Developing Performance Indicators for Managing Maintenance

1. 1Define wrench time for your site before you measure it. Direct, value-adding work on the asset counts. Waiting on or looking for parts, rental equipment, equipment shutdown, drawings or tools, plus travel and getting instructions, do not. The passages give no wrench-time percentage or benchmark, so your first measurement is your own baseline.
2. 2Build a delay-cause list from the source categories: parts, rental equipment, equipment not shut down, drawings, tools. Add travel and instructions. Put these on the work order or time log so a technician can tag lost time in one line instead of writing a narrative.
3. 3Track Percentage of Maintenance Work Orders Planned: work orders planned divided by total work orders received. Run it by department, craft line and individual planner. The source says these cuts 'can reveal some interesting trends', such as crafts with a lower planned percentage.
4. 4Match each delay cause to the planning step that removes it. Parts delay means the planner confirms and stages parts before release. Shutdown delay means the planner coordinates the window with operations. Drawings and tools delay means the planner attaches them to the job package. Rental equipment or contractor delay means the planner arranges it ahead of the job.
5. 5Test each planning investment against its payoff. A plan is worth the effort only if the planner's hour saves more than an hour of labor, or reduces downtime or the equipment required. Planning every job is not the goal.
6. 6Set the target planned level as a management decision. Weigh labor savings, downtime savings and reduced costs together. Expect the indicator to stay below 100% and watch it closely to confirm the right jobs are being planned.
7. 7Integrate planned jobs into the operations schedule so equipment is actually available when the crew arrives. Without that, the source says, equipment utilization cannot be optimized. This is the pit-crew test: the race is lost by not keeping the car on the track.
8. 8Keep the data tracking proportionate. Use a card file, white board or log book if the equipment count is small. Move to a computerized work order system if paperwork buries the workforce. Either way the data must be complete and analyzable.

The source supplies the decision arithmetic. A planner spends one hour planning a job. Case A: the plan saves one hour of labor and does not reduce the required downtime. The source judges this 'not worth the effort', because you spent an hour to save an hour and gained no capacity. Case B: the same one planning hour saves two or more labor hours, or cuts downtime or the equipment required. The plan is worth it. Apply this to wrench time. An unplanned job loses technician time to logistic delays: getting the equipment shut down, organizing labor, finding and delivering spare parts, and coordinating contractors. A planned job removes those delays before the crew starts, so more of the crew's hours go to direct work. Compare two jobs of the same type on your own log: one planned, one reactive. Tag every lost hour by cause, then compare the totals. Planning effort goes to the job types where the gap is largest. It does not go to every job.

• If a job's planning time is expected to save less labor than the hour spent, and downtime and equipment needs stay the same, then skip detailed planning and run it as minor or reactive work.
• If planning saves two or more labor hours, or reduces downtime or required equipment, then plan it.
• If one craft line or planner shows a noticeably lower planned percentage than peers, then investigate that craft or planner before changing the site-wide target.
• If delay tags cluster on waiting for or looking for parts, drawings or tools, then fix the planner's job-package checklist and staging before adding labor.
• If paperwork and file cabinets of equipment data burden the workforce and nobody has time to analyze them, then computerize the work order system. If the equipment count is small and tracking is easy, keep the manual system.

• Pushing the planned percentage toward 100%. The tell: planners spend hours on short jobs that save little labor and no downtime, so the planning effort costs more than it returns.
• Planning jobs without integrating them into the operations schedule. The tell: crews arrive with parts and drawings but the equipment is not shut down, so they still wait.
• Treating wrench time as a technician discipline issue. The tell: lost hours trace to parts, tools, drawings, rental equipment or shutdowns, which are logistics failures the planner controls.
• Tracking a site-wide planned percentage only. The tell: the average looks acceptable while particular craft lines or planners lag, which a department, craft and planner breakdown would show.
• Collecting data that is incomplete or unanalyzable, on cards, boards or logs. The tell: you cannot produce top-ten problem equipment or percent reactive vs. proactive reports, and delay causes stay anecdotal.

• ▸Work order number and craft line: identify the job and the trade performing it.
• ▸Planned or unplanned: mark whether the job was planned before release, for the planned-percentage calculation.
• ▸Planner name: record who planned it so planner-level trends are visible.
• ▸Hours on direct work: record the time the technician spent doing value-adding work on the asset.
• ▸Delay hours by cause: split lost time into parts, rental equipment, shutdown, drawings, tools, travel and instructions.
• ▸Planning hours spent: record the planner's time on this job.
• ▸Labor hours saved vs. estimate: record the labor saving attributed to the plan.
• ▸Downtime or equipment reduction: note any cut in downtime or required equipment from planning.
• ▸Worth-it check: mark yes if labor saved is two or more hours per planning hour, or downtime or equipment was reduced; otherwise mark no.

