An in depth look at which maintenance metrics matter, how to measure them and what you can gain from optimising.
In an earlier article, we looked at 'What good looks like?' in the world of maintenance and more particularly, within the world of maintenance metrics.
That article focused on the area of contractor management (i.e. how to monitor and measure the performance of your supply chain). Here, we have a more in depth look at the maintenance metrics relevant to plant and equipment.
In this article, we look at 'How to save money, reduce downtime and make your life easier as a facilities manager'. Rather than looking at how technology can save you money as a facilities manager, here we examine what sort of things you can measure in the pursuit of optimising your maintenance spend.
Whilst part of the positive effects of our data driven maintenance strategy will be qualitative in nature (e.g. hopefully you will be less frantic), the benefits should also be demonstrable to other stakeholders. In order to clearly illustrate the effects, you first need to create a baseline.
For more information on how to capture this data, and what can be done with it once it is gathered, have a look at our guide on the benefits of digitalising your facilities management operations. If anything is still not clear, just react out - we would be happy to help.
Whilst you can look at industry-wide data to benchmark your own operations, ideally you have access to a set of data that represents your plant and equipment because without it, you really are flying blind.
KPIs are a great way of measuring the performance of your contractors (or projects, or people). KPIs are generally used as a goal toward which these parties should strive. In general KPIs are an aggregation of underlying data points, and very often these data are metrics. For example, you may set a downtime objective (KPI) of less than 1% of operating time for your maintenance engineers. This level of downtime is in turn affected by such things as time to site; percentage of first time fixes; percentage of jobs fixed with available parts; time for parts delivery etc. All of these are metrics.
Metrics are therefore an output of actions, rather than an objective of actions. It is a subtle difference, but hopefully it will become clearer as we provide a few more examples.
We will also look at the constituent parts of these metrics above: asset, operations and inventory based metrics. Let's jump in ...
Whether you operate in the hospitality, hospital or manufacturing sectors, your asset infrastructure is absolutely critical to delivering your product or service to customers. Without your plant and equipment functioning properly, your ability for your company to earn revenues is either compromised or eliminated. As such, plant and equipment downtime has a cost. The cost changes from one company to the next but you can calculate this by looking at the lost productivity or sales that occur when plant or equipment is not functioning as it should. It will likely differ from one piece of equipment to the next.
So, what are some of the metrics that can look at in assessing asset performance?
As the name suggests, this is a measure of the average period from one instance of equipment failure to the next. You can look at the mean time between failures either in aggregate (the time between any type of failure) or for specific failure types to examine is there is an isolated type of issue that should be investigated further. Where failures result in specific issues, for example those related to health and safety, may also warrant explicit investigation as these will have potentially bigger implications.
It should be noted that mean time between failures does not apply to planned or scheduled downtime of equipment, for example for regular servicing, regulatory inspections or preventative maintenance such as parts replacements.
Mean time between failures is calculated by dividing the number of operational hours for a piece of equipment by the number of failures. This is looked at on a periodic basis, for example a calendar year.
For example, a piece of machinery may operate for 4,000 hours in a year. Over this time, the equipment failed 5 times. The mean time between failure is therefore 800 operating hours.
Often O&M manuals or maintenance schedules will give estimates of mean time between failure. Whilst obviously better than nothing, this is usually based on equipment that operates in a wide variety of environmental and production environments. As such, it is useful only as a rough guide. Much better would be to track this for your own plant and equipment and see how it is trending over time. To do this, obviously you need your own data which takes us back to Step 1. Measure.
Mean time between failures is used to create budgets and inform maintenance schedules. As above, it is most useful when looking at the cause of the failure and see if there is something deeper that needs to be investigated. It can also be used as an input in creating a predictive maintenance strategy where other factors such as the variability of the times between failures (i.e. how regularly do failures occur) will also be used to generate predictive works orders.
Now that we know how long (on average) it is between failures, now the question is 'How long does it take to get the issue fixed?'. For this, we can look at the mean time to repair. The mean time to repair is a measure of how long (on average) is required to become aware of the failure, identify the cause of the issue, and repair the failed equipment. It is essentially a measure from when the equipment went down until the moment the equipment is returned to production.
