort, material and times taken for doing maintenance</li><li>A history of when maintenance was done, and how long it took and what resources were necessary</li><li>Budgetary and cost data.</li></ul><p>We need other data including</p><ul><li>Machinery criticality, several factors may be taken into account, but the major factors for maintenance are the likelihood or frequency of failure, impact or consequences of failure and detectability of the failures occurring. These three basic factors are calculated in an FMEA RPN to determine the criticality of possible failures of components.</li><li>A justification for the maintenance tasks we do. The data recorded in FMEAs and RCM along with RCM age analysis provides this justification</li><li>Root cause analysis data, which also forms the basis of defect elimination</li><li>Documents and drawing data</li><li>Data used in and derived from On-condition maintenance or Predictive maintenance</li><li>Failure recording, including how failure presents, sudden, intermittent or gradual. What if any symptoms are available?</li><li>Reliability, availability and maintainability data. Using several methods including Weibull, FMEA</li><li>Effects and consequences of failure (FMEA data). A scale showing SHEL (Safety Health, Environmental &amp; legislative), Operational, economic or low significance should be classified for each functional failure</li></ul><p>In a previous blog, we covered the possible use of graph type data management solutions that suit the nature of data necessary for maintenance, because the data is rich in relationships.</p><p>A mapping between maintenance type tasks in existence or planned, and what data is necessary to justify that task is necessary, the completeness of the data could be given a simple scale of completeness</p><h3>Is the right maintenance being done?</h3><p>We have previously shown that the shape parameter in the Weibull analysis is a great indicator to help us decide what maintenance tasks are applicable. The Table is reproduced here</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rQ_0EyieNdaHfeA5NCPMew.png" /><figcaption>Failure patterns define Weibull shape and the applicable maintenance strategies</figcaption></figure><p>Besides the Weibull shape determinant, further prerequisites are necessary to apply maintenance task types as follows:</p><ul><li><strong>On condition.</strong></li><li>Is there a time delay between the inception of failure and functional failure?</li><li>Is the detection of the failure possible, leaving sufficient time (P-F interval) to avoid failure consequences, ideally to plan and predispose resources for recovery?</li><li>Is the variance in P-F interval small enough to be practical?</li><li>Is the task practical and cost-effective? Cost-effectiveness is a balance between the costs of failure, versus the costs of inspection, sampling or processing data? The frequency of inspection is a fraction of the P-F interval. The downside costs of failure are determined by the frequency of failure.</li><li><strong>Scheduled replacement or restoration</strong></li><li>Does the failure have a wear-out failure pattern?</li><li>Is the task cost-effective? The costs of avoided failure are a combination of frequency of failure and impact of failure. The cost of conducting maintenance is the planned change costs, plus if the task is a scheduled replacement, the discarded remaining life of the part. The formula for calculating this may be: 1 — (Weibull scale age / the part’s age at change) * the material cost of the part</li></ul><p>The metric is to determine the percentage of active tasks that have the appropriate technical justifications and evidence present, along with the calculations that determine the currently used periodicity.</p><h3>Is the maintenance being done right? (maintenance compliance)</h3><p>In previous blogs in this thread, we have discussed the costs of timeliness of the scheduled maintenance. For scheduled replacement or restoration, we used the concept of the timeliness waste matrix.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Rtie7xpkYwoAUwpNQnHbeg.png" /><figcaption>Scheduled PM timeliness waste matrix</figcaption></figure><p>For on-condition, we used a confusion matrix for the accuracy of diagnostics and looked at a cone of accuracy over the prognostic remaining useful life. Timeliness of the on-condition maintenance is also important, if inspections or samples are done late (periodicities &gt; ½ the P-F interval) then detection may be too late to plan recovery or maybe missed entirely. This incurs the costs of an avoidable unplanned failure with the associated disruption.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cbWQ2yVqRJ2GAZi4cnUPgg.png" /><figcaption>Systems for on-condition diagnostics and prognostics</figcaption></figure><p>All of these metrics can be arithmetically combined and supplied to a higher level for maintenance compliance</p><h3>Reliability and maintainability</h3><p>Many maintenance systems keep trends of MTBF and MTTR to measure reliability and maintainability trends.</p><p>The Mean Time Before Failure is the average of the times to failure of a repairable component or machine. By repairable we can replace a machine’s broken part with a new one, and we assume the newly fitted component has the same intrinsic reliability as the previous component. The Mean Time To Repair is the average of all the times taken to repair or replace the non-functional part.</p><p>These measures have the advantages of being relatively simple to calculate and are universally well known. However, there are disadvantages and limitations that all maintenance and reliability professional should be aware of.</p><h3>Problems with using mean or average measures</h3><p>Both MTBF and MTTR metrics use the mean or average of the time data they collect, but we need to ask ourselves is using the mean always the most appropriate measure for the ‘middle tendency’ or ‘expectation’ for sets of numeric or continuous data? For example, why do government authorities worldwide use the median, instead of the mean, when presenting wages data?</p><p>The answer is that the mean measure is sensitive to outliers and is not as robust where the data set is skewed (not uniformly distributed, or the distribution is not symmetrical about the mean). Comparatively, the Median measure is less sensitive to outliers and more robust in these circumstances. If we trend an MTBF we need to include a window over several previous readings to calculate the average. If the underlying reliability changes and the time to failures are different from the most recent to the previous set of data, then the trend has a lag before settling down to the new truer reading. An exponentially weighted moving average could be used to compensate for this, but this complicates the calculation and does not entirely solve the lag problem. I prefer to use the Crow-AMSAA reliability growth chart, which shows a log scale of the cumulative trend of the age against the log of the event count. An increasing slope of the trend for the latest events compared to the historical events shows worsening reliability (as shown in the figure below). If the slope gets shallower it indicates reliability is improving. I believe we need to wait for at least three consecutive similar events or data points need to be recorded before we can have adequate certainty that a new trend is established. This is the quickest way to observe a new trend compared with the lag involved with a moving window average calculation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_wY1vOIIPs0pGo7vLAZ3FA.png" /><figcaption>How to use Crow-AMSAA to determine reliability trends</figcaption></figure><blockquote>Some people believe that the MTBF is the age that a component will last, this is not the case.</blockquote><p>If we are given an MTBF without any other information, we would need to assume a constant failure rate for the times to failure. This does not align data with the normal (or gaussian) distribution; it aligns with the exponential distribution where the meantime to failure equals the point at which 63% of a population of parts will likely have failed.</p><p>Because given a raw MTBF tells us nothing about the variance of the times to failure, the Maintenance Guru prefers to use Weibull analysis or Crow-AMSAA. We have discussed the use of Weibull in a previous blog here. A more meaningful measure of reliability for the working maintenance or reliability engineer would be to use the B-20 likelihood of failure derived from the Weibull Cumulative Distribution Function illustrated in the figure below. The B-20 measure is the time to failure where 20% of the population of the parts are likely to have failed. The B-20 is also useful because if the shape characteristic of the Weibull is small &lt; 1, then the reading accentuates how much worse premature failure is compared with a wear-out failure (Weibull shape &gt; 2.5). This measure then helps the reliability engineer focus more on eliminating the causes of premature failure.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*sAYUCZ2UHZkPEqWIrsepig.png" /><figcaption>Using B-20 instead of MTBF is more informative</figcaption></figure><blockquote>Einstein said “we should strive for the simplest solution, but no simpler”. The maintenance Guru believes MTBF is an oversimplification for reliability engineering.</blockquote><p>The blog would recommend that MTBF or MTTR may be retained if changes to more meaningful metrics may be too difficult to persuade others who expect them, but that specialist maintenance and reliability folk use Weibull and B-20. As explained in previous blogs, the MTBF should never be used to calculate the periodicities of scheduled tasks. Using the Median measurement for MTTR will reduce the impact of outliers and provide a better measure for estimating future maintenance work durations for planning and budgeting.</p><h3>Operational Equipment Effectiveness (OEE)</h3><p>OEE is a metric made up of the product of availability, performance and quality. Availability is the measure of time that an asset can be profitably utilised, minus planned outage time against that time when the asset is unavailable for productive work for other reasons. Other reasons may include:</p><ul><li>Avoidable delays in the operational cycle</li><li>Unavailability due to unplanned failure (Failures that have an operational effect)</li></ul><p>Not all assets can be run profitably 24/7, consider the profitability whole fleet of busses running at 4.00 am, but a small minority of the fleet will still run. The definition of the baseline availability therefore may need to be carefully defined. It may be a case that the total scheduled running hours of profitable utilisation should be used against the actual number of hours of profitable hours achieved is used.</p><p>Availability is also impacted by inefficiencies in unplanned failure recovery. The availability of resources and logistics delay times may need to be broken down and recorded so that analysis and efficiency improvements may be conducted later.</p><p>Performance can be defined by the capacity to do useful work. Capacity may be reduced by the inefficiency of machinery or ageing effects that reduce performance. Predictive maintenance is often applied to measuring machinery performance and efficiency. Fuel or energy costs per unit of output is an effective way of measuring machinery efficiency. Efficiency can also be restored by scheduled restoration tasks.</p><p>Quality is the easiest to think about for a production line. Any product that is out of specification and rejected (possibly by a customer) is a low-quality issue that means the work completed and value-added in production is wasted. If a customer rejects a product because of low quality, it may involve loss of reputation and avoidable costs due to compensation.</p><p>In some situations, the quality of the product may be difficult to define. In mining, the quality of the ore may be specified and mixing higher grade ore with lower grades may be necessary to meet customer specifications. In freight transport, it may equate to the condition of whatever is being carried at its destination after being exposed to the environment</p><h3>Maintenance resource utilisation, planning and budgets</h3><p>The maintenance team utilisation may be split between reactive corrective and planned preventative work. The aim should be to increase planned preventative work which should decrease the overall maintenance spend.</p><p>Over time we are looking to achieve a high utilisation rate with the maintenance team, with as little variance as possible whilst achieving the required safety.</p><p>Overtime is a means of dealing with unexpected demands, or maintenance may be subcontracted, especially for major machinery outages such as a machine mid-life update. Overreliance on overtime indicates workload may be too high, or estimations for work planning may be too low. Overtime should be recorded and trended.</p><p>In a mature maintenance management system, we should be able to exploit the statistics derived from the Weibull analysis, providing the data for the probability of failure over the age of the components. This is input into RAM (Reliability, Availability, Maintainability) simulation. This is where a discrete-event-monte-carlo simulation can be used to forecast future failure and maintenance events. If the simulated events are extracted costs can be assigned as a Life-Cycle cost capability. Using simulations is far more accurate than assuming fixed failure rates used in a spreadsheet. We can discuss the use of RAM and LCC simulations in a future blog.</p><p>Another output of the RAM modelling is to provide a forecast for maintenance resources over time. If peaks and troughs are indicated in the simulation forecast proactive actions by the planners can be taken to bring forward, stagger or defer work to flatten the demand out. Actual demand can be trended with forecast demand, and any deviations can trigger analysis and investigation.</p><p>The use of simulations does not provide a crystal ball to accurately predict that a single component is going to fail (with high certainty) on a certain date.</p><blockquote>All models are wrong, but some are useful.</blockquote><p>The RAM and LCC simulations should be used to determine likely demand for whole populations of components or assets, split into time buckets (monthly) for the likely number of events in those periods. The more granular the population of components and the smaller the time slices, the more inaccurate will be the forecast. A simple way to find a good trade-off is to simulate the previous year and see how accurate the predictions are to the actual events, although past performance is not a guarantee of future accuracy.</p><p>The thinking and use of the simulations have emergent benefits, as it enables sensitivity analysis to be conducted to determine priorities for improvement, conduct what-if scenarios and the process of running the simulation capability increases knowledge of how the operations and maintenance systems work and interact.</p><p>The planning and LCC simulations can be used to help set the maintenance budgets and then track spend against the plan and the forecast. Deviation could trigger investigative action to see what is underlying the deviation. The organisation will likely set a tolerance that the maintenance manager could work inside.</p><h3>Spare Parts and inventory management</h3><p>Spares management and logistics is a separate discipline to maintenance management, but there is a close dependency and reliance. Common metrics for spares and inventory management include:</p><ul><li>trending any stockouts and resultant delays this causes</li><li>Inventory record accuracy</li><li>Turnover times, to minimise sunk capital invested in static inventory</li><li>Managing lead times for supply, especially when this is associated with planning horizons and the P-F intervals in on-condition maintenance.</li></ul><p>In this blog we have presented several commonly used maintenance metrics, highlighted problems and improvements that may be made suggested some extra metrics and shown how to combine the lower-level maintenance-task metrics to higher levels. We have again shown the importance of having master data and as we have added further detail to the data, it should reinforce the desirability to use graph-based data management systems we covered in this blog.</p><p>I would like to hear about your experiences with metrics, anything novel and valuable you have experienced and used. What other metrics do you use in your organisation?</p><p>In the next blog, we will cover a useful chart type called the Jack-knife diagram that plots a scatter chart of an asset’s components based on FMEA RPN or risk criteria. This chart may be used to help prioritise improvement work and complements the metrics we have covered here.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5c0f012c02e4" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ox-mountain/maintenance-metrics-5c0f012c02e4">Maintenance Metrics</a> was originally published in <a href="https://medium.com/ox-mountain">The Maintenance Guru</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Maintenance Metrics]]></title>
            <link>https://medium.com/ox-mountain/maintenance-metrics-bacdde9aac55?source=rss-53d6f9846bf7------2</link>
            <guid isPermaLink="false">https://medium.com/p/bacdde9aac55</guid>
            <category><![CDATA[prognosis]]></category>
            <category><![CDATA[diagnosis]]></category>
            <category><![CDATA[predictive-maintenance]]></category>