1. 1Fix the definition before you measure. Uptime is the time an asset is available to perform its designed function; downtime is the period it is not performing to design specification. That covers a full breakdown and an asset running below design specification. Write both conditions into the metric definition so nobody counts a degraded asset as 'up'.
2. 2Define 'required for service' per asset. Availability is measured against the time the asset should be operating, not total calendar time. Record the scheduled-required hours for each asset, so equipment that is down when it should be running is the only thing counted against it.
3. 3Log every downtime event with its start time, end time, and reason: breakdown, emergency, or idling and minor stoppage. The glossary notes that small interruptions are easy to fix or reset but add up to significant losses, so capture them rather than only the large failures.
4. 4Calculate uptime as the required hours less downtime hours, divided by required hours, per asset and per period. Report downtime as its inverse. Check the baseline against the source's benchmark: some companies see downtime average 10-to-20 percent or more.
5. 5Attach a cost to downtime. Start with the base items the source lists: utility costs, idle production or operations personnel, late delivery, and overtime to make up lost production. Then add the true cost, which is the lost sales dollars of product not made on time and is significantly higher than the base items.
6. 6Rank assets by their contribution to downtime. Use a Pareto analysis to find the roughly 20% of assets that create 80% of the problems, and target them first.
7. 7Route downtime causes to the practices that address them. Operations feeds back how the equipment performs, maintenance acts on it, and engineering buys and installs equipment that is not inferior. The source's NASCAR analogy holds that the whole team has to work together to get maximum capacity from the asset.
8. 8Review uptime as a forward indicator on a fixed cadence and use it to strengthen best practices or eliminate poor ones. The source contrasts this with financial statements, which are 'damage reports' that tell you only after the fact that you performed poorly.

The source supplies its own framing rather than a plant dataset, so this example applies its figures. A company's equipment downtime averages 10-to-20 percent, so the equipment is unavailable for that share of the time it should be running. Product deliverable to the market is restricted by that amount. Some companies respond by buying backup or redundant equipment. The source warns that this hurts the return on net assets indicator and lowers investment ratings in the financial community. The better route is to measure downtime by asset, price it, and reduce it. Pricing starts with utilities, idle personnel, late delivery, and overtime. It then adds the lost sales dollars, which the source says are significantly higher than those base costs. Because the source states that the pure maintenance contribution to profitability is dwarfed by gains from added capacity (availability), recovering even part of the downtime outweighs the maintenance savings. For example, 1/3 of a $3M maintenance labor budget wasted equals $1M. Even in markets capped on volume, downtime still raises costs and blocks the low-cost-supplier goal.

• If an asset is running but below design specification, then count it as downtime. The definition covers breakdown and failure to produce at design specification.
• If someone says downtime has no cost, then price the four base items (utilities, idle personnel, late delivery, overtime) and then the lost sales dollars.
• If downtime is being covered by buying backup or redundant equipment, then challenge it first. It lowers return on net assets and investment ratings, so fix the availability of the existing asset before adding capital.
• If many small stoppages are being dismissed as trivial, then accumulate them over time. The glossary notes that small causes become significant losses in total.
• If operations, maintenance, and engineering each blame the others for downtime, then treat it as a team failure. Require operations feedback to reach maintenance and engineering to stop buying inferior equipment.

• Counting only breakdowns as downtime. The tell is that an asset reports high uptime while producing below design rate. The definition also covers failing to produce at design specification.
• Measuring against calendar hours instead of required-for-service hours. The tell is availability figures that do not match what operations experienced when it needed the asset.
• Ignoring small stoppages. The tell is a log that shows only major failures while the line is frequently reset or idled.
• Declaring downtime free, or costing only the base items. The tell is a downtime figure made up of utilities and overtime with no lost-sales line, which understates the true cost.
• Using uptime only as an after-the-fact report. The tell is that the figure is reviewed monthly like a financial statement and prompts no change to a practice. The source calls such lagging measures damage reports.