As can be seen above, there are a few factors which influence mean time to repair. The first is how long it takes to notify the appropriate parties (such as maintenance technicians) and in turn how long it takes them to arrive on site or remotely diagnose the issue. It may also be affected by the time to remove the equipment from its place in the production process, have it cool down, disassemble components, test etc.. Generally, this metric only relates to instances where equipment can be repaired without parts or based on stock inventory. Where parts need to be ordered this is typically treated in downtime costs, but should be tracked in any event. Mean time to repair is designed to capture a company (or its supply chain's) responsiveness to issues.
Mean time to repair is calculated as the sum of the total time from when issues were reported to when the job was successfully completed and the equipment was online once more. This is then divided by the number of failures.
For example, perhaps there were 2 instances of failure during a calendar year. On each occasion it took the engineers 1.5 hours to arrive on site and the works took 3 hours until the equipment was working again. The total maintenance time is therefore 2 x (1.5 +3) = 9 hours and the mean time to repair is 9 / 2 = 4.5 hours.
First rate responsiveness is often deemed to be in the 4 to 5 hour range with a remote supply chain but can be required to be as low as 15 minutes with on site staff on critical infrastructure such as in a data centre.
Like many of these metrics, mean time to repair needs to be looked at in context in order to assess whether there are issues that need to be addressed. For example is the equipment design such that repair times are longer than desirable? Would you be better off with on site engineers given your equipment location? These metrics can inform such discussion.
As above, one of the benefits of collecting these metrics is the ability to strategically consider whether plant and equipment choices are the right ones and how these data are trending over time. Whilst nominal maintenance costs are important, it is often useful to think of these in the context of whether you should replace equipment: maintenance as a percent of replacement value allows you to do just this.
This metric informs you as to how much you are spending on maintenance compared to the value of an asset. At its extreme you can use the example of a toaster. It would not be worth calling out an appliance technician to repair a faulty element in a kitchen toaster. The callout costs (if required), parts and labour would be a multiple of the cost of replacing the toaster.
Whilst this is a simplistic example, the same principle applies when looking at complex plant and machinery. Are maintenance costs increasing year over year? Have they jumped versus last year? These sorts of questions can allow you to optimise your asset replacement strategy. You may also need to include other factors such as disposal, lead times, installation costs etc. but it is still valuable information.
Maintenance as a percentage of replacement value is calculated by dividing the total amount spent on maintenance over a period (e.g. a calendar year) by the replacement cost of the asset. The maintenance costs should include components such as parts, labour and any other expenses. Replacement value should be the amount required to install an equivalent piece of equipment.
For example, imagine you spent £80 on maintenance in total, on a piece of equipment that would cost £2,000 to replace. The maintenance as a percent of replacement value is therefore 4% (80 / 2000 * 100%).
Well maintained equipment as part of considered asset strategy should target maintenance as a percentage of replacement value level spend of between 2 - 3%.
As with these other metrics, this is something that will give you information that warrants further investigation. Should you replace a certain piece of equipment? Should you change your maintenance strategy to 'run to destruction' or is it worth investing in more preventative maintenance? This metric can help you make these decisions.
This is more of a catchall metric - and one more relevant for those operating in manufacturing or production environments - but an important one nonetheless. It is a summary metric for looking at how often your equipment is performing in the way in which it was designed to (or in the way you want it to). In most instances, this is a measure of the percentage of time that the equipment was available and being used productively. As such it is a function of three components: asset availability, production quality and asset performance.
It may be easiest to look at this in terms of extremes. An overall equipment efficacy metric of 100% would mean that an asset is always available (has no downtime) and is producing acceptable quality output (within tolerances) at an acceptable (or optimum) speed.
In order to calculate overall equipment efficacy, you therefore need to look at each of the elements. Availability is calculated by dividing total run time by total production time. It excludes planned maintenance periods, holidays, shutdowns due to lack of demand etc.
Performance is calculated by dividing the system throughput by the maximum possible throughput.
Quality is in turn calculated by dividing the number of usable pieces of work produced by the total units started.
We can therefore calculate the overall equipment efficacy as: Availability (%) x Performance (%) x Quality (%).
World class efficacy rates are in the region of 85%.
As with many of the other metrics, this is a first order number and this means that the number will lead you to investigation before action. Which of the inputs are causing the decline (or low) efficacy levels? Are there links between performance and quality levels? These may be worth exploring.