            <category><![CDATA[predictive-analytics]]></category>
            <category><![CDATA[engineering]]></category>
            <dc:creator><![CDATA[The Maintenance Guru]]></dc:creator>
            <pubDate>Fri, 26 Mar 2021 07:59:47 GMT</pubDate>
            <atom:updated>2021-04-03T10:03:09.345Z</atom:updated>
            <content:encoded><![CDATA[<h3>Part 2. The underlying metrics for On-condition and failure finding maintenance tasks</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*CGQkdaohuUhOxq4J" /><figcaption>Photo by <a href="https://unsplash.com/@frieauffcom?utm_source=medium&amp;utm_medium=referral">Markus Frieauff</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Metrics are measurements that help us manage a system. Metrics should be understandable, and business-led. We should also recognise metrics influences the behaviour of our people and lastly, too many metrics may blur our focus. The conclusion to be drawn from this is that metrics need to be selected very carefully.</p><p>This blog is the second in a thread that will discuss low-level metrics to ensure that the applicability, effectiveness and efficiency of maintenance is measured and reported.</p><p>The taxonomy of maintenance we discussed in a previous blog identifies on-condition and failure finding as types of task done as preventative maintenance. We have also described on-condition maintenance in some detail <a href="https://medium.com/ox-mountain/why-are-there-so-many-inspections-with-on-condition-maintenance-76a34fc0f417">here</a>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lTWOHm_3t7WlAlyS-ckKAQ.png" /><figcaption>Taxonomy of maintenance task types</figcaption></figure><h3>On-condition tasks</h3><p>There are two broad measures of effectiveness for on-condition maintenance tasks. The first measure is associated with diagnosing that a failure exists and the second is success in prognostics.</p><p>We split on-condition task types into two models in a previous <a href="https://medium.com/ox-mountain/why-are-there-so-many-inspections-with-on-condition-maintenance-76a34fc0f417">blog</a>, the first model is <strong>long term deterioration (LTD)</strong>, and the second is <strong>incipient failure detection (IFD)</strong>. Diagnostics has a looser association with LTD; if we are able to measure the condition at any time in the deterioration lifecycle, there is no point of diagnosis in this case and we apply prognostics. For LTD where it is only possible to measure condition beginning at some point in the deterioration lifecycle, we can say the diagnosis is applicable. Diagnosis is necessary for IFD to determine a component is failing.</p><h3>Diagnostics</h3><p>Taking diagnostics first. Diagnostics is the activity that determines whether we are in a state of failure, where the consequences are so far minimal. Diagnostics has a set of degrees of success or accuracy, from where we can recognise something unusual is happening to recognise our asset is definitely failing. We introduce a set of keywords to classify the degrees of diagnostic certainty ranging from:</p><ul><li>Detecting a <strong>novelty</strong>. Novel behaviour that may or may not be symptomatic of failure. Novel behaviour may be indicative of a region of normal behaviour that is not known. This implies we have knowledge of what we regard as normal and abnormal asset behaviour.</li><li>An <strong>anomaly</strong> that is recognised to be non-normal behaviour (a deviation from normal), which may be indicative of failure, but as yet is unknown.</li><li>A <strong>symptom</strong> is a deviation from normal, which is a known indicator of failure. In some on-condition systems, a symptom is known as a <strong>condition indicator</strong> or <strong>CI</strong></li><li>A <strong>diagnosis</strong> that is made up of one or more symptoms. A diagnosis is indicative, with a high degree of confidence, of a failure mode on a component.</li></ul><p>The goals of on-condition diagnosis are to:</p><ul><li>Detect the incipient failure, ideally isolated to a failure mode and a component(s) that are damaged by the failure mode (that may require repair or replacement).</li><li>To detect the incipient failure within sufficient time to allow planning and predisposal of resources to gracefully withdraw the affected machine from service and recover it quickly and efficiently. This should minimise operational disruption.</li></ul><p>Taking each goal and breaking them down further</p><p>We need to understand our requirements for isolating the fault condition. For example, what are our levels of line replacement for components? Do we replace a whole electric motor as a line replaceable unit, or would we replace the bearings on the motor in situ? The on-condition fault isolation may only be required at the whole motor level for the first situation or fault isolation is required at the motor bearing level if we had to recover the motor bearings. Isolating the failure mode helps inform us what type of damage is occurring and helps us to determine what spares may be necessary. This ‘level of recovery’ information needs to be recorded in the maintenance master data linked to the associated on-condition task. It is desirable to develop diagnostics so we can trust it enough to avoid having to manually troubleshoot or confirm the diagnosis.</p><p>The second goal makes us consider what the required recovery lead times may be for planning and predisposition for recovery. If we are to gain maximum benefit from on-condition maintenance, then the diagnosis ideally needs to detect the onset of failure before that lead time.</p><p>The degree of system conformance with these universal goals should form metrics for on-condition tasks.</p><p>Not all on-condition systems are optimal, partial value from on-condition may be realised by late detection of an anomaly and shutting down a machine to avoid the end point of failure thereby avoiding secondary induced damage. This situation is better than allowing the machine to run to final failure, partial value is achieved.</p><p>Looking at how we have broken down on-condition to varying degrees of accuracy:</p><p>If our system detects a novelty, we need action to confirm that this is either an unrecognised part of normal behaviour, or whether it is possibly an anomaly. Our action would be to recognise a novel area of benign normal operation or classify an anomaly.</p><p>If the detection is a known anomaly, then we may need to initiate further action to task a maintainer to troubleshoot the affected machinery to determine what may be going wrong. The resulting information that the troubleshooting may find needs to be captured and recorded in the maintenance master data about that machine, so improvements in the on-condition diagnosis may be made. In this case, an anomaly may be promoted to a symptom or Condition Indicator. The difference between a symptom and an anomaly is that we are sure the symptom is associated with known failure mechanisms or failure modes, where an anomaly is not.</p><p>If the detection is a symptom, we may still need a trouble-shooter to manually determine the condition of the affected part, or we may have sufficient confidence in the symptom to start to plan recovery. If we consistently trigger corrective work from single symptoms (effectively treating a single symptom as a diagnosis), we may suffer from a situation known as a ‘false positive’ A false positive is where we think we are detecting a failure, but none exists on the machine. This may result in taking nugatory action and possibly taking a machine offline, which is wasteful. We will discuss what we mean as a ‘false positive’ below when we cover the confusion matrix.</p><p>If we have a diagnosis, which may occur after a trouble-shooter has confirmed an anomaly or single symptom is a failure, or we trust a single symptom, or we trust multiple symptoms are a failure, then we should have a lot more confidence that we are indeed suffering from a failure and act accordingly. Symptoms may also present themselves at different times in the failure cycle, which helps us to quantify prognostics and the time to functional failure.</p><p>If we generalise, essentially Diagnosis is a type of classification problem. Diagnosis asks “are we in the normal or failing state?” The best way of measuring a classifier is with a confusion matrix. The aim of the system is to correctly classify the onset of failure, avoiding misclassifications.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/931/1*ypHOx_ZYPAFq_TL-Xb3WPA.png" /><figcaption>Confusion matrix to measure the performance of a binary classifier</figcaption></figure><p>The semantics associated with the confusion matrix can take a few minutes to understand especially when our classifier is positive in its ability to detect failure. Usually, we naturally regard a failure as a negative, but not in this case with a confusion matrix</p><p>Where on-condition involves manual inspections with expert and qualified inspectors, the chances of false positives is generally low. The confusion matrix is probably more useful in predictive maintenance where diagnostics and prognostics is data-centric and has more automation. It is more likely misclassification occurs in a predictive maintenance system.</p><p>Metrics can measure the occasions the classifier triggers, and whether this is a true or false detection, and the occasions where the classifier misses a failure (False negative). In the matrix above green are desirable results and red are not.</p><p>It is also advantageous to think in terms of specificity and sensitivity as you may tune a diagnostic classifier. Tuning is possible in a predictive maintenance situation (where sensor data is being used) for increased sensitivity to reduce the numbers of false negatives or tune to higher specificity to lower the number of false positives. In a system where failure consequences are high, you may tune to be more sensitive and accept more false positives.</p><h3>Prognostics</h3><p>Beyond diagnostics, we need to address prognostics. Prognostics might not have to follow on from diagnosis if we are dealing with LTD where we can measure condition at any time in the deterioration lifecycle. The goals of prognostics are to:</p><p>· Provide an accurate prediction of the remaining useful life of a failing component</p><p>· Provide a forecast of the likely impact and consequences of failure at any point during the remaining useful life, functional or final failure. This should include the time and effort required for recovery</p><p>Much seminal work on prognostics has been completed by NASA. Many peer-reviewed papers and textbook chapters may be found <a href="https://ti.arc.nasa.gov/publications/?sattr=prognostics&amp;stype=title">here</a>.</p><p>To fully understand the scope of prognostics we can use a model that breaks down the types of prognostics possible:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*r6WJWLnCsTq5lyeepXKMnQ.png" /><figcaption>Types of prognostic approaches</figcaption></figure><p>The Type 1 models use historical age at the time of failure data to determine the probability of failure over the age of the component. Type 2 models are able to build on type 1 and introduce other variables that may influence the speed of deterioration. Type 3 models are able to utilise digital twins and other data-driven or physical models to infer prognostics. Type 4 models use sensing that can directly measure either the effects of or the failure mechanisms directly and use other models such as particle filtering to determine RUL.</p><p>When considering IFD, type 1 models can predict the onset of failure events, but the model is not relevant to the RUL after the inception of failure and diagnosis. Weibull could be used in a separate model for RUL, but IFD RUL Weibull models are not the same as Weibull’s modelling the full age of an asset and its probability of failure. When looking at the specificity of the model types, Type 1 is more general for a population of the same component for a given failure mode, where type 4 is specific to an instance of a failing component.</p><p>Where diagnosis was a type of classification problem, prognosis is a type of regression problem and different techniques and models are used to measure each problem.</p><p>Prognosis is difficult because of high degrees of uncertainty, and that we need to predict the future operating and environmental influences that may or may not occur. If we have historically recovered all previous instances of failure events before the point of functional failure, we have little historical failure data that may be used to decrease uncertainty.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/993/1*7PRPQXvLeI9C6yWMyQku_w.png" /></figure><p>The practical measurements for calculating RUL are based on the prognostic horizon where any prediction of RUL should be long enough to allow recovery to be prepared before end-of-life or functional failure. The allowable tolerance, alpha a is determined by the time needed to prepare recovery and reduces in a cone as operating time passes, with lambda reducing the allowable error the closer to EoL. The points of estimation using an algorithm should have uncertainty deriving from measurement and other factors to provide measures of uncertainty.</p><p>The second type of maintenance task we discuss in this blog is Failure Finding tasks.</p><h3>Failure Finding tasks</h3><p>A failure finding task is necessary where components may fail unnoticed by the operators or maintainers conducting their normal duties. Before LED lights, with lots of redundancy were introduced into cars, a good example of hidden failure was when a car rear brake light failed. The failure went unnoticed by a driver in their normal activity of driving. The braking lights function is to inform other drivers following you that you were slowing down by applying brakes, so rear-end collisions could be avoided. Such circumstances are called ‘hidden failures’ in the RCM process. Hidden failures are particularly important, where the component or parts are safety devices. A hidden failure leaves no or reduced protection that the component is intended to prevent. The purpose of a failure finding task is to determine whether the component is in the failed state or not. This does not rule out conducting on-condition monitoring if it is effective.</p><p>Note: if standard operating procedures are changed, or staff are reduced, then a review of the maintenance regime should be reviewed to ensure there are no new hidden failures resulting.</p><p>Examples of safety protection devices include pressure-relief valves where a valve failing to the shut position would be hidden.</p><p>Most safety protection devices are designed to fail in the ‘safe’ or ‘noticeable’ state, but some cant be. Alarms and warning systems should fail-safe. Redundancy is often employed to help improve responsiveness for safety protection. Pressure vessels are often fitted with two or more pressure relief valves. Other protection systems using sensors may also be configured using majority voting logic. An example is where four sensors may be used to detect rotating machinery overspeed, it will take three out of the four sensors to sense overspeed for the protection to trip the machine. If the overspeed sensors individually failed to the trip state (which should be indicated to the machine operator — so it is not hidden) then the protection system will be very reliable and also responsive. The supplies to the sensors and their infrastructure need to be separate and independent of each other to avoid common modes of failure that may affect all four sensors.</p><p>Failure finding tasks usually involve physically testing the trip functions. Overspeed tests can be conducted on rotating machinery if speed control is possible. Examples may include turbines. Other failure finding tasks may include removing some components and bench testing them. Pressure-relief valves may be checked in this way. Other protection devices may also be tested by isolating the device from its normal system, and applying local pressure, temperature or electrical signals that functionally tests the protection device. For example, an induction motor overcurrent trips may be tested for instantaneous overcurrent (protecting against a short circuit condition) or slow overcurrent trips (protecting against single phasing conditions) by injecting currents through the secondary coils supplying the trip devices, that are fitted around the primary supply cables.</p><p>Another example of hidden failures is with standby machinery. There is a concept of hot and cold standby where hot standby is where the standby machinery is always running and is ready to take the load if the primary machine trips. Cold standby is where the standby machine is normally shut down and automatically runs up to take the load of a tripped duty machine. The hidden failures for either are: will the standby machinery take the duty load, and in the case of the cold standby machine, will it fail to run upon demand. It is possible to exercise cut-ins and load taking by switching off duty machines and checking the standby runs and takes the load.</p><h3>Failure finding task periodicities</h3><p>The principles for failure finding tasks are to understand the probability of failure over time and choose a periodicity that reduces the risk of failure to an acceptable level. Often in a safety situation, Safety Integrity Levels (SIL) are used to specify the maximum required probability of failure, and the task intervals are adjusted accordingly. In some systems, international standards may be adopted that set required periodicities. I have worked in regimes where the American Society of Mechanical Engineers (ASME) <a href="https://www.asme.org/codes-standards/find-codes-standards/ptc-25-pressure-relief-devices">codes</a> for pressure relief valves were used. In steam systems, pressure relief valves were functionally tested once per year.