• ▸Asset / ID: Name the specific equipment item being measured.
• ▸Required-for-service hours (period): Enter the hours the asset should be operating, not calendar hours.
• ▸Downtime hours - breakdown: Enter hours the asset was down from failure or emergency.
• ▸Downtime hours - below design specification / idling and minor stoppage: Enter hours the asset ran but not at design specification, including accumulated small stoppages.
• ▸Uptime % = (required hours - total downtime hours) / required hours: Calculate it and record the result; downtime % is its inverse.
• ▸Base downtime cost: Sum utility costs, idle production/operations personnel, late-delivery cost, and overtime to make up lost production.
• ▸Lost sales dollars: Enter the value of product not made on time; this is the true cost.
• ▸Pareto rank: Mark whether the asset falls in the roughly 20% of assets driving 80% of problems.
• ▸Cause and owner: Note the cause category and which function (operations, maintenance, engineering) owns the corrective action.
• ▸Practice to strengthen or eliminate: Name the best practice to reinforce or the poor practice to drop, plus the review date.

1. 1Fix the baseline first. Write down the design specification for each asset: designed speed, capacity and rate of output. The source defines downtime as any time the asset is not performing to design specification, so a running asset below spec is already in a loss state. Without this number you cannot measure efficiency.
2. 2Separate the loss categories. Breakdowns are unexpected interruptions of service, and they belong to availability. Reduced speed and minor stoppages while the asset is running belong to performance efficiency. Keep them in different columns, because OEE compares availability, performance efficiency and quality rate as three distinct components.
3. 3Measure the loss that is usually left unmeasured. TPM-related Japanese studies cited in the source say efficiency losses are always greater than pure downtime losses, and most are never measured or reported. Start recording actual rate against design rate on every asset you track, not only the stop events.
4. 4Rank assets with a Pareto. Use the source's definition: find the roughly 20% of assets that create 80% of the problems. Start the efficiency work there, because chronic losses concentrate on a few assets.
5. 5Convert each efficiency gap to dollars. Take the shortfall against design and value it in lost production per year. The compressor example prices a 61%-efficiency gap at about $5.4M annually. A dollar value lets you compare the loss with the cost of fixing it.
6. 6Find the root cause using maintenance records together with financial data. The source says chronic efficiency problems surface only when accurate maintenance records are kept and combined with financial data. Check that labor and work are actually recorded in the CMMS/EAM, for example by reconciling CMMS maintenance labor cost to accounting.
7. 7Cost the fix completely and test payback. Include labor, materials and the production lost while the asset is down for the work. Compare that total to the annual value of the efficiency loss.
8. 8Sequence the work to protect output. If the fix means an overhaul, check whether it can be done asset by asset (serially) to avoid a total shutdown. Then track efficiency after restart to confirm the recovered output is real.

A company examined the efficiency of the gas compressors on its off-shore operation. Because of age and internal wear, the compressors were running at only 61% efficiency. The company valued that gap at about $5.4M a year. The overhaul was costed at $450K, and that figure included labor, materials and the production lost to downtime. Management chose to overhaul the compressors one at a time to avoid a total shutdown. The overhaul paid for itself 28.1 days after restart, and the $5.4M of increased production was realized over the next 12 months. The sequence to copy is: measure efficiency against design, put a dollar value on the gap, cost the fix including lost production, and schedule the work to keep the plant running.

• If you track only breakdowns and downtime, assume you are missing the larger loss. The source says efficiency losses always exceed pure downtime losses and are mostly unreported, so begin measuring running-rate shortfalls.
• If the annual dollar value of the efficiency gap far exceeds the full overhaul cost (labor, materials and lost production), then fund the overhaul. The compressors had a $5.4M annual loss against a $450K fix.
• If a full shutdown for the fix would itself destroy output, then stage the work serially, as the compressor overhaul did.
• If the loss could stem from raw materials, production scheduling, quality control or operator error, then do not assign it all to maintenance. The source lists these as other causes of downtime, so isolate the maintenance-attributable share before sizing savings.
• If the loss is chronic and nothing has broken, then pull maintenance records and financial data together to find the root cause. Do not wait for a breakdown to force the issue.

• Equating asset performance with uptime. The tell: reports show only breakdown hours, and an asset that runs slowly all shift looks fine.
• No design-specification baseline. The tell: nobody can state the asset's designed rate, so reduced-speed running is judged against 'what it usually does'.
• Letting chronic losses persist until a breakdown. The tell: the efficiency problem is first discovered during a failure investigation, not from routine data.
• Costing the fix without lost production. The tell: the overhaul estimate lists only labor and materials, so payback looks better than it will be.
• Incomplete maintenance data. The tell: CMMS labor costs do not reconcile to accounting (the source notes many companies reach only about 50% CMMS utilization), so root-cause analysis lacks records to work from.