Now that we have explored the equipment side of the maintenance process, we must also consider that real life humans are an important part of any productive environment. Any maintenance strategy requires the involvement of people but in the pursuit of efficiency you can also measure how they are performing in terms of decision making and its impacts.
Here are some of the ways in which you can measure the performance of people in facilities management.
One of the most impactful decisions that is made by facilities managers is the type of maintenance strategy to use. For many, there are two types of maintenance: planned and unplanned. At Homyze, we are also incorporating predictive maintenance which uses real time data to determine maintenance requirements based on representative datasets but this is not yet widely used outside the manufacturing space.
We have outlined elsewhere on this site what constitutes planned maintenance. The objective of planned maintenance is to keep plant and equipment in a functional state. There are of course tradeoffs to be made between the costs of maintenance and their benefit, but this metric allows you to look at how long is spent on planned maintenance versus unplanned maintenance. We will look at the costs elsewhere but this metric gives you insight into the effectiveness of the planned maintenance in avoiding failures.
Planned maintenance percentage is calculated by dividing the total number of hours spent on planned maintenance versus the total number of hours of maintenance (both planned and unplanned) over the same period.
The higher the planned maintenance percentage the better. This means that this is having the effect of avoiding serious issues. Target levels should be a mix of 70% planned maintenance and 30% reactive (excluding any component of predictive maintenance).
The planned maintenance percentage outlines the effectiveness of planned maintenance but where it is really felt is in the overall maintenance spend. We will look at this below, but in short the reason it is so important is the cost of unplanned maintenance in relation to planned maintenance. Studies indicate it can cost as much as 8x what planned maintenance costs. From a facilities manager's perspective, this is only half the problem. It is much more time consuming to organise and leads to very unhappy clients.
Depending on the industry in which you operate, compliance can have different levels of importance. When one considers care homes, pharmaceutical companies and the like, you can imagine compliance plays a pretty important role. It is one thing to identify and outline what constitutes compliance. Next, you need to ensure that these compliance jobs are being done.
Compliance jobs often involve the submission of reports following site attendance and may require off site testing and analysis. As such, you should only include jobs where every required step in the process has been completed.
Compliance levels can then be calculated as follows: number of completed compliance works orders / total number of compliance works orders scheduled for the period x 100.
For example, if there were 10 compliance jobs required during the month and only 8 were successfully completed (for example you were still waiting on documentation for 1 job, and the other was missed completely), the compliance level would be 80%. It should also be noted that often remedial works are required following compliance jobs and that this should constitute part satisfactory completion.
It may be useful to look at compliance levels at individual properties or types of compliance work (e.g. fire door surveys versus emergency light testing) to decide on the appropriate actions but it is a metric that is worth monitoring.
Obviously the targeted compliance level needs judgement depending on the potential impacts of non-compliance but the best run compliance programs would be looking for levels of 90% or more.
Similar to the compliance levels, the open works order percentage is a metric that is designed to give insight into how well works orders are being managed. Time has been spent on creating maintenance schedules, compliance schedules and the resultant work orders. This time and effort is wasted unless jobs are attended, reports are obtained and the appropriate follow up actions are taken.
In order to ensure that these appropriate actions are taken, the preceding steps such as site attendance, diagnosis, job reports, invoicing etc. should have occurred. Until these have been done, jobs cannot be marked as completed.
So how is this calculated? It should be available within any CAFM, but the metric is obtained using the following formula: number of works orders not in a completed state divided by the total number of works orders issued for attendance during the period.
There may need to be a delay before this can be calculated based on SLAs that you have agreed with your supply chain (for example that job reports or invoices are submitted within 48 hours of attendance) but a well managed maintenance program should be shooting for an open works orders percentage of below 10%.
When considering the impacts of these decisions taken by facilities managers such as which maintenance strategy to employ and which suppliers to engage the most obvious all encompassing metric is that of total maintenance spend.
Thankfully, it is one of the easiest to calculate (but requires the submission of all relevant information). Again, this should be a data point (metric) that can be pulled from any comprehensive CAFM and can be filtered based on individual asset types, or properties or service lines. Given this condition (or filter) it is then calculated as the total amount spend on maintenance (both planned and unplanned) within a given period of time.