</p><p>Note: How many maintenance systems explicitly identify safety-critical parts where maintenance periodicities have been set by standards, legislation or a safety case. How would you avoid a latent failure and inadvertently change the periodicity?</p><p>Many maintenance texts recommend using Mean Time Before Failure (MTBF) to determine failure finding periodicities. I am highly sceptical of this process unless you know that the underlying failure pattern is random. If you do not know the failure pattern or distribution using MTBF is not suitable and if the impact or consequences of failure are SHEL implicated, it could be dangerous.</p><p>Some methods of determining periodicity take into account both the probability of failure of the component plus the probability that the event that requires protection may occur whilst the protection is negated.</p><p>Thinking about master data and metrics</p><ul><li>In an FMEA or an equipment register have we classified equipment functional failures as having SHEL, Operational, Economic or low significance?</li><li>In an FMEA or equipment register have we identified equipment as having hidden failure modes, and are there associated hidden failure tasks?</li><li>Have maintenance items that are subject to SHEL or standards criteria been identified, and is the maintenance regime compliant with their stipulations?</li><li>Has the CMMS got a warning flag that scheduled maintenance items, that are required by SHEL or by standards, that the periodicities must not be altered and that these maintenance items cannot be deferred?</li></ul><p>It is worth introducing a segue here to discuss the classification of SHEL criticality. What we have so far discussed in this blog has been associated with machinery, how it has been integrated into systems as its <strong><em>intrinsic </em></strong>SHEL concerns. In other words, the SHEL concerns that the machinery has from its design, manufacture and integration in its plant or operation. This is not enough; we must also think about functional safety and how safety changes as we use equipment in its operating context and environment.</p><p>For example, if we take a mining truck, that transports ore from an open cast pit, then if a failure occurs that may normally be operationally implicated, that causes the truck to not be driveable, if the failure occurs whilst the truck is in the pit, then the consequences are probably promoted to SHEL. This may be because the exposure of the maintenance staff to recover the truck in a hazardous environment carries higher safety risks. This poses the question:</p><blockquote>Has the operational context, environment or any abnormal operational procedures cause operational risks to increase, and have these been mitigated with plans or processes? Has this knowledge been captured in the maintenance and asset management master data?</blockquote><p>In the previous <a href="https://medium.com/ox-mountain/maintenance-metrics-b87502230104">blog</a>, we saw the metrics required for timeliness of scheduled maintenance, using a timeliness matrix. The matrix is not applicable for failure finding, but It is important that failure finding tasks associated with safety functions are completed on time and are not deferred.</p><p>The timeliness is an important metric, as well as looking at the failure rate and making sure the statistics that assume acceptable reductions in the risk of failure and constantly validated is the other metric</p><p>The task is intended to discover failures. In the Weibull analysis, if we conduct a failure finding task and physically check the function and find the failure, we cannot determine the exact age of failure. We know the failure occurred at some time between the last task and the latest one. The analysis of data may involve increases in uncertainty.</p><p>In this blog, we have covered deriving metrics for on-condition failure finding types of maintenance tasks. We discussed the applicability of diagnostics and prognostics to the two models of on-condition we introduced in a previous blog here. We applied a confusion matrix to diagnostics which we presented as a type of binary classification problem and then summarised NASA’s work in measuring prognostic regression to predict End of Life. We discussed the importance of failure finding tasks especially when they are associated with a protection, alarm or warning system. Many safety-related failures finding task periodicities may be set by standards or with the organisation’s own safety rules. This means changes need to be justified with robust cases.</p><p>In the next blog, we will finish this series of three blogs to show how the metrics we have discussed in the first two blogs in this series. We will show how the metrics should be automated and can feed a more generalised higher-level metric that ensures we can avoid being overwhelmed with too many metrics.</p><p>We have focused on metrics associated with maintenance tasks, we would like to hear about other general metrics associated with maintainability, availability and reliability. Please share your stories and experience.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bacdde9aac55" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ox-mountain/maintenance-metrics-bacdde9aac55">Maintenance Metrics</a> was originally published in <a href="https://medium.com/ox-mountain">The Maintenance Guru</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Maintenance metrics]]></title>
            <link>https://medium.com/ox-mountain/maintenance-metrics-b87502230104?source=rss-53d6f9846bf7------2</link>
            <guid isPermaLink="false">https://medium.com/p/b87502230104</guid>
            <category><![CDATA[planning]]></category>
            <category><![CDATA[maintenance]]></category>
            <category><![CDATA[analytics]]></category>
            <category><![CDATA[metrics]]></category>
            <category><![CDATA[optimization]]></category>
            <dc:creator><![CDATA[The Maintenance Guru]]></dc:creator>
            <pubDate>Thu, 04 Mar 2021 16:44:19 GMT</pubDate>
            <atom:updated>2021-03-04T16:44:19.100Z</atom:updated>
            <content:encoded><![CDATA[<h4>Part 1. The underlying metrics for scheduled replacement or restoration type tasks.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*qpETPCY4hFSMfVOO" /><figcaption>Photo by <a href="https://unsplash.com/@spacexuan?utm_source=medium&amp;utm_medium=referral">Crystal Kwok</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>In previous blogs, we defined a breakdown of maintenance task types and later on discussed how Weibull analysis helped us identify some of the conditions for applying the types of task. We will build on those foundations.</p><p>This blog will start to look at metrics for each maintenance task type in terms of their effectiveness and timeliness. Each different maintenance task type requires separate metrics to measure their applicability, effectiveness and efficiency. Where metrics are needed, we will mark these with a METRIC: with a small explanation in the text.</p><p>The taxonomy of maintenance task types we described in a previous <a href="https://medium.com/ox-mountain/maintenance-task-classification-29083df05038">blog</a> is this:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lTWOHm_3t7WlAlyS-ckKAQ.png" /><figcaption>Maintenance Task Type Taxonomy</figcaption></figure><p>We are going to look at scheduled restoration and replacement type tasks in this thread. (replacement is sometimes termed scheduled discard)</p><p>First, we are going to look at the pre-requisites for applying scheduled replacement tasks.</p><ul><li>The failure pattern needs to be aged, where the Weibull shape &gt; 1.3, preferably &gt; 2.6. With an age failure characteristic, there is a period of useful life, where the probability of failure is low before the probability of failure substantially increases with increasing age.</li><li>Where “1.3 &lt; shape &lt; 2.6”, it is a ‘weak’ age characteristic and there may be an option to check the practicality and apply an on-condition task.</li><li>Is the task practical and cost effective?</li></ul><p>We discussed the Weibull methods used in Reliability engineering in a previous <a href="https://medium.com/ox-mountain/selecting-maintenance-tasks-using-weibull-analysis-7545ea5fba9d">blog</a>.</p><p>This immediately identifies the first possible metric.</p><blockquote>METRIC: Are we doing the correct type of task? Is there a correlation between the Weibull shape and the type of maintenance we apply, for all maintenance tasks?</blockquote><p>The next important consideration is for each candidate failure mode whether the consequences of failure are classified as:</p><ul><li>Safety — Health — Environmental or Legislative (SHEL),</li><li>operational,</li><li>economic</li><li>minor significance</li></ul><p>We will mostly focus on operational or economic consequences as this is where there is most leverage that maintenance organisations can exercise, even with the understanding that safety is paramount.</p><p>If the failure consequences fall into the SHEL bracket, the aim of the associated scheduled tasks will be eliminating or reducing unplanned failure events to an acceptable minimum probability. This implies that the cost optimisation method discussed below may not be applicable.</p><p>With some safety-critical parts, the optimal time for changing out components may be determined by using a fraction of their validated useful life. Examples include the rotating components in gas turbines or landing gear in aircraft. This is where the probability of failure can be demonstrated to be lower than a set acceptable level. Safety Integrity Level (SIL) policies/standards that apply in many industries are used to determine what the acceptable limits are. Further information about these processes can be found <a href="https://amog.consulting/our-work/safety-risk-engineering/319-safety-integrity-level-sil-assessments-in-mining#:~:text=Analysis%20of%20hazard%20risk%20leads,level%20of%20dependability%20or%20integrity.">here</a>.</p><p>If the failure consequences are classed as operational or economic, we can look at two methods for determining periodicity involving Weibull analysis.</p><blockquote>Note: As an aside, using MTBF as a measure to define periodicity without knowing the underlying distribution of failure events can lead to costly mistakes being made. MTBF assumes a constant failure rate.</blockquote><p>The first, simpler method, is to use the Weibull Cumulative Distribution Function graph where the shape is greater than 1.3, and we are looking to apply scheduled replacement. We can choose an acceptable probability of failure and read off the age using the graph, and use this as the periodicity. It is common to talk about the “B-xx” age using Weibull where B-20 is the age where it is likely that 20% of the population of components may have failed. With the lack of cost optimisation capability, an organisation may choose to select the B-10 or B-20 age as their periodicity for scheduled replacements, depending on the severity of operational or economic failure consequences. The figure below illustrates the use:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*p2Gyqh10oODUrr73cUD-5g.png" /><figcaption>Weibull CDF, showing age failure pattern, and how periodicity could be selected</figcaption></figure><p>The second method is preferable as it optimises maintenance through-life-costs using simple simulation. This approach samples the Weibull distribution for random failures against different scheduled replacement periodicities. If the random age exceeds the periodicity, the simulation generates a scheduled replacement event, if not it generates a failure event. The simulation aggregates and then summarises a number of runs calculating the cost difference between applying the preventative and corrective maintenance over a population of the same components. The costs to take into account are:</p><ul><li>Material and logistics costs</li><li>Labour costs</li><li>Any lost production times associated with unplanned recovery.</li><li>Any costs associated with lower product quality caused by an unplanned failure.</li></ul><p>The simulation needs to account for the timings for recovery from an unplanned failure. If the recovery is urgent, because it has severe operational consequences, then the likelihood is that lost production time or reduced product quality. If the failure is economic then it may be possible to wait until the next planned outage to conduct the corrective maintenance. In this case, operational consequences are minimal. The type of corrective maintenance should be termed “Immediate (unplanned) Corrective” or “Deferred (planned) Corrective”. We can see these two types of corrective maintenance in the taxonomy diagram above. An example of failures with economic impact only may be the recovery of standby or redundant machinery.</p><p>In these optimisation calculations, it is usual practice to not include lost production times associated with planned replacements or deferred corrective, because:</p><ol><li>the planned outages have already been financially accounted for in the larger operational and maintenance plans for the whole site</li><li>the planned outage combines many tasks besides the replacement or deferred corrective tasks we are interested in. Lumping all of the outage time costs on one item distorts the calculation.</li></ol><blockquote>Note: if planned outage time is over or under-estimated for the work that needs to be done, then corrections need to be applied to planned outage time and budget forecasting may be skewed.</blockquote><p>These considerations result in the need for the following metrics:</p><blockquote>METRIC: Have we classified the failure mode consequences adequately between SHEL, Operational, Economic and low significance classes, and determined periodicity appropriately. These classifications are usually identified in Failure Modes &amp; Effects Analysis (FMEA).</blockquote><blockquote>METRIC: Are planned outage times, and their periodicity optimal and are the outage times fully utilised?</blockquote><blockquote>METRIC: What is the percentage of immediate or deferred recoveries?</blockquote><blockquote>METRIC: What is the variance of planned and corrective (immediate and deferred) tasks, and are the right figures included in maintenance planning estimations. (some systems predefine maintenance in templates or task lists)</blockquote><p>Labour costs may be higher for unplanned failure because troubleshooting to isolate the fault condition and fixing other secondary damage caused by the primary failure is likely.</p><p>If the results from the simulation were plotted, you would see a cost optimisation chart similar to the following:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9raM7iw7li952QcIb9nzRQ.png" /><figcaption>A plot of the results of a simulation to optimise scheduled replacement periodicities</figcaption></figure><p>We run the simulation through a series of PM periodicities (ages), where these are short, the cost of doing more scheduled removals is high. The cost of unplanned removals may also be relatively high at this point because over-maintaining a component is very likely to cause premature failures. In other texts showing optimisation of the scheduled maintenance costs, the tendency to show the effects of premature failure is often overlooked. As the periodicity is increased the scheduled removals grow less frequent and their costs fall, the cost of unplanned failure bottoms out and then increases. The optimal periodicity is where the sum of the planned and unplanned costs is at a minimum, indicated by dotted lines in the diagram above.</p><p>Getting a mathematically optimal periodicity is not the end of the story, we now need to consider ‘Maintenance Packaging’. We need to round the periodicity up or down to an underlying maintenance cadence, so this task when scheduled in a plan will coincide with other related tasks. This allows planned outages and packages of work to be sensibly scheduled. We can illustrate this using a simple example of an engine and turbocharger.</p><p>An engine may be swapped for new at a designated half-life of a mobile asset. When the engine is changed, so is the turbocharger. The turbocharger is also changed out by itself, more frequently than the engine. The periodicities of the engine and the turbocharger should be integer factors of each other. If the engine change-out periodicity is 20,000 operating hours, then the Turbo periodicity needs to be 10,000 or 5,000 hours so that wasted work is minimised. If our simulation studies indicate that turbo’s mathematical optimal is 11, 506 hours and that the assets most frequent common periodicity over every maintenance task fr the whole asset is 1000 hours, then the consideration should be to extend both the engine and the turbo to 22,000 and 11,000 hours, alternatively, keep them both at their previous periodicities.</p><blockquote>METRIC: Are all task periodicities packaged so they are in line with planned outage cadences and periodicities for associated components. The underlying set of periodicities should align to minimise through-life costs.