• ▸Asset / Pareto rank: name the asset and its rank among the roughly 20% of assets driving 80% of problems.
• ▸Design specification: record designed speed, capacity and output rate.
• ▸Actual running rate: record measured output rate while the asset is running.
• ▸Efficiency vs design (%): actual divided by design, as in the 61% compressor figure.
• ▸Loss log: list reduced-speed periods and minor stoppages, kept apart from breakdowns.
• ▸Annual value of the loss ($): lost production valued per year.
• ▸Root cause and evidence: state the cause and the maintenance records and financial data supporting it; note any non-maintenance contributors (raw materials, scheduling, quality, operator error).
• ▸Fix cost ($): sum labor, materials and production lost during the work.
• ▸Payback and sequencing: calculate days to recover fix cost after restart, and say whether the work can be done serially to avoid total shutdown.

1. 1Split the maintenance budget into its two main divisions, labor and materials, then add maintenance contracting as a third line. The source treats these as the direct-cost base, and each needs its own baseline.
2. 2Break labor spend into overtime, outside contractors, and deferred maintenance performed without additional expenditure. These are the places the source says savings appear, and they are not necessarily headcount reductions.
3. 3Size the opportunity with the source's rule of thumb: about 1/3 of maintenance expenditures are wasted through inefficient and ineffective use of resources. Treat 1/3 as a hypothesis to test against your own records, not a result.
4. 4Apply the same 1/3 test to the materials budget. The source says materials are typically a similar amount to labor, so the combined savings can approach 2x the labor savings.
5. 5Discount the contracting line. Contractors already use labor and materials more efficiently than reactive in-house crews, so expect savings there, but smaller than in-house.
6. 6Shift the work mix from reactive to planned and proactive. Support it with the structure the source names: work flow system, CMMS/EAM, stores and procurement, predictive maintenance, and operations involvement. Reactive-to-proactive conversion is where the waste is removed.
7. 7Set a 3-to-5 year timeline, not a fiscal-year target. The source says the transition is not technically hard; the time goes to changing the culture so maintenance is treated as a core business process.
8. 8Record savings as expense dollars no longer required, and track them against the baseline by line: overtime, contractors, materials, deferred work. Keep this separate from capacity and downtime gains, which the source says are a different and larger contribution.

The source gives a $3M annual maintenance labor budget. If 1/3 is wasted, $1M can be saved each year. That saving need not come from headcount. It can come from less overtime, less use of outside contractors, or deferred maintenance done without added spend. Studies cited in the source put the materials budget at a similar $3M. A 1/3 reduction there brings combined savings close to $2M per year, and the source says these are real expense dollars that translate to profit. The source also warns the savings are not immediate. Converting a reactive organization to a proactive best-practice one takes 3-to-5 years, so a plan should phase the $2M rather than book it in year one.

• If a savings claim depends on cutting headcount, then re-test it: the source locates savings in overtime, contractor use, and deferred work, not necessarily people.
• If the organization is mostly reactive, then budget a 3-to-5 year ramp and do not expect full savings immediately.
• If you are estimating contractor-line savings, then set them below the in-house estimate because contractors are already usually more efficient than reactive in-house crews.
• If you cannot measure maintenance cost by labor, materials, and contract lines, then fix the measurement first, since the source says what is not measured cannot be managed.
• If a proposed cost cut reduces availability or efficiency, then weigh it against the downtime cost, which the source says dwarfs the pure maintenance savings.

• Treating the 1/3 waste figure as a guaranteed result. The tell: a budget cut is set at 33% with no baseline analysis of overtime, contractor, or materials spend.
• Equating cost reduction with downsizing. The glossary calls downsizing a commonly misused quick method of cutting expenses. The tell: the plan is a headcount number rather than a change in work mix.
• Counting only labor and ignoring materials. The tell: the savings case is about $1M when the source's logic implies materials could bring it near $2M.
• Expecting the savings in year one. The tell: a reactive shop is judged a failure at month 12 because the full savings have not appeared, though the source says 3-to-5 years.
• Judging maintenance only as an expense line. The tell: spending falls while downtime, which the source says can run 10-to-20 percent or more, rises and costs more than was saved.