The total maintenance spend is not a terribly useful number in isolation. To be of most benefit to facilities managers and other stakeholders, these data should be looked at by comparing periods in order to try and isolate the impact of decisions taken by people.
Has a switch of supplier improved our compliance levels? With this change in maintenance strategy from preventative to run to failure, have we increased or decreased our total maintenance spend? The answers to these questions can give you insight into how your people are performing.
Unlike some of the other metrics mentioned above, inventory ratios and metrics tend to be looked at over a longer period of time, often a year. Depending on who has responsibility for inventory spend it may also be something that falls on your suppliers or may be something that you are managing for yourself. Either way, it is another opportunity to to improve spend efficiency and deliver returns for stakeholders. Having the right parts in the right locations at the right times improves a number of the other metrics and can give you information about the root cause if other metrics are blow target levels.
Here are ways in which you can think about inventory spend.
The stock turn ratio is used as an indicator of the efficiency of spend on inventory. Captured in this metric is whether you are investing in parts that are not being used and therefore money is tied up in stock sitting on shelves. It is calculated on the basis of the value of inventory which can lead to some distortions but in general higher cost parts are less likely to be kept in stock.
Specifically, the stock turn ratio is calculated by dividing the value of inventory issued in the past year by the value of stock on hand (inventory) at the point at which the ratio is calculated.
As an example, if a company has bought £1,000,000 worth of inventory over the year and at the point at which the measure is taken it has £500,000 worth of inventory. This would be a stock turn ratio of 2.0x (£1,000,000 / £500,000 = 2).
The most efficient managers of inventory should be looking for stock turn ratios of 2.5x but it is not uncommon to see ratios less than 1.0x. Like these other metrics, it is used to inform discussion and identify where to investigate opportunities for improvement. If inventory is a significant investment for your company and you have a low stock turn ratio such as 0.5x, this may be something that needs to be addressed.
What is worse than slow moving parts inventory is redundant or obsolete parts inventory. This is where parts are no longer able to be used for repairs or replacements because they are no longer compatible or have passed their useful life period. This is wasted money and while some elements are out of an operator's control, for example if customers update their plant and equipment and parts are no longer appropriate, it is something that should be tracked closely.
Unfortunately, this is also a completely backward looking metric in that there is little to anything that can be done about this wasted spend, but it may prompt purchasing processes to be reconsidered.
The redundant inventory percentage is calculated by dividing the amount spent on inventory that must be written off or down (the wasted spend) divided by the total amount spend on inventory during the period e.g. a year. For example, if a company spent £100,000 on inventory that had to be disposed of and the total inventory spend during the year was £1,000,000 this would be a redundant inventory percentage of 10% (£100,000 / £1,000,000 = 10%).
As is no doubt clear from the above, maintenance metrics are only the means to an end. They are important to monitor so that you can be made aware of issues that need to be investigated further. If they have spiked or cratered, it can be an indication of something that needs to be changed.
Different companies will weight each of these metrics in a different order. For example as a client side facilities manager who has outsourced their maintenance on a TFM basis, stock turn ratio will clearly not be a consideration.
Having the information available is the most important first step. Once you are tracking these data you will get a sense of where they are and be alert to numbers that fall outside of trends.
As may also be clear, you can amend these metrics as appropriate to make them more relevant to your company. If you can get the information, you can create a metric around it and if it is something that is important to how your facilities function you are the best person to make these judgements.
Collecting data is not the objective. Driving improved outcomes is actually the goal. Having the numbers available is of no use unless it drives change at your company. The change can be used to target different KPIs such as reduced downtime, higher profits, improved health and safety or the like. What you can be sure of is that the effectiveness of your decisions will be increased with the availability of metrics.
Collecting data should not be intimidating. Ideally, it should be maintained within your existing systems and software. You may need to set up some processes, but with this, Homyze is happy to help.
If instead you just want to talk about efficiency savings in FM, reach out to us - this is one of our favourite topics!
Homyze provides a guide on what is planned preventative maintenance (PPM) and how to go about implementing this maintenance strategy
Smart FM means different things to different people. Here we explain what smart facilities management means to us at Homyze.
Homyze looks at what predictive maintenance is, what is needed to calculate it, and what are the benefits of using
Radio Homyze: AM / FM (Asset Maintenance & Facilities Management) provides you with the latest industry information.