</blockquote><p>Where the consequences of failure are insignificant then we should classify this finding against the part in our FMEA, to show that it has been considered. This may seem obvious, but it does provide an explicit indication that we should not expend any further resources or time analysing these parts unless there is a major change made either to the asset or its use.</p><p>Let us assume that the simulation we have described has been completed and that the optimal periodicity for planned removals or restoration has been calculated and applied before we proceed to the next stage. We should now look at the timeliness of the scheduled tasks.</p><h3>Timeliness of scheduled tasks</h3><p>We can look at two dimensions</p><ol><li>Whether the task was completed early, on time or late.</li><li>What was the state of the component at the replacement event, had it failed or was it still operational?</li></ol><p>This is best viewed as a matrix:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Rtie7xpkYwoAUwpNQnHbeg.png" /><figcaption>A matrix for classifying waste in scheduled maintenance</figcaption></figure><p>Taking each quadrant in the matrix as numbered in the figure above:</p><p><strong>1.</strong> <strong>Possible premature failure. </strong>It is possible that there will be a small number of early failures, given time to failure is a distribution, where the tails never reach absolute zero. However, if the numbers of events start to creep up, we should use the Crow-AMSAA technique to determine whether we are observing a growing quality problem. Then if the severity of the failure warrants, conduct root-cause analysis (RCA) to determine a preventable set of causes. If we look up the probability of failure for the periodicity age on a weibull cumulative density graph, we can expect that a small percentage of the components are likely to fail. but if we observe the percentage of actual failures grows above this figure, coupled a change of reliability trend seen in the Crow-AMSAA then it is likely we are experiencing a new development of premature failure. The cost of premature failures is the loss of economic life, increased rate of corrective maintenance and possible lost production.</p><p><strong>2.</strong> <strong>Avoidable unplanned failure.</strong> This is where the component has failed after its age has exceeded its periodicity time, and the planned replacement is too late. The cost is the difference between an unplanned and planned removal event. This is a process violation. Sometimes it is necessary to defer planned maintenance, but the costs and risks should be balanced when making these decisions.</p><p><strong>3.</strong> <strong>Wasted Economic Life.</strong> This is where a working component has been changed before the due periodicity. The cost is a proportion of the discarded economic life of the changed component, and if this practice of early change-outs is continued it increases the through-life costs with too much work being done. A known cause of premature failure is over-maintenance, and so there may be a relationship here that may be found during RCA.</p><p><strong>4.</strong> <strong>Gambling.</strong> This is where the planned replacement is late, on an operational component. This might be thought of as a win as we have squeezed some more economic life from this particular component. However, when taken over a fleet of components throughout their lives, and we trust our optimisation, it is more likely we will lose out and increase through life costs. Betting against the house inevitably loses.</p><blockquote>METRIC: Scheduled Replacement timeliness shall be measured and converted to costs deviating from optimal cost by applying the matrix rules.</blockquote><h3>Weighting the cost function</h3><p>Calculating the cost of the matrix process variation may also be finessed. This is best illustrated by considering that if you executed the maintenance at 5% of the timeliness target the penalty would be uniformly proportional to 50%. This method of apportioning cost is not optimal. There should be a graduated scale of cost versus deviance. The use of a sigmoid function provides a means of weighting the cost of variance to lessen penalisation closer to optimal and maximise penalisation the greater the variance. This is illustrated below.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/987/1*r0BrbVSE_T32QikKGkmQ6w.png" /></figure><p>Where x equals zero on the x-axis, is equivalent to executing the maintenance at its optimised periodicity. The sigmoid function can be applied to weight the cost of variation, an explanation can be found <a href="https://en.wikipedia.org/wiki/Sigmoid_function">here</a>.</p><h4>Managing to many metrics</h4><p>We have outlined and suggested a number of low-level metrics for measuring the scheduled replacement or restoration tasks. One risk is that as we cover the whole taxonomy of maintenance tasks, we will generate a significant number of metrics, that become difficult to manage. There are two mitigations to this</p><ol><li>These metrics need to be automated and derived automatically from the maintenance system, without manual effort.</li><li>The metrics need to be implemented in a hierarchical manner with higher levels summarising lower-level metrics we have defined here.</li></ol><p>How we summarise and produce a hierarchy of metrics will be included in the end blog of this thread.</p><p>In this blog, we have started the validation for why we should classify our maintenance tasks in a CMMS according to the taxonomy in the diagram above. This taxonomy was discussed in greater detail in an earlier <a href="https://medium.com/ox-mountain/maintenance-task-classification-29083df05038">blog</a>. The reasons why we need to break down tasks are:</p><ul><li>The periodicities for each type of task are calculated differently</li><li>The metrics for the low-level applicability, effectiveness and efficiency are different for each type of task</li><li>The applicability for tasks also depends on the patterns of failure, as discussed in the previous blog on Weibull analysis.</li></ul><p>This blog is part of a thread proposing a number of low-level metrics that should be applied to measuring the applicability, effectiveness and efficiency of scheduled restoration of restoration type maintenance tasks. Please feedback any other metrics you use for different types of maintenance task</p><p>In the next blog, we will look at metrics used for on-condition maintenance, including predictive maintenance.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b87502230104" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ox-mountain/maintenance-metrics-b87502230104">Maintenance metrics</a> was originally published in <a href="https://medium.com/ox-mountain">The Maintenance Guru</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why are there so many inspections with on-condition maintenance?]]></title>
            <link>https://medium.com/ox-mountain/why-are-there-so-many-inspections-with-on-condition-maintenance-76a34fc0f417?source=rss-53d6f9846bf7------2</link>
            <guid isPermaLink="false">https://medium.com/p/76a34fc0f417</guid>
            <category><![CDATA[condition-monitoring]]></category>
            <category><![CDATA[maintenance]]></category>
            <category><![CDATA[asset-management]]></category>
            <category><![CDATA[risk-management]]></category>
            <category><![CDATA[predictive-maintenance]]></category>
            <dc:creator><![CDATA[The Maintenance Guru]]></dc:creator>
            <pubDate>Fri, 19 Feb 2021 16:11:45 GMT</pubDate>
            <atom:updated>2021-02-22T12:59:05.450Z</atom:updated>
            <content:encoded><![CDATA[<h4>why can it seem so wasteful?</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*oJ55p_ebvYmhpm3n" /><figcaption>Photo by <a href="https://unsplash.com/@spacexuan?utm_source=medium&amp;utm_medium=referral">Crystal Kwok</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>In my career, I have heard several people criticise on-condition maintenance as being wasteful. To paraphrase</p><blockquote>“There are so many inspections and they hardly ever find a failure, especially after conducting an RCM project”.</blockquote><p>In this blog, we will explore this issue by discussing some of the principles behind on-condition maintenance and showing why there may be so many inspections. We can also show techniques for their reduction whilst still retaining the integrity of the maintenance to avoid failure consequences.</p><p>In order to address this criticism, we need to establish the basic principles behind on-condition maintenance. I believe we need to use two models to describe On-condition maintenance. Most other textbooks do not differentiate between these two models, which I think can lead to misunderstandings. The first model is where deterioration of the component’s condition begins either after manufacture (such as rubber hoses curing) or when the component is introduced into its operating environment (such as a ship launch into seawater). We call this model <strong>long-term deterioration</strong>.</p><p>The second model is where rapid deterioration starts after an initiating failure event. The initiating event can occur at any stage during the service life of a component. The failure initiation event may be caused by a shock or sudden overload situation overstressing the component. An example may be a bearing failure that is caused by a sudden overload. A car wheel bearing failing after driving over a really deep pothole. We call this model <strong>incipient-failure detection</strong>. We can use two basic graphs to illustrate the two models, which we will reuse and add more detail to in the deeper explanations.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0vfOt9SxhOHP7JBcgEvpjQ.png" /><figcaption>The basic on-condition maintenance model for Long-Term Deterioration</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xVB-HB1OGmrqpvXReL0mYw.png" /><figcaption>The basic on-condition maintenance model for Incipient Failure Detection</figcaption></figure><p>The major differences between the two models is the length of time the condition deteriorates over and the start of the degradation process.</p><p>The two diagrams above show degradation that accelerates over time, this is for illustrative purposes only, as some degradation may be constant and others noisier because of other changing physical properties. We should expect much noisier degradation trends in the real world. Before we dive deeper into each model, it is worth explaining some of the pre-requisites for applying on-condition maintenance to set the context for the rest of this blog.</p><p>On-condition maintenance is only applicable when:</p><ol><li>There is a time-lapse between either initiation of a failure or the start of degradation until functional failure.</li><li>That the deterioration is detectable at a time before functional failure (Usually called Diagnostics) that allows:</li></ol><p>a. the functional failure to be avoided.</p><p>b. the recovery of the failing part may be planned with resources predisposed</p><p>c. Additionally, if possible, the time to functional failure should give some leeway to time the recovery withdrawal of service to minimise operational impact</p><p>3. The detection:</p><p>a. Is highly certain (avoiding false positives or false negatives)</p><p>b. Isolates the failure modes</p><p>c. Isolates the components to those requiring direct corrective maintenance</p><p>4. The variance of times to functional failure, for different instances of the same failure mode on the same component is reasonably tight and dependable.</p><p>5. The prediction of remaining useful life (usually called prognostics) is possible</p><blockquote>A subtlety in maintenance semantics is worth pointing out here. Although on-condition maintenance is in the family of preventative-maintenance, it does not prevent failures. On-condition enables the avoidance of failure consequences. On-condition still requires us to take corrective action at the later stages of deterioration.</blockquote><h3>Long term deterioration</h3><p>We will first look at <strong>long-term deterioration</strong>. Where there is long term deterioration throughout the operational life, we need to consider two further sub-divisions of how and when the deteriorated condition can be detected, measured or inferred. The first subcase is when we can measure condition at any time but may need to consider other constraints. The second subcase is where we can only measure the condition late in the deterioration lifecycle, and there may be many earlier, seeming nugatory, inspections.</p><p>Regardless of when the condition measurement is possible it needs to be able to inform us what the remaining useful life (RUL) is before the component we are monitoring is unable to deliver its functions (functional failure).</p><p>Sometimes it is not possible to quantify the condition until it is late in the degradation lifecycle. This constrains what can be done. An example of where initial detection is only possible relatively late in the degradation cycle is the detection of fatigue cracking in metallic structures such as vehicle chassis, or aircraft fuselages and critical components like landing gear. If we draw the model diagram, we can introduce the idea of the P-F Interval.</p><p>We have introduced two terms that seem to describe similar things, Remaining Useful Life (RUL) and the P-F interval. What is the difference? The P-F interval is the period of time from being able to detect and diagnose an incipient failure, to the time of functional failure. The RUL is the time after diagnosis of a failure until functional failure. RUL counts down the time to functional failure as time goes by, the P-F Interval is fixed.</p><p>The failure mechanism of fatigue relieves stress by micro-cracking within the structure where the cracks join and then propagate until they reach a critical size where the structure is deemed unsafe. It is only possible to detect the cracks at a certain threshold size, and RUL is then usually calculated using crack-propagation models. Many inspection techniques may only detect cracks that appear on the surface of structures as well. Inspection techniques often referred to as NDE or NDT (Non-destructive examination or testing).</p><p>Techniques used to detect the cracking include ultrasonics, eddy-current, dye-penetrant and others. A useful blog can be found <a href="https://blog.foerstergroup.com/en/component-testing/the-6-most-important-methods-for-crack-testing-in-non-destructive-material-testing">here</a></p><p>The RCM process then instructs that the inspection periodicity needs to be a fraction of the P-F interval. The result may be a large number of inspections that early on in the degradation lifecycle will have very little chance of detecting cracking. We also need to apply the pre-requisite rule #2 above, in order that the inspection is valid and will add value. This is where some of the criticisms of too many nugatory inspections may be valid.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*T8nh2REdH3Q4L5T4uuuxuQ.png" /><figcaption>Long Term Deterioration — quantification of condition is only possible late in the failure cycle</figcaption></figure><p>An alternative to using the P-F interval to determine the frequency of inspections for metallic and most composite structures is to adopt the approach invented in aerospace after a Dan-Air cargo <a href="https://en.wikipedia.org/wiki/1977_Dan-Air_Boeing_707_crash">aircraft crash</a> in 1977. This is termed the ‘damage tolerance’ approach, which relies on the frequency of robust inspections being able to detect cracking such that cracks cannot develop to critical lengths between the inspections.</p><p>The essential pre-requisites for a damage tolerance approach of inspections for structural cracking are:</p><ol><li>An understanding of crack growth rates in relation to utilisation and cyclic stressors in the operating environment (thermal cycling, vibration etc)</li><li>An understanding of rogue manufacturing and accidental environmental defects</li><li>An understanding of the smallest size of defects that can be reliably detected</li><li>The determination of the largest defect size for safe operation</li></ol><p>In any safety regulated industry or application, all of these factors would need to be demonstrated and validated.</p><p>This results in the type of inspection schedule illustrated below. This addresses the basic criticism of too many inspections but requires rigour in being able to model and justify the failure mechanisms and how they are influenced by the operating context.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UTLVlIXMsS5lmM17o-IYHg.png" /><figcaption>Long Term Deterioration — Damage Tolerance inspection technique</figcaption></figure><p>We can also look at other examples where we can measure and quantify the deterioration at any time in the lifecycle, but in order to inspect we may need to interrupt operations or the sheer amount we have to inspect is extreme and this poses practical problems. An operational interruption example may be a gas turbine internal inspection to determine the rate of dissipation of the thermal barrier coating in the hot end. Thermal barrier coatings allow hotter combustion that improves thermal efficiency with temperatures exceeding the melting point of the guide vanes and blades. The inspection requires the engine to be shut down and cooled off enough to use borescopes.