• ▸Annual labor budget: enter total maintenance labor spend as the baseline.
• ▸Annual materials budget: enter total maintenance materials spend.
• ▸Annual contractor spend: enter total contracted maintenance spend, tracked separately.
• ▸Overtime / outside contractor / deferred-work split: break labor into these three savings sources.
• ▸Estimated waste % by line: start at the source's 1/3 and replace it with your own record-based estimate.
• ▸Projected annual savings: waste % times baseline for each line; discount the contractor line below in-house.
• ▸Reactive vs. planned work mix: record the current percentage and the target percentage.
• ▸Transition timeline: phase the savings across years 1 to 5 rather than booking them at once.
• ▸Actual savings to date: record expense dollars no longer required, by line, against baseline.

1. 1Fix the definition before measuring. The source defines OEE as a measurement that compares availability, performance efficiency, and quality rate of an asset. Write down one definition for each of the three factors and one rule for combining them. The source does not prescribe the arithmetic. The conventional method is to multiply the three ratios, so pick that or another method, document it, and never change it mid-series.
2. 2Define 'design specification' for the asset. The source's downtime definition is any period when the asset is 'not performing to design specification', whether in breakdown or merely unable to produce at spec. Record the design rate, the design output per time period (equipment capacity), and what counts as good product. Every later number depends on this baseline.
3. 3Separate the loss categories so each lands in exactly one factor. Because the source's downtime includes running below spec, decide up front that stops and breakdowns go to availability, slow running goes to performance efficiency, and rejects go to quality rate. Otherwise the same loss is counted twice.
4. 4Capture idling and minor stoppages explicitly. The glossary names them as a distinct equipment loss: small interruptions, easy to reset, significant when accumulated. Operators must log them, because they rarely show up as formal downtime.
5. 5Set up data collection per shift: scheduled time, uptime, downtime with a cause, output against capacity, and rejects. The source treats data collection as the step that supports a performance indicator. Assign the people who run the equipment to log it, which is operator involvement in practice.
6. 6Calculate availability, performance efficiency, and quality rate separately, then combine them. Always report the three factors next to the composite. A single number hides which factor is failing.
7. 7Run a Pareto on the losses and on the equipment. The source's Pareto logic is that about 20% of assets or causes drive about 80% of problems. Rank the loss causes and the assets by OEE shortfall, and attack the top few first.
8. 8Review OEE jointly across operations, maintenance, and engineering, and act on it. The NASCAR analogy in the source makes the point: capacity from the asset comes from the whole team listening to each other. TPM is the strategy the source ties to maximizing OEE, and it requires full organizational participation.

The source gives no numeric OEE case, so the numbers below are hypothetical and illustrate the method only. A packaging line has a design rate of 100 units/min. In a 480-minute shift it runs 420 minutes (availability 420/480 = 87.5%). In those 420 minutes it should make 42,000 units at design rate but makes 37,800 (performance efficiency 90%), a gap that includes accumulated idling and minor stoppages. Of the 37,800, 36,288 are good (quality rate 96%). Multiplying the factors gives 0.875 × 0.90 × 0.96 = 75.6%. The composite alone says '76%'. The breakdown shows performance efficiency, not breakdowns, is the cheapest gap to attack: the minor stoppages that the source warns are small to fix but large in aggregate. This matches the source's view that measures should point to what to improve tomorrow, unlike financial statements, which it calls 'damage reports'. The NASCAR analogy applies to the follow-up: operators who see the stoppages, maintenance who can fix them, and engineering who can design them out must share the data.

• If the composite OEE moves but you cannot say which factor moved, then stop and report the three factors separately before taking any action.
• If a loss could be counted as both downtime (below spec) and slow running, then assign it to one factor by written rule, because the source's downtime definition covers both and invites double counting.
• If stoppages are short and easy to reset, then log them as idling/minor stoppage anyway, because the source says they accumulate into significant losses.
• If several assets have poor OEE, then Pareto the shortfall and start with the few assets or causes producing most of the loss rather than spreading effort evenly.
• If operations and maintenance disagree on what the OEE data means, then treat it as a team-process issue and resolve it jointly, since the source links asset capacity to organization-wide cooperation.

• Reporting only the composite number. The tell: the team debates the percentage but cannot name the biggest loss.
• Double-counting losses between availability and performance because 'downtime' was defined as any below-spec running. The tell: the factors multiply to a figure lower than what the shop floor believes the asset produces.
• Missing minor stoppages because only breakdowns are logged. The tell: availability looks high while output per hour is chronically below capacity.
• Changing the design spec, definitions, or arithmetic between periods. The tell: OEE jumps without any change in how the equipment runs.
• Treating OEE as a maintenance-only scorecard. The tell: operators do not collect the data, and operations and maintenance blame each other, the dysfunction the NASCAR analogy warns against.