</p><p>Another example of a huge amount of infrastructure to inspect is pipelines. Pipes are often used to convey liquids or gasses for considerable distances. Trans Canada is an example. A pipe leak may be catastrophic and cause considerable damage to the environment. Among the pipe failure modes, pipe wall thickness tends to degrade and reduce over their economic life and are subject to internal and external corrosion. We will find it prohibitively expensive to continuously monitor all of the pipes. The use of <a href="https://en.wikipedia.org/wiki/Pigging">pigging scrapers</a> enables lengths of pipe to be measured whilst still maintaining flow. The use of drones can also be used for inspecting external issues, with drones being extensively used in other distribution infrastructures such as electricity transmission and distribution. With advances in technology advanced instrumentation and data systems are getting more effective for surveillance and inspections, but their use and scheduling has to be carefully planned to ensure the inherent rules of on-condition maintenance are adhered to.</p><p>Another consideration for inspections should be accessibility and the consequences of failure. An example may be inspecting pipes and the reactor pressure vessel inside a nuclear power station. Reactor pressure vessels are subject to work-hardening embrittlement because of the Neutron bombardment from the nuclear fission process. Cracking becomes more prevalent the more work hardened the pressure vessel becomes. Inspections are necessary for a radio-active environment where the ALARP principles (As low as reasonably practicable) apply to reduce the radiation dosage that workers may receive. Robotics for inspections are therefore commonly used.</p><h3>How can we further seek to reduce the extent of what we inspect?</h3><p>In such systems simulation and damage accumulation modelling that infers the condition are frequently used. These simulation techniques add a corroborative back-up to the inspection results and can be used in determining the risk of failure. Where models are used in conjunction with safety-critical components, the amount of validation &amp; verification, testing and proof used is considerable, and may also be cited in licences to operate the plant. As software tools emerge the ability to utilise simulations become easier and their adoption may become more widespread.</p><p>We can also exercise engineering knowledge and use a risk-based approach to determine which parts of our infrastructure are most likely to fail first. In a vehicle chassis, we know where the most stressed parts of the structure exist, and we can focus inspections on these parts. Welding used in structures also increases failure risks with heat-affected zones. In pipelines, we may have bends to allow thermal expansion of contraction or low points where moisture may collect. These areas may be more prone to failure compared with straight runs of pipe and are selectively inspected because they should degrade more rapidly. The inspection system then needs to ensure that inspections are repeated at the same location.</p><p>Given the systems we can quantify condition at any time, there is no P-F interval. How can we schedule the inspections?</p><p>If we can quantify the variation of times to functional failure, from introduction into service, using techniques such as Weibull, we can determine worst-case deterioration in the left-hand tail of the Weibull distribution. If we can further back this up with modelling for damage accumulation, we should be able to determine a scheduling regime roughly based on the half-life of the remaining deterioration life. This is illustrated below:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*K-BwCWACDP-aKIbb_xmjvw.png" /></figure><p>The determination of when to switch from half-life to regularly spaced inspections can be based on a risk approach to determine what is the most credible shock event that could occur that accelerates the degradation of condition to functional failure. The periodicities may be based on the minimum time of propagation of the fault after this shock. It is at this time that remedial action should be taken when the risks of instant functional failure from ever smaller shock conditions arise.</p><p>The Half-life approach to scheduling where we have long term deterioration where the condition can be quantified at any time in the deterioration lifecycle ensures that the number of inspections is minimised, and partially addresses the criticisms addressed at on-condition maintenance. The RCM process does not explicitly describe this method of scheduling.</p><p>Now we turn to the second model, Incipient failure detection.</p><h3>Incipient failure detection</h3><p>Incipient failure detection is where components have a distinct failure initialisation event that may happen at any time in their service or operating lives. After the initiation event, the condition deteriorates over a relatively shorter time than Long Term Deterioration until functional failure or worse is arrived at. The physical time lag between initiation and functional failure aligns with the first pre-requisite for on-condition applicability quoted above.</p><p>In order to be effective, the other pre-requisites need to be met, very important is the detectability that we are failing, at a time before functional failure that enables us to prepare for taking the component out of service, minimising operational disruption and apply the corrective maintenance.</p><p>If we take a rolling element bearing failure as an example, we can illustrate the principles. If a bearing is specified so it has a margin of load bearing over the maximum loads including most credible overload situations for its intended use, that it is manufactured with high degrees of quality assurance, that lubrication and protection from environmental stressors (e.g. corrosion) is maintained appropriately, and operations stay within specified limits then the probability of bearing failure is minimal and rare. However, lapses in any of these may result in an initiation of failure.</p><p>Bearing failures may be detected by observing a number of symptoms and techniques throughout the degradation cycle. The symptoms may become observable earlier or later. This is illustrated below</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zCQLjxIhfLGAW5UCifsBog.png" /><figcaption>Different types of detection for IFD, changes the P-F Interval</figcaption></figure><p>It is worth pointing out that it is possible to use fixed sensors that may measure parameters such as vibration and/or temperature and using software to automate the processes we can avoid periodic inspections or sampling because the monitoring is continuous in a predictive maintenance system. Dependent on how data is acquired from the sensors the diagnosis and prognosis can be rapidly alerted to the system user. This helps enable gaining more RUL to plan and prepare for corrective maintenance.</p><p>Often oil-debris relies on taking samples from magnetic chip detectors or filters, and vibration may be captured using handheld portable equipment. If a fixed temperature sensor is not fitted, it may be possible to measure temperature using a thermal camera or portable touch temperature sensor. In all cases, the sampling equates to an inspection and needs to be done according to the RCM process at less than half of the P-F interval.</p><p>If we think about the failure rate of bearings, we may only experience one or two failure events a year, where we may have an inventory of hundreds of machines with bearings. In order to ensure we capture all of the failure events, we must sample each at half of the P-F. If we assume the P-F interval is two months, for fifty machines each with two bearings we would have to sample twelve hundred times to detect two or three failures. This is why on-condition might at face value seems so wasteful.</p><p>A common mistake made is to extend the periodicity of inspections, without reference to the P-F intervals. This implies only a fraction of the true failure events may be detected, and some of those may not provide enough RUL to be beneficial. This may result in on-condition maintenance becoming a cost, with little or no benefit.</p><p>The vibration and oil debris sampling regime might be too expensive, but the maintenance supervisor should also consider the benefit of having an experienced maintenance person or suitably trained experienced operator carry out walk-downs of all the plant on a shift or daily basis. The five human senses coupled with an experienced human brain are very powerful on-condition sensing devices. Although the RUL may be quite small when detecting a bearing failure by touching the bearing housing, it may be enough to avoid functional failure consequences. Engineers conducting walk-downs are not usually captured or formalised in a CMMS system, but they are a vitally important part of the overall maintenance regime and recognised as a best practice among most experienced maintenance engineers.</p><h4>Cost justification for on-condition maintenance</h4><p>Unlike the long-term degradation, there are no smart techniques to reduce inspections or sampling tasks in Incipient Failure Detection, except to adopt predictive maintenance. In order to justify the cost of frequently sampling or inspecting we need to conduct a cost-benefit analysis which should be done as part of the RCM process. Recording such an analysis and showing the benefit may be used to defend the inspection regime from those who may mistakenly think it is wasteful.</p><p>We need to capture the following data for the analysis</p><ul><li>The P-F interval and possible variation of the P-F</li><li>The failure rate for the particular failure modes being addressed</li><li>The cost of the inspections, samples, and their analysis</li><li>The costs of incurring unexpected functional failure and unplanned corrective action</li><li>The costs of avoiding functional failure consequences, but including planned corrective action</li></ul><p>A spreadsheet could easily be used to determine the cost benefits. If there is a Weibull distribution available and current ages of all of the bearings, then the failure rate can be predicted over the next few years in a simple Discrete-Event-Monte-Carlo simulation using Python and some specialised libraries. This simulation should only take a few lines of Python script.</p><p>Where new equipment is involved or there is no existing data on failure rates or the P-F interval, the RCM process uses the knowledge of the most experienced maintainers and operators to glean these data, as these people best know their machinery. There are also other manufacturers or industry data that may be used as an initial value. As data is subsequently gathered, the business case may be re-evaluated.</p><p>It would also be prudent to think about the quality of the on-condition diagnosis, sometimes potential failures may be missed, and at other times a diagnosis may be made of failure, where none actually exists (commonly known as a ‘false positive’). A false positive may result in nugatory corrective work.</p><p>If the business case holds, and there is a benefit, then we should accept that the number of inspections to actually detecting failure is a natural, normal and acceptable consequence of conducting maintenance.</p><p>It may also be possible to set this business case against investment to develop a predictive maintenance system with fitting new sensors, where the initial investment can be set against the vastly reduced costs of the inspections themselves.</p><p>In summary, we have broken down on-condition monitoring into two basic models and described why the ratio of the number of inspections to detecting failures is high. We have explained some techniques where inspections can be reduced and outlined how on-condition may be cost-justified. The cost justification should be integral in the RCM process to demonstrate it’s a requirement for ‘practicality and cost-effectiveness. If RCM is done correctly then the many inspections for a few failures is justified.</p><p>Please let us know if you have tried RCM and found the process to be too onerous, have you experienced where the periodicities of inspections have been changed for cost reasons with no technical linkage to the driving principles of the P-F curve?</p><p>In the next blog, we will start a series looking at maintenance metrics and waste. This blog had feedback after we introduced the taxonomy of maintenance task types. One of the reasons this is necessary is that effectiveness, efficiency is different in every case, and we will need to measure different things. In the first of a thread, we will look at metrics for Scheduled Renewal or Replacement, and Scheduled Restoration maintenance.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=76a34fc0f417" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ox-mountain/why-are-there-so-many-inspections-with-on-condition-maintenance-76a34fc0f417">Why are there so many inspections with on-condition maintenance?</a> was originally published in <a href="https://medium.com/ox-mountain">The Maintenance Guru</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Exploring new methods of data exploration in the maintenance domain]]></title>
            <link>https://medium.com/ox-mountain/exploring-new-methods-of-data-exploration-in-the-maintenance-domain-891fcb85f76b?source=rss-53d6f9846bf7------2</link>
            <guid isPermaLink="false">https://medium.com/p/891fcb85f76b</guid>
            <category><![CDATA[maintenance]]></category>
            <category><![CDATA[rdf]]></category>
            <category><![CDATA[linked-data]]></category>
            <category><![CDATA[engineering]]></category>
            <category><![CDATA[rdfox]]></category>
            <dc:creator><![CDATA[The Maintenance Guru]]></dc:creator>
            <pubDate>Mon, 01 Feb 2021 16:57:13 GMT</pubDate>
            <atom:updated>2021-02-05T13:43:21.814Z</atom:updated>
            <content:encoded><![CDATA[<p>Part 2: A vision for the future:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*ODU6nkWjjEMxEoLH" /><figcaption>Photo by <a href="https://unsplash.com/@sortino?utm_source=medium&amp;utm_medium=referral">Joshua Sortino</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>In Part 1 we discussed how important data for maintenance may be held in a number of different silos. This data may be held in incompatible formats with different labels and names that describe common entities. Traditional database management systems also contain structured data. There is a whole raft of other unstructured data in documents, images and video footage that may also be exploited. This situation increases costs and obscures latent value that may be realised by having a shared and common data repository. Some existing systems for combining data include data lakes, with offerings such as SAP Hana, or adopting emerging standards such as <a href="https://www.mimosa.org/mimosa-osa-eai/">Mimosa</a> or the expansion of ISO 10303 (<a href="http://docs.oasis-open.org/plcs/plcslib/v1.0/cs01/help/plcslib_overview_content.html">PLCS</a>) to a full lifecycle perspective. There may be another option we discuss in this blog.</p><p>In this Maintenance Guru blog, we will show how some disparate data held in differing formats could be converted into an RDFox triple store, so we can realise new insights that would not be otherwise possible. We can also describe the ‘reasoning’ capability in RDFox to use logic to derive new relationships in the data with scripted rules.</p><p>This blog will avoid some of the scripting and programming details, to focus on the maintenance domain. There will be another version of this blog released under the <a href="https://www.oxfordsemantic.tech/">Oxford Semantic Technologies</a> Blog, that shows the scripting detail.</p><p>A data triple may be expressed as three elements, a subject -&gt; a predicate -&gt; an object. This can be illustrated in a graph format as:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/239/1*vBEqDwW7jW5GX1dzdJBd9w.png" /><figcaption>Fig 1: An RDF Triple</figcaption></figure><p>The triple forms a small directed graph, which makes sense semantically and logically and provides a more natural way of structuring data and its relationships compared with a relational database.</p><p>If we look at how we may use graphs to model what is in a database table, we can re-use a diagram we first saw in the previous blog:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*elVpg5BJ1zL89walQvRQnw.png" /><figcaption>Conversion of a SQL table to a triple graph</figcaption></figure><p>The part node corresponds to the data table that represents an ‘entity’. This may be a ‘type’ or a ‘class’ in the graph. Instances of the rows in the table appear in pink in the graph with the relationships between nodes representing the column names or attributes of the entity. The data table may have a primary key that uniquely identifies the row of values, which is represented as the ID referencing all the other values in a single row in the graph. This means data contained in relational database tables may be automatically converted into triples if the SQL tables or flat-file spreadsheets (or csv) are structured correctly.</p><h3>Systems views</h3><p>To illustrate the power of using triple stores, we will first describe three separate views of a fixed plant subsystem, which are normally in different information systems.</p><p>The system of interest has the following operating context. It is an automated filling system for a liquid holding vessel which has an electrically driven pump to supply liquid to a holding tank. The liquid in the holding tank is maintained between lower and upper levels from a level indication system. There is a system controller that starts and stops the pump via its electric motor starter.</p><p>The controller also suppresses continued starts if they occur too frequently, the motor winding temperatures may rise and exceed a threshold value. Induction motors draw 8 times normal running current when they start, the heat spike is marked. The danger is that the motor temperature increase starts to degrade the lacquer insulation on the windings that risk short circuit conditions, cascading to eventual motor burn out.</p><h3>The electrical system view</h3><p>Below is a diagram of the electrical distribution supplying the vessel filling system elements. The system is normally supplied via the national grid supply of three-phase 440v AC, via a grid feeder air-circuit breaker (a breaker is essentially a large switch). The diesel-generator is used as a back-up if the grid supply fails with its own supply breaker. An electrical system has under and over voltage and over current protection devices on all the breakers and starters. Fuses are also included for over current.</p><p>The value of the trip protection is set by rules of ‘electrical discrimination’ such that if a fault occurs on a downstream device the protection on the next upstream device will trip before any other, thereby protecting as much of the fault-free circuitry as possible. This rule helps us model the dependencies in the system.</p><p>Having an electrical distribution drawing makes it possible to describe how power is distributed from sources via isolation devices such as breakers fuses to their end services, such as machinery actuators or electronics.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*g-1fPEDkqsY6V5jQPFBuKQ.png" /><figcaption>Electrical distribution to the filling system</figcaption></figure><p>The electrical distribution diagram is already a type of graph format equipment is connected by supply cabling from sources (Grid and Diesel Generator) to end equipment such as the fill motor and sensors. These relationships should be able to be extracted from CAD drawings or other specialised drawing applications.</p><h3>The process view</h3><p>The following diagram shows a state machine diagram that describes the filling process for the tank. State machines are a commonly used tool in <a href="https://www.incose.org/systems-engineering">Systems Engineering</a> in a modelling methodology called <a href="https://www.omgsysml.org/what-is-sysml.htm">SysML</a>. Control system software specifications are often based on state machines and sequence diagrams.</p><p>This diagram essentially reproduces the logic inside the fill-system controller, but clearly shows the cyclic process of our system of interest under normal and abnormal states. Many SysML models should be able to be decomposed into data elements that can be structured and imported into a triple store.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1002/1*XvvRr8wTDV06-Gdufl4QBw.png" /><figcaption>Simplified State Machine diagram for the filling system controller</figcaption></figure><p>The diagram shows the controller starting and stopping the motor-pump through the opening and closing its starter air-circuit contactor in normal operation. The contactor is tripped if the motor temperature or a motor overcurrent situation exists and puts the starter contactor into a tripped state (which is open, but not cocked to reclose). The breaker cannot be reset to the open position in the idle state until the trip conditions are cleared, allowing the normal operational cycle to recommence. Each of the steps and the associated equipment it involves has an ‘effect’ on the filling process.</p><p>We can build a generalised ontology that models the relationships between processes and the equipment and use this to populate the filling system data. We can use an ‘affects’ predicate between the process and equipment. For example, the Filling process is affected by the level sensor, starter contactor, motor and pump. The data elements are contained in the state machine diagram that may be converted to the triple format. Part of the affects -&gt; process view can be seen in the figure below.</p><h3>The equipment’s view</h3><p>The following diagram shows a physical breakdown of the tank filling system showing the functionally significant equipment items at the lower levels. This type of breakdown is often seen in FMEAs. The data in this view is separated from, but has a correspondence to the electrical view, with the inclusion of functional location identifiers. Each node on the right of the taxonomy below is a maintainable item.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1022/1*AXp4OEcQSt58DSXZb0f-1w.png" /><figcaption>Equipment breakdown for the fill system</figcaption></figure><p>The equipment view contains electrically supplied components that have similar equipment breakdowns, and dependency data as shown in the electrical system view.</p><p>This is a graph view of the equipment data. Because the equipment view was originally held as SQL, it was not difficult in writing a python script to convert it to a triple format. The colour coding in the figure below shows the relationships between RDF types and instances. If we drew all the relationships within the figure, it would look like spaghetti.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_6oAdldjKgNZM0tDtxaAiw.png" /><figcaption>A graph view of the equipment breakdown structure</figcaption></figure><p>Now that we have seen and described the three views we want to combine, we now turn to the objectives we set.</p><p><strong>The first objective is to take these three views and combine them into a single triple data store</strong>.</p><p>Having converted each of the views into Turtle scripts and ensuring we have defined common RDF types to classify the same data attributes in all three views we can import the data into a common data store. Much of the work in combining electrical and mechanical equipment is already achieved.</p><p><strong>The second objective is to use rules to extend the data and use a query to address many fundamental questions.</strong></p><p>For example, a question might be:</p><p><em>What system processes are affected if I safely isolated the fill system pump motor for maintenance?</em></p><p>This question also relates to data we would want in a FMEA, the effects of the pump failure on the processes and what would be the functional failures, given the close relationship between required processes and functions.</p><p>The approach to answering this question is understanding the motor and the relevant relationships in the triple store. This query also relies on another business rule that is not yet expressed in the triple store. Safely isolating the motor from its electrical supplies may have an internal safety rule that applies that says <em>electrical machinery must be isolated by removing the supply fuses and tagging associated operating switches</em>. This rule ensures that we avoid a possible latent-process failure and significantly reduce the chance of a maintainer being accidentally electrocuted.</p><p>The following diagram shows the graph that is directly relevant to this query, where the <em>Supplied_by</em> chain from the motor to its nearest upstream set of fuses is shown.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hQciKR6RIy4krhGZ1wQU0A.png" /><figcaption>Part of the triple data showing the electrical safe-to-work boundary for the fill pump motor (to the nearest isolation fuse)</figcaption></figure><p>We electrically isolate the motor by removing the next upstream fuses. We will need to consider other machinery that has also been isolated, the starter, and the controller. These other <em>Fitted_equipment</em> items may have <em>affects</em> on other processes. The Starter affects the <em>Filling, Idle and Trip</em> processes whilst the motor <em>affects</em> the <em>Filling</em> process</p><p><em>It is worthy of note that in the previous paragraph we have used the English language to describe business rules and the relationships that very closely match the semantics that the triples represent. The triples are both computer and human friendly.</em></p><p>At this point, it is worth introducing the idea of transient relationships in the graph. We can use this to our advantage in making transitive relationships in the database using rules, to help query performance.</p><p>If we use the graph directly above, we can see that the motor links to the starter through an <em>isSuppliedBy</em> relationship. The starter is also similarly linked to the controller and main fuses. The motor is then transitively linked to the motor main fuses and the controller. We can take advantage of this by defining the transient relationships via a rule.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1018/1*Hpd0IruZI-loilrPL5mDmQ.png" /><figcaption>Shows motor isSuppliedBy transitive relationships</figcaption></figure><p>The transitive relationships shown from the motor, in red in the graph above, make the querying faster and easier to find the isolating fuses, than having to transit through several linkages of <em>isSuppliedBy</em> relationships. All of the new relationships were formed using a single rule.</p><h3>So what? What have we achieved?</h3><p>This small demonstration uses three separate sources of data about a subsystem and shows the following:</p><ul><li>through the conversion of the given data in three separate formats, we have formed a baseline triple store that contains the data in a single database.</li><li>We have used reasoning and rules to infer new relationships in the triple store that have linked data from different views of a system</li><li>We have queried the resulting triple store to extract information that is valuable in real-life maintenance circumstances</li><li>The triple store could be further developed to form the basis of an FMEA and other useful engineering frameworks. We have shown that using the <em>affects </em>property, we can determine the processes affected by a machine, which indicates the effects of failure in an FMEA.</li></ul><p>All of this might have been achieved through more conventional means using relational databases, but the level of effort required would have been greater than using the triple store. The inherent reasoning to derive new relationships and reuse of the data is game-changing. The triple store data schemas are inherent in the data itself, there are no primary or foreign keys of the SQL relational model. Extending the data and relations does not impact the original schema and therefore updates are so much easier and efficient.</p><p>The use of RDFox also allows the system to naturally <strong>scale to many millions of triples</strong> while retaining high performance.</p><p>The downsides of using the graph approach include a learning curve to understand how to use SPARQL as a query language and exploiting the RDF and RDFS libraries. RDFox does have an add-on module called a faceted search that could shield un inexperienced user from the complexities of SPARQL, which can be found <a href="https://medium.com/oxford-semantic-technologies/faceted-search-52625d25dbff">here</a>.</p><p><strong>In the future, the use of triple stores could be significantly enhanced if the engineering and academic communities could define template ontologies as open standards to describe equipment and engineered products within its full lifecycle context.</strong></p><p>I would be interested in being involved in defining an open standard for an FMEA ontology that is suitable for defining maintenance as I think<strong> industrial adoption of the graph technology could be simplified </strong>by having access to open-source ontologies<strong>.</strong></p><p>Additionally, I would welcome any reader’s experience using OWL, RDF triples, linked data or ontologies in the maintenance engineering domain? A future blog may return to this subject and show how we can derive other FMEA type data using this data set using rules and reasoning.</p><p>In the next blog, we will describe on-condition maintenance in a deeper way that will address some common misunderstandings. A common criticism I have heard about conducting RCM which may recommend a lot more on-condition inspection tasks, is why do so many inspections get done that do not find any failures — isn’t this wasteful? Does this waste question the effectiveness of RCM? We can answer this question and refute the criticism as well as describe how Predictive Maintenance belongs in the family of on-condition maintenance.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=891fcb85f76b" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ox-mountain/exploring-new-methods-of-data-exploration-in-the-maintenance-domain-891fcb85f76b">Exploring new methods of data exploration in the maintenance domain</a> was originally published in <a href="https://medium.com/ox-mountain">The Maintenance Guru</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Exploring new methods of data exploitation in the maintenance domain]]></title>
            <link>https://medium.com/ox-mountain/exploring-new-methods-of-data-exploitation-in-the-maintenance-domain-8a51b6bda464?source=rss-53d6f9846bf7------2</link>
            <guid isPermaLink="false">https://medium.com/p/8a51b6bda464</guid>
            <category><![CDATA[graph-database]]></category>
            <category><![CDATA[engineering]]></category>
            <category><![CDATA[database]]></category>
            <category><![CDATA[maintenance]]></category>
            <category><![CDATA[reliability]]></category>
            <dc:creator><![CDATA[The Maintenance Guru]]></dc:creator>
            <pubDate>Tue, 12 Jan 2021 15:44:19 GMT</pubDate>
            <atom:updated>2021-04-14T08:00:00.148Z</atom:updated>
            <content:encoded><![CDATA[<h4>Part One: The Current Method</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5rXejhcKLorK7JiXnQH-tg.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@sctgrhm?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Scott Graham</a> on <a href="https://unsplash.com/s/photos/data?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><p>In maintenance many processes we conduct rely on data. Some of this data may need to be elicited from experts and may not be encoded. Codified data may be split between different disparate systems and may not be easily linked. Each different information system may also hold data about the same entities stored differently that makes it incompatible with loss of interoperability.</p><p><em>For example:</em></p><ul><li>The Computerised Maintenance Management System (CMMS) may contain an equipment registry with physical location &amp; equipment breakdown structures, maintenance plans, work orders, notifications and task lists.</li><li>A machinery surveillance system may have the data to define a process diagram, with tag identifiers for sensors and their data, but the tag identifiers may not align with the sensor identifiers in the original engineering drawings.</li><li>There may often be Failure Modes Effects Analysis (FMEA) data that does not share the same physical breakdown of the same equipment as the CMMS.</li></ul><p>However, what if we could make these systems <strong>more succinct</strong>? Ensure that data is not stored in silos or hard to access?</p><p>This article will focus on framing the problem of disparate data and how we can rethink our classical ways of managing data, to improve our current systems. In the second blog in this series, we will introduce an emerging data management technology and show how this can centralise our data and use inherent reasoning capabilities to extend it and make it more amenable to ask general questions.</p><h3>FMEA Data Exploitation</h3><p>FMEA usually contains a physical breakdown of each system of interest or machine using a hierarchy, decomposing the machinery to constituent parts. The FMEA then looks to define functions, functional failures, failure modes, effects and risks. to populate the data linked back to the original breakdown.</p><p><em>A major reason why we use FMEA in the maintenance domain is that it contains the fundamental information we need to design and produce a maintenance regime. According to the international standard SAE JA1011, the use of an FMEA is mandatory if we conduct an RCM process.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*VKurIGQJc-x0FSZ9" /><figcaption>Photo by <a href="https://unsplash.com/@theblowup?utm_source=medium&amp;utm_medium=referral">the blowup</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h3>Fundamental problems</h3><p>There are some fundamental problems with a hierarchical model of the machinery.</p><ol><li>A single hierarchical equipment breakdown structure does not lend itself easily to model physical or functional dependencies in systems that use shared and distributed elements. Examples include electrical and hydraulic within a system of other mechanical components. We can model the electrical or hydraulic components and connecting pipes and cables, but the connections from the supplies to the actuators that sit in other branches of the hierarchies is problematic.</li><li>In an earlier blog in this series, we have argued that the purpose of maintenance is to preserve functions required of machinery in a defined context by its stakeholders. Why is it that most FMEAs are written against a physical breakdown structure? Surely a functional view of our equipment that links to the physical breakdown view would be more useful.</li><li>The FMEA is often created using spreadsheets, with columns and rows to record the data attributes described above as one-to-many hierarchical relationships. But this does not adequately model the real-world situation, for example, many failure modes may share the same failure mechanisms and vice versa. The true relationships are many-to-many. Additionally, if we misspell a shared entry in one cell compared with others it causes problems with filtering and grouping data.