• ▸Asset / line and period: name the one asset and the shift or date range covered.
• ▸Design specification: record the design rate (output per time) and the definition of good product.
• ▸Scheduled time vs. uptime: enter minutes available and minutes actually running at spec, then compute availability.
• ▸Downtime log with cause: list each stop or below-spec period, its duration, and its cause, assigned to one factor only.
• ▸Idling and minor stoppages: tally count and total minutes of small interruptions.
• ▸Output vs. capacity: enter actual output and the output expected at design rate for running time, then compute performance efficiency.
• ▸Rejects: enter total units and good units, then compute quality rate.
• ▸Combination method and composite OEE: state the written rule used (e.g., multiply the three) and the resulting figure.
• ▸Top three losses (Pareto) and owner: rank the biggest loss causes, name an operations, maintenance, or engineering owner for each, and set a review date.

1. 1Define the profit claim as three separate lines: expense reduction (labor, materials, contracting, energy), capacity gain (availability and efficiency), and avoided asset investment. Don't report 'maintenance savings' as one number, because the sources show the three lines differ greatly in size and timing.
2. 2Baseline the expense lines. Record annual maintenance labor, materials, and contractor spend. Split labor into overtime, outside contractors, and deferred maintenance. The source notes that materials spend is typically close to the labor budget, so test that against your own ledger.
3. 3Baseline the share of work done reactively. Breakdown work typically costs two to four times the same work done in planned mode. Each job moved from breakdown to planned is a saving that translates directly to profit dollars.
4. 4Put a dollar value on downtime. Include utility costs, idle production or operations personnel, late-delivery cost, and overtime to make up lost production. Then price the real cost: the lost sales dollars of product not made on time. Never accept 'downtime has no cost.'
5. 5Measure efficiency, not only uptime. Compare running speed and output to design speed and capacity. Some plants discover equipment running at only 50-to-60 percent of capacity. Efficiency losses are always greater than pure downtime losses and are mostly never measured.
6. 6Use maintenance records joined to financial data to find chronic efficiency problems and fix their root causes. This requires accurate records and real CMMS use (many companies reach only about 50% utilization), so audit data entry before you promise savings.
7. 7Add the energy line. Effective maintenance/asset management is credited with a 5-to-11 percent reduction in energy consumption. Document it with specific programs such as steam-leak or air-leak reduction.
8. 8Build a phased profit forecast over 3-to-5 years, not one year. Converting a reactive organization to a proactive one takes that long, mainly because of culture change. Report profit with the corporate indicators (ROFA, return on net assets) that the financial community uses.

The source's own case: maintenance labor is $3M a year, and materials are typically a similar amount. Labor savings come from less overtime, less use of outside contractors, and deferred maintenance done without extra spend. If materials are also cut by one-third, about $1M, the combined savings from better maintenance efficiency and effectiveness can approach $2M a year. Those are expense dollars no longer required, so they translate to profit dollars. The source warns that the savings are not immediate. A reactive shop needs 3-to-5 years to become proactive, mostly because the culture has to treat maintenance as a core business process. On top of this, energy use falls 5-to-11 percent, and the capacity side dwarfs the pure cost savings. Where downtime averages 10-to-20 percent, recovering it adds sellable product without buying redundant equipment. In one cited case, $5.4M in increased production was realized in the 12 months after the change. So the forecast shows roughly $2M of annual cost savings phased in over several years, with the larger capacity gain stated separately.

• If a savings claim has no baseline (spend, downtime hours, design-speed output), then do not book it as profit; build the baseline first.
• If the company is buying backup or redundant equipment to cover downtime, then treat availability recovery as the cheaper profit lever, since the redundancy hurts return on net assets.
• If uptime looks acceptable but output is below design speed, then investigate efficiency losses first; they exceed downtime losses and usually go unmeasured.
• If someone proposes downsizing the maintenance workforce to cut expense quickly, then reject it as a profit strategy; the source calls this a common misuse. Pursue overtime, contractor, materials, and breakdown-cost reductions instead.
• If the proposed payback is under a year for a reactive-to-proactive conversion, then extend the timeline to 3-to-5 years or show which savings are immediate and which depend on culture change.

• Counting maintenance as only an expense line and ignoring capacity gains. The tell: the business case stops at labor and materials, even though the source says the pure maintenance contribution is dwarfed by availability and efficiency gains.
• Assigning downtime a zero or generic cost. The tell: no entry for utilities, idle personnel, late delivery, overtime, or lost sales.
• Tracking only uptime. The tell: 95% 'availability' reported while equipment runs at 50-to-60 percent of design capacity.
• Expecting full savings in year one. The tell: the budget cuts spend before the proactive practices (PM, planning, training) exist, so breakdowns rise.
• Weak data discipline: buying the CMMS but using about half of it, or buying lowest-cost equipment with no maintenance-cost data. The tell: chronic problems persist until a breakdown, and purchasing chooses on price rather than life-cycle cost.
• Reducing inventory spend carelessly or leaving spare-parts waste in place. The tell: excess stock, expedited deliveries, expired shelf life, single-line purchase orders, vanished parts.

• ▸Baseline labor spend ($/yr): split into regular, overtime, outside contractors, deferred maintenance.
• ▸Baseline materials spend ($/yr): and its ratio to labor spend.
• ▸Reactive vs. planned work share: percent of work orders done as breakdown work, with the 2-to-4x cost multiplier applied.
• ▸Energy baseline and target: current consumption and cost, plus the 5-to-11 percent reduction expected, with the specific leak programs behind it.
• ▸Downtime rate and cost per hour: percent downtime, utilities, idle personnel, late delivery, overtime, and lost sales dollars.
• ▸Efficiency gap: actual speed/output versus design speed/capacity, with the source of the measurement.
• ▸Avoided asset investment: backup or redundant equipment not needed, and its effect on ROFA or return on net assets.
• ▸Phased forecast: expected profit dollars in years 1 through 5, noting which depend on culture change.
• ▸Data readiness: CMMS utilization percent and whether the records can support root-cause analysis.

1. 1Write the scope in one sentence using the passages' definition: PM is activity designed to eliminate or reduce wear on asset or equipment systems. Every task you add must name the wear it addresses (inspection, lubrication, adjustment, or proactive component replacement). A task with no named wear mechanism is a habit and does not belong in the program.
2. 2Rank assets before writing tasks. The source notes that a portion of the equipment creates or contributes to 80% of a company's problems. Pull breakdown history from work orders, list the worst offenders, and write PM for them first. Do not spread thin effort evenly across every asset.
3. 3Build each task as a planned and scheduled job. Per the glossary, that means resources are determined in advance, time is estimated, and the work sits in a weekly schedule. Record craft, hours, parts, and tools on the task card so a planner can schedule it without further research.
4. 4Standardize routine lubrication as its own task family. Define the lubricant, point, frequency, and the person responsible, and put it on a regular schedule. Lubrication is the cheapest wear-reduction action in the program, so it must not depend on memory.
5. 5Add operator start-up inspections where operations can see the asset first. Use visual systems (pressure, flow, fluid-level indicators) so a reading is easy to take and a mistaken reading is less likely. This is the operations-involvement link the source stresses with its NASCAR driver/pit-crew analogy: operations reports on how the asset is performing, and maintenance acts on it.
6. 6Run every PM through the work flow: request, plan, schedule, track, report on a work order. A PM that does not produce a closed work order leaves no history, and with no history you cannot rank assets or measure the program.
7. 7Pre-stage parts for scheduled tasks. List the routine spares each task consumes and check the storeroom service level (the percentage of time parts are in stock when required). A stock out on a scheduled PM turns a planned job into a deferred one.
8. 8Review the program against failures. When a breakdown occurs on an asset that has PM, run root cause analysis and change the task, its interval, or the part. Where the source's RCM technique applies, use it to improve the PM program. Add predictive tools (vibration analysis, thermography, ultrasonic, wear particles) only after the basic PM is stable, since PDM is described as an advanced PM technique.

The source gives no task-level PM case, so this example uses its stated cost and downtime figures and only simple arithmetic. The glossary says reactive work arrives with little or no notice, interrupts the weekly schedule, and costs two-to-four times as much as work that is planned and scheduled. Take a pump that fails at its bearing. Reactive repair costs 2 to 4 units for every 1 unit the same job costs when planned. Now put it on a PM program: a routine lubrication task on the bearing plus a start-up inspection of the pump's pressure and flow indicators, both on the weekly schedule with parts pre-staged. If the program converts even one breakdown into a planned bearing replacement, you save between 1 and 3 units of that cost, before counting avoided downtime. The source notes that equipment downtime in some companies averages 10-to-20 percent or more, and downtime carries utility costs and idle-production costs. The saving from avoided downtime therefore comes on top of the direct repair saving. The source also says effective maintenance strategies reduce energy consumption by 5-to-11 percent, with air and steam leak reduction as the cited example. A PM inspection route that finds leaks contributes to that saving.

• If an asset is among those driving most of your breakdown work orders, then give it PM tasks first and review them after every failure.
• If a PM task cannot name the wear mechanism it prevents, then delete it or rewrite it until it can.
• If a breakdown happens on an asset that already has PM, then run root cause analysis and adjust the task, interval, or parts before you add more tasks.
• If PM jobs repeatedly slip because parts are missing, then fix the storeroom service level and routine spares list before adding scheduled hours.
• If basic inspections and lubrication are reliably executed, then add predictive technologies (vibration analysis, thermography, ultrasonic, wear particles) to detect wear before failure; if they are not executed, do not add them yet.

• PM exists on paper but jobs are unplanned: no estimated hours, no parts, no slot on the weekly schedule. The tell is a low completion rate and PM work that is the first thing bumped by reactive jobs.
• Reactive work consumes the schedule. Reactive jobs interrupt the weekly schedule and cost two-to-four times planned work, so the program starves. The tell is that the schedule is rebuilt daily and PM work orders age open.
• PM is spread evenly across all assets instead of focused on the worst offenders. The tell is heavy PM hours on assets with few breakdowns while repeat failures continue on a small group of assets.
• Operations is excluded from inspections and lubrication. The tell is that abnormal conditions are noticed late, readings are mistaken or missing, and maintenance and operations blame each other, the NASCAR driver/pit-crew breakdown the source describes.
• No feedback loop: failures are repaired, not analyzed. The tell is the same failure mode recurring on the same asset with the same PM task unchanged.

• ▸Asset ID and name: identify the asset and note whether it is among your repeat-breakdown assets.
• ▸Wear mechanism addressed: state the specific wear this task eliminates or reduces.
• ▸Task type: choose inspection, lubrication, adjustment, or proactive component replacement.
• ▸Task description and lubricant/parts: list the exact steps, lubricant, and routine spares required.
• ▸Craft, estimated hours, and tools: record the resources determined in advance so the planner can schedule it.
• ▸Frequency and weekly schedule slot: set the interval and the planned week and day.
• ▸Operator start-up inspection and visual indicators: list the readings operations checks (pressure, flow, fluid level) and normal ranges.
• ▸Parts availability check: confirm the part is stocked and note any stock out that delayed the job.
• ▸Work order number and completion findings: record the closed work order and what was found.
• ▸Post-failure review: if the asset failed since the last review, record the root cause and the change made to this task.

1. 1Route every maintenance activity through a work order. The source defines the work order as the document used to request, plan, schedule, track, and report on all maintenance activity. Work done without one cannot be planned, costed, or counted.
2. 2Separate emergencies from everything else at intake. The source defines an emergency as a breakdown needing immediate attention, usually involving safety, health, or process integrity. Only work that meets that test bypasses the planning queue, and everything else enters it.
3. 3Plan each job before it is scheduled. Fill in the labor, materials, and contractor blanks on the work order. The source says planning and work-method adjustments reduce the labor and materials a job needs. Planning also lets stores confirm parts before a craftsperson is assigned.
4. 4Schedule on a weekly cycle. The source's benchmark is 80-90 percent of all maintenance activities planned and scheduled weekly, at which point there is very little waste. Set that range as the target for a reactive organization.
5. 5Execute against the plan and record actuals on the same work order. Capture labor hours, materials, contractor costs, and the duration of the repair. Without actuals, you cannot compare plan to result.
6. 6Record run-time between repairs along with repair duration. The source says MTBF and MTTR depend on accurate records of each repair, its duration, and the run-time between repairs. Do this whether or not you use a computer; a CMMS only helps carry the volume.
7. 7Roll costs up from the work order. The source says cost can be rolled up from equipment to line to department to area to total plant only if it is captured at work order level. That same data also gives you costs by equipment type, craft, and cost center.
8. 8Measure the effect. The source's targets are wrench time moving from about 25 percent toward as much as 60 percent, less overtime and outside contracting, and less reactive waste. Collect before-and-after figures, since the source says quantified improvements build management support.

The source describes a reactive maintenance organization with a $1,000,000 budget. In that mode it may be wasting about $300,000, the 'up to one-third' the source cites. The fix is planning and scheduling: move to the point where 80-90 percent of activities are planned and scheduled weekly. Technician wrench time rises from about 25 percent to as much as 60 percent. By simple arithmetic on those figures, each paid hour then yields about 2.4 times as many productive hours (60/25). The same crew can then