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ehmnVKyIv37wFVszoT6R5A.png" /><figcaption>Typical FMEA flat file report format</figcaption></figure><h4><strong>Use of SQL for FMEA data:</strong></h4><ol><li>FMEAs could use relational SQL databases, which can represent many-to-many relationships using link tables to solve this issue. Relational databases can be normalised so that all of the data is only entered once, (to at least strong third normal form) which avoids misspelling errors and other anomalies in the data possible in a spreadsheet flat-file approach.</li><li>However, querying normalised SQL databases through the link tables raises other problems in query complexity and performance. De-normalising SQL schemas is a common practice to improve performance when the data is linked to a user interface application. This change re-opens the risks of data anomalies unless extra verification code is added which can over-complicate application software.</li><li>Adapting and updating a database also presents issues in requiring changes to the underlying schema. If a user interface is built on the database with tight coupling, then this could involve major updates to the whole application. Changes are difficult and expensive and take too much time.</li><li>One important element of an FMEA which does not lend itself well to a spreadsheet tabular view is a description of the operating context of the system of interest. In my experience, an FMEA without an operating context description has marginal value for the design of a maintenance regime.</li></ol><p>We do need to remember that the spreadsheet tabular view of FMEA data is familiar and human friendly to many engineering people. Any FMEA application must be able to be reproduced in this report format, but this report format should not constrain the underlying data structure or schema.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*elVpg5BJ1zL89walQvRQnw.png" /><figcaption>Converting existing structured data to a graph schema</figcaption></figure><h3>A vision of what should be possible</h3><p>To capture the vital aspects of our systems efficiently and correctly we need a richer representation of our systems and their data. This would enable us to ask fundamental questions that we want to be answered.</p><p><em>Such questions could be:</em></p><ol><li>If I lost an electrical supply at any point in its distribution what system functions would I lose? This question relates to populating the Effects of a functional failure in an FMEA</li><li>What other components should I isolate, to form a trustable safe-to-work boundary around a component? This is a frequent requirement from maintenance to safely work on a component with the rest of the system still working, or isolating sources of stored energy, or hazardous materials. We could use our data to automatically populate task list instructions, or check the accuracy of existing task lists.</li><li>What effects on the functionality of the whole system will the failure of one, two or many components have? This goes beyond the limitations of an FMEA, in only analysing one failure mode at a time. We should be able to incorporate the data and relationships to reproduce the functionality of a Fault Tree or HAZOPs analysis.</li><li>What are the functional requirements of a system or machine including their performance standards? We want to have many views of our data from different perspectives including Physical breakdown and dependencies, functional dependencies, electrical/hydraulic perspectives, process perspective, and others.</li></ol><p>RDF Graph databases would provide a data management system that enables this kind of multi-perspective view based on common data. Linkages are expressed using data triples and the relationships between system entities can be much richer than a SQL database.</p><p>The <strong>most powerful </strong>aspect is the ability to use <strong>reasoning</strong> to enrich the data. This is where we have existing data and relationships, but using the inherent logic and rules, we can derive more meaningful relationships which allows us to gain more insights and derive more value.</p><p><strong>Part Two</strong> delves into how this can be done using RDFox a new reputable RDF triple store or database, to illustrate how all this can. To read Part Two of this blog, tune in in a fortnight. Part 2 can be found <a href="https://medium.com/ox-mountain/exploring-new-methods-of-data-exploration-in-the-maintenance-domain-891fcb85f76b">here</a>.</p><p>Has anybody used Graph, triple store or linked data databases in the Maintenance Context? Let us know about your experiences. Which graph database did you use and did you manage to merge different views of your data?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8a51b6bda464" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ox-mountain/exploring-new-methods-of-data-exploitation-in-the-maintenance-domain-8a51b6bda464">Exploring new methods of data exploitation in the maintenance domain</a> was originally published in <a href="https://medium.com/ox-mountain">The Maintenance Guru</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Selecting maintenance tasks using Weibull Analysis]]></title>
            <link>https://medium.com/ox-mountain/selecting-maintenance-tasks-using-weibull-analysis-7545ea5fba9d?source=rss-53d6f9846bf7------2</link>
            <guid isPermaLink="false">https://medium.com/p/7545ea5fba9d</guid>
            <category><![CDATA[engineering]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[maintenance]]></category>
            <category><![CDATA[advice]]></category>
            <category><![CDATA[statistics]]></category>
            <dc:creator><![CDATA[The Maintenance Guru]]></dc:creator>
            <pubDate>Thu, 10 Dec 2020 09:49:58 GMT</pubDate>
            <atom:updated>2020-12-10T15:57:20.341Z</atom:updated>
            <content:encoded><![CDATA[<h4>The data, distribution and what this means for maintenance</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FmT4XPM2dF0G5U9-OI4Jiw.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@inspirationfeed?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Inspirationfeed</a> on <a href="https://unsplash.com/s/photos/aeroplane-engines?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><p>In my career I have often found that Weibull analysis has been treated as if it is far too difficult and expensive to conduct. This might have been true when computing power and digital technology for gathering and processing data was uncommon. Today, the old reasons for dismissing Weibull analysis have largely evaporated.</p><p>The Weibull distribution is a continuous data distribution in the exponential distribution family. It enables us to determine whether components are failing and by which pattern of failure. This blog will outline the data required to conduct a Weibull Analysis, how to plot the Weibull distribution and what this means for maintenance.</p><h3>Why the Weibull distribution?</h3><p>Weibull was invented seventy years ago, and has been used extensively for life or failure analysis. Other distributions can be used such as the exponential or lognormal, however, these are normally used for specialised situations.</p><h3>What data do we need?</h3><p>For each part that we want to carry out the analysis on, we need its failure and change event data. We also need to know the operating age of the components at the time they’re changed.</p><p>The Weibull process is enhanced if we can use age data for parts that have been subject to scheduled replacement, when the removed part is still working. These functional component changes are termed ‘censored’ or ‘suspended’ data. Because suspended parts have not achieved their failure age at removal it adds evidence to the analysis because of their survival and makes the results more accurate.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*n1SuhGo-kP-uO88ixPAXVQ.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@matthew_t_rader?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Matthew T Rader</a> on <a href="https://unsplash.com/s/photos/train?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><h4><strong>How to define age</strong></h4><p>If machinery operates or is exposed 24/7, age could be elapsed calendar time. If not it is best to use operating hours, or some kind of cyclic count.</p><p>For example, the life of a medium voltage air circuit breaker’s main contactors is proportional to the number of on-off cycles, as electric arcing dissipates contactor material. The age of the breaker contactors is expressed in switching cycles.</p><p>As well as components that have been subject to planned removals, the age of the currently fitted parts should also be included in the data as ‘suspended’ events.</p><h4><strong>Fitting age data to a Weibull distribution</strong></h4><p>There are a number of well-known estimation methods for fitting age data to a Weibull distribution. Ranked Regression (using the median ranks on the y values) Maximum Likelihood Estimation (MLE), and the Method of Moments. The latter only takes into account the failure events, and so is not as useful.</p><p>A useful web site using excel can be found at <a href="http://www.real-statistics.com/">www.real-statistics.com</a>. Years ago Reliability engineers used specially printed log-log (probability plot) graph paper to manually plot the failure ages and fit a regression line by eye, before reading off the parameters that represented the Weibull distribution. The ‘goodness of fit’ of the regression line was seen by eye. Today specialised software is used to calculate the estimates.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/802/1*PxGEUDgI_f6FCOqr6hSBWw.png" /><figcaption>source: <a href="https://weibull.com/hotwire/issue8/relbasics8.htm">https://weibull.com/hotwire/issue8/relbasics8.htm</a></figcaption></figure><p>Usually people with an engineering background prefer the ranked regression method, because the probability chart allows the data and the fit to be visualised. It is easy to see whether the data suffers from mixed failure modes, or whether it is appropriate to use the location parameter from the visual line fit.</p><p>Those from a statistical background usually prefer the MLE method as in general terms it is more accurate. My experience suggests that there are circumstances where you may prefer either. The MLE method has an inherent bias, which is worse when the shape parameter is &lt; 1 and the number of data points is &lt; 15. The dispersion of data is greater the lower the shape. In these circumstances I generally use the ranked regression method. MLE has separate methods for bias correction.</p><h4><strong>Parameters</strong></h4><p>The estimations can yield 1, 2 or 3 parameters that describe a Weibull distribution. The Parameters are the shape, scale and location. Most often a 2 parameter Weibull is used with the shape and scale, the location projects the distribution left or right on the x-axis. If deterioration begins and there is a probability of failure after manufacturing with components in stock before they are fitted to an operating machine. Shape and scale are often referred to using Greek letters beta and eta respectively.</p><h3>Let’s plot!</h3><p>We will take three sets of simulated data, to represent the three fundamental patterns of failure we introduced above, premature, random and wear-out. The number of simulated failure events (500) is higher than would normally be encountered in the field, so we can show these examples clearly. The first three charts show a histogram of the event age data for the three cases:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*39MDaR_DGIGBTr7GoOc02w.png" /></figure><p>We now pass the three data sets into the Ranked Regression Weibull estimator, and see the probability plots (checked by eye) are a reasonable fit for the Weibull distribution.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YWFjtmLKUZP1nXJ_CAQseg.png" /></figure><p>There are two other views of the Weibull distribution, the Probability distribution function (PDF) and the Cumulative Distribution Function (CDF). You will note that each chart above provides the estimates of the shape and scale for each case. The shape and scale parameters are needed to plot the PDF and CDF below.</p><p>The PDF plots should be very similar to the histograms above. The premature PDF seems to be confined to the bottom left-hand corner of the chart, this is because the distribution is so dispersed, stretching out the values in x-axis.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*L2onD9FycUcfh0RP1xbvag.png" /></figure><p>The Cumulative Distribution Function (CDF) is useful as it can tell us the probability of failure of the components at a specific age. The Weibull Scale parameter is always equivalent to the age or cycles equal to the probability of 0.63. Probability is a number that is always between 0 and 1, and so is easily mentally converted to percentages. The Scale is the age at which 63% of the components are likely to have failed.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*57-oDOe29G66R1_DZ8FWkA.png" /></figure><p>A convention often used by reliability engineers is to quote reliability as “B-xx”, so using the scale factor, this is equivalent to ‘B-63’. Each case was generated from a Weibull distribution with a scale of 1000, but to ensure each of the charts were readable any sample ages above 6000 cycles were not included in the data set. This is why the estimated scale of the premature data set is 843.</p><p>Interestingly, if we look at the B-20 figures for each of the cases, where 20% of the components are likely to have failed, the difference between the three cases with similar scale parameters is strikingly different. This illustrates why premature failure is so bad:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1JB8ebVm8wi7Xr0swYIgLw.png" /></figure><h3>So what does this mean for maintenance?</h3><p>If we reflect on <a href="https://medium.com/ox-mountain/maintenance-task-classification-29083df05038?source=friends_link&amp;sk=addc79aadaec9c5f58db5e679e0cf1ac">our previous blog</a> where we align failure patterns with maintenance task prerequisites, we can now link Weibull analysis and the Weibull shape parameter.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AhcWpPPoB5TTB_V0HfALFA.png" /></figure><blockquote>Just a cautionary note, there are other pre-requisites and conditions that need to be present for the various maintenance task types to be applicable, we can discuss these in a later blog.</blockquote><p>Weibull analysis is not only extremely useful, but viable given the advancement in digital technology for gathering and processing the appropriate data. The shape analysis informs us what maintenance interventions and task types are applicable to the three basic patterns of failure. Thus, Weibull analysis helps justify our maintenance regime.</p><blockquote>What is your experience using Weibull? Have you thought it too difficult or expensive to conduct or have you embraced it? Please let us know your experiences and stories.</blockquote><p><strong>The previous blogs</strong> have discussed defined maintenance, and classified maintenance types, applicability and other attributes. You can find them <a href="https://medium.com/ox-mountain">here..</a></p><p><strong>In the next blog</strong> we are going to explore and demonstrate how a new emerging database technology can be used to better store, retrieve process and more importantly <strong><em>reason</em></strong> on the data we hold. We will look at the traditional methods and information systems that we use in maintenance and show how we can transform, innovate and improve effectiveness using new technology.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7545ea5fba9d" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ox-mountain/selecting-maintenance-tasks-using-weibull-analysis-7545ea5fba9d">Selecting maintenance tasks using Weibull Analysis</a> was originally published in <a href="https://medium.com/ox-mountain">The Maintenance Guru</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Maintenance Task Classification]]></title>
            <link>https://medium.com/ox-mountain/maintenance-task-classification-29083df05038?source=rss-53d6f9846bf7------2</link>
            <guid isPermaLink="false">https://medium.com/p/29083df05038</guid>
            <category><![CDATA[application]]></category>
            <category><![CDATA[advice]]></category>
            <category><![CDATA[maintenance]]></category>
            <category><![CDATA[engineering]]></category>
            <category><![CDATA[clarity]]></category>
            <dc:creator><![CDATA[The Maintenance Guru]]></dc:creator>
            <pubDate>Thu, 26 Nov 2020 10:08:00 GMT</pubDate>
            <atom:updated>2020-11-26T10:42:58.933Z</atom:updated>
            <content:encoded><![CDATA[<h4>Clarifying the Terminology</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_1qkAfp99Qb0Xu5I99UYgg.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@mihaly_koles?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Mihály Köles</a> on <a href="https://unsplash.com/s/photos/industry-plant?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><p>Maintenance can be a complex domain. There are several categories of maintenance tasks with varying descriptions, including: preventative, corrective, predictive, proactive, scheduled, unscheduled, condition based, and reliability centred maintenance.</p><p><strong>In this blog</strong> we will put forward a <strong>taxonomy</strong> of maintenance tasks that not only includes a description of each category, but also when and why they are applicable.</p><h4>International standards</h4><p>‘BS 13306:2017’ is an international standard that defines maintenance terminology. It provides taxonomies of both maintenance types and activities; however, it is not definitive as it does not fully describe defect elimination or failure finding tasks, nor does it recognise that conducting no-scheduled maintenance (NSM) is a deliberate choice.</p><p>This taxonomy is justified as the scheduled elements are provided with an idea of how periodicities for scheduling the tasks are derived, and which tasks are applicable to specific failure patterns. This breakdown allows the measurement of effectiveness and efficiency of the tasks.</p><p>The following diagram shows the taxonomy used, which will be further explained below…</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/941/1*k6XVthYuDa60DOGdg2nkfQ.png" /><figcaption>source: author</figcaption></figure><h3>Preventative maintenance</h3><p>The scheduling of preventative maintenance implies repeating the task at a periodic interval. This could be:</p><ul><li>Calendar based. This may be relevant if machinery is run 24/7, or we are dealing with failure modes that are driven by environmental exposure. Maintenance can be set daily, weekly, monthly or annually as appropriate.</li><li>An operational usage basis. At its simplest elapsed operating hours, or cyclic based.</li><li>A hybrid. More than one influence on the deterioration of condition can be blended. An example might include a mix of exposure and operating age.</li></ul><p>We will now look at each type of preventative maintenance…</p><h4>Scheduled Restoration</h4><p>This refers to a periodically scheduled maintenance intervention that allows the condition of a component to be restored to a nominal value. The condition should slowly deteriorate over operating time, and the periodicity is set before machinery functions are seriously degraded.</p><blockquote>An example of a scheduled restoration task includes cleaning carbon dust from rotating electrical machines’ carbon brushes used in commutators or slip rings when the task falls due. The carbon dust accumulates and gradually reduces the insulation resistance in the electrical windings. The task of cleaning the carbon dust restores nominal values of insulation resistance.</blockquote><h4>Scheduled Replacement</h4><p>This refers to a periodically scheduled maintenance intervention that relies on the condition deteriorating over time or age, where the probability of functional failure substantially increases after a period of useful life. The component cannot be restored on the maintenance line and is swapped out for a new or an off-line refurbished component. The old component is replaced regardless if it is still functional, which implies the condition of the component is not measurable when it is in service.</p><p>Scheduled restoration or replacements are only applicable if the component failure patterns are wear-out. There needs to be an initial period of useful life before the component finally wears out and there is a rapid increase in the probability of failure. The periodicity for timing the task is set to equal the end of useful life. Many maintenance regimes mistakenly apply scheduled replacement tasks where the component’s failure pattern is either random or premature. This is wasteful and increases costs for no benefit.</p><blockquote>An example of scheduled replacement includes filters that cannot be restored on-line, which are changed out for new after a period of operation.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0DRNtTfd-C3bcw4NnpmIaw.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@dominik_photography?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Dominik Vanyi</a> on <a href="https://unsplash.com/s/photos/mine-truck?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><h4>Condition based — Predictive Maintenance</h4><p>Condition based maintenance does not prevent failure; instead it detects the onset of failure or tracks gradual deterioration of condition so we can take early recovery action. It enables us to avoid the worst consequences of full failure.</p><p>Predictive maintenance is a type of condition based maintenance where we rely on fixed sensors and have access to the sensor time series data. Because the data monitoring and analysis is continuous there is no scheduling. The data acquisition and processing provides a high degree of automation so beyond the initial investment in the software the ongoing costs are low. The data is analysed in order to:</p><ul><li>Diagnose that a component is failing. At any point during operational life, a failure initiation can trigger the rapid deterioration of a component to a point of functional failure. The failure is detected using the time series data to spot the symptoms or causes.</li></ul><blockquote>An example may be a bearing that suffers a shock overload event that initiates failure that could then be detected using fixed vibration and temperature sensors.</blockquote><ul><li>For longer term degradation. Where there is no initiation of failure, and the deterioration starts as soon as the equipment enters service. The deterioration occurs because of exposure to stress, wear or environmental factors. There may be a point where the degradation is possible to measure. This may be as soon as the asset starts its operational life or it might be later.</li></ul><blockquote>Examples include structural cracking due to cyclic fatigue — this may only be practically measurable when cracks grow to the surface of structures where they can be conveniently detected. Additionally, corrosion in accessible places may be observed and measured at any time during the service life of the asset.</blockquote><ul><li>After diagnosis or the earliest point that degradation can be measured, we can prognose the interval until functional failure, referred to as the ‘Remaining Useful Life.’ Prognosis is aided by the captured sensor time series data, and can be enhanced as deterioration progresses and new symptoms develop.</li></ul><blockquote>An example includes the bearing failure. The initial vibration based symptom that is used to diagnose the failure may be followed, after some time, by the bearing temperature increasing. As the bearing deteriorates it generates more heat. The time gap between different symptoms presenting should be predictable and yields prognostic information.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*znT2xvMPPJY-f7uBkDX6kw.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@gautamkrishnan?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Gautam Krishnan</a> on <a href="https://unsplash.com/s/photos/freight-rail?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><h4>Condition based — scheduled inspections, sampling, NDE or NDT*</h4><p>With scheduled condition based maintenance the periodicity of the task is not related to useful life; instead, it is a function of the P-F interval. This is the time from initial diagnosis (designating a Potential failure, the P point) to the time of Functional failure, the F point.</p><p>The periodicity for non-safety implicated failure modes is half the P-F interval; whilst safety implicated failure should be one third of the P-F interval. When used effectively, once the failure is diagnosed there should be sufficient time (remaining useful life) to plan and predispose resources for recovery. The time for taking the machine out of service may also be optimised, before final failure, minimising operational disruption. The recovery task effectively becomes a deferred corrective maintenance task.</p><p>Condition based maintenance is applicable for all of patterns of failure and is the only practical choice for the random failure pattern. For the premature failure pattern, the primary maintenance response should be to conduct Root-Cause Analysis, but on-condition maintenance may also be done as a palliative action. This attempts to avoid final failure consequences before the causes of premature failure are eliminated.</p><p>*NDE is Non-destructive Examination, NDT is Non-destructive testing</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Op8H4yaK4L_DfwGua5jfWA.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@dominik_photography?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Dominik Vanyi</a> on <a href="https://unsplash.com/s/photos/mine?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><h4>Failure finding — scheduled</h4><p>This task is especially important to conduct for protection alarm or interlock systems that can fail but go unnoticed by the operating and maintenance crew conducting their normal duties. If the equipment suffers a failure, whilst the protection device which usually protects against such failure, has itself failed, then the consequences of the two failures is likely to be grave.</p><blockquote>An example being if a home fire or smoke detection system fails, and a subsequent fire causes casualties through smoke inhalation because there was no alarm.</blockquote><p>The periodicity of conducting failure finding tasks may be determined by calculating the joint probability that both failures occur in a set time period. The resultant probability is then reduced to an acceptable level (usually set by an organisation’s safety policy). Failure finding tasks are usually functional checks, introduced periodically to reduce the risks of undetected failures.</p><blockquote>For the example of the fire alarm there will usually be a test button, or the function can be tested with a match. Manufacturers will recommend the periodicity of the checks.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KaVlyuufHaOJhWzbFHlBhA.jpeg" /></figure><p>During product design a great deal of effort is spent to reduce intrinsic safety risks to an acceptable minimum. Safety and protection devices are usually designed to ‘fail safe’ or fail in an obvious way, so it can be easily noticed. Sometimes designing safe is not possible, and redundancy is introduced with many separate protection systems which do not have common failure modes working in a voting system. This might mean two out of three protection devices have to fail or detect a dangerous measurement to trip the equipment.</p><p>In the Maintenance gurus experience, within many organisations failure finding tasks are not recognised for how vital they are for safety. Many regimes carry out physical safety checks under a different management system from a CMMS and resultantly, these checks are not recognised as maintenance.</p><blockquote>In a previous blog we used an example of a pilot walking around their aircraft checking for any possible failures. Although it is not usually recognised as such these are classic failure finding maintenance tasks.</blockquote><p><strong>How many maintenance regimes clearly designate which of their tasks are failure finding, so they do not mistakenly change the periodicity of checks, because they rarely find problems thereby inadvertently introducing a safety risk?</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fXntoM5zcg68pem8QoI6Jg.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@kangterbang?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Pandu Agus Wismoyo</a> on <a href="https://unsplash.com/s/photos/maintenance?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><h4>Servicing and zonal checks</h4><p>These checks are usually fairly frequent; many are done of a daily or shift basis. The tasks include checking and replenishing consumables, lubrication and cleaning. Cleaning is very important as the risks of contamination or fire increase in machinery spaces if cleaning is neglected.</p><p>Zonal checks usually involve experienced operators or maintainers walking around sections of their machinery, using their five human senses and experience to look for anything out of the ordinary. This is normal working practice for maintainers. Experienced maintainers develop a tacit skill to notice anomalies which can then be acted on. The proportion of incipient failure found by these practices is surprisingly high. However, these tasks are not captured in the CMMS and are unlikely to be accounted for financially. This can result in an unexpected decrease in reliability if drastic manpower reduction is implemented.</p><h4>No Scheduled Maintenance (NSM)</h4><p>This inclusion may seem strange; however, it is important to record that components have been analysed and the risks of failure are low, coupled with a lack of maintenance tasks that are practical or cost effective. This is an active decision and is justified by the analysis work. If the risks and consequences of failure are unacceptable and maintenance isn’t practical, further modifications might be required to reduce risks to acceptable levels.</p><h3>Corrective Maintenance</h3><h4>Deferred corrective maintenance</h4><p>This is where a failure has occurred, but the consequences may be small enabling operations to continue, possibly at a reduced performance until recovery can be planned and resources predisposed.</p><blockquote>An example may be the failure of redundant equipment, or if the condition-based maintenance has enough ‘Remaining useful life’ to reach a planned outage.</blockquote><p>Usually, equipment has a routine of being periodically released for planned maintenance, including short, frequent, planned outages or shutdowns. Many deferred corrective maintenance tasks can be completed during these periods.</p><h4>Immediate corrective maintenance</h4><p>This is where the recovery from a failure is urgent as the consequences of the failure may cause operational disruption. For mobile equipment the operating context may dictate that recovery is urgent and must be continuously worked on until completed.</p><blockquote>A ship in a storm with loss of propulsion would be an example.</blockquote><p>Components that require immediate recovery may be better candidates for predictive maintenance, or deferred corrective maintenance if the RUL is sufficient.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qtWtkAwtWewP33HVJLPiAg.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@tdederichs?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Torsten Dederichs</a> on <a href="https://unsplash.com/s/photos/ship-in-storm?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><h4>Modification</h4><p>Modifications can include design change of components to avoids or protect against failures, increasing intrinsic reliability. Modifications can also include changing maintenance or operations. Perhaps the cadence of frequent planned outages is not optimal for reliability, or the duty cycle is putting too much stress on the equipment. The operating environment may also cause problems and may require modifications to enclose equipment and protect it from environmental stressors.</p><blockquote>An example of maintenance that modifies the operating environment: in an open cast mine site suppressing dust and maintaining smoother mine road surfaces to reduce vibration for mobile equipment.</blockquote><h3>Defect Elimination</h3><p>Defect elimination is sometimes called proactive maintenance. It is a constant improvement effort that looks for common preventable causes of failure. These causes can be eliminated by modification or improving quality. If the causes are problems with design or manufacturing, then suppliers may be contacted to determine if Warranty claims are appropriate, or whether modifications from the manufacturer are possible.</p><p>Defect elimination could be run in two basic ways:</p><p>1. Root Cause Analysis on all components with a premature failure pattern, or random pattern, where the engineering staff believe the component should have a wear-out pattern.</p><p>2. A quality management project approach looking at how operations, maintenance or logistics handling can be improved. This could be part of adopting a ‘Total Quality Maintenance’ approach to maintenance management.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*d4AYpTOmDf0hRMG-DObyXw.jpeg" /><figcaption>pixabay</figcaption></figure><p>This blog has attempted to clarify maintenance terminology, by delving into a task classification taxonomy and providing explanations of applicability.</p><blockquote>We would love to hear how you <strong>classify your maintenance tasks</strong> in a CMMS system? How do you know if the <strong>task types are applicable</strong> to the failure behaviour of the components? How do you <strong>appropriately set periodicities</strong> using a principled method for each type of task?</blockquote><p>How could we know which pattern our component failure modes are failing in? The answer is in one of the most powerful techniques we can deploy in reliability engineering, involving Weibull analysis.</p><p>In the <strong>next blog</strong> we will discuss the <strong>Weibull process</strong>, and further pre-requisites for applying the different types of maintenance we have covered here. We aim to discuss how Reliability Engineering forms a solid foundation for mature maintenance management.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=29083df05038" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ox-mountain/maintenance-task-classification-29083df05038">Maintenance Task Classification</a> was originally published in <a href="https://medium.com/ox-mountain">The Maintenance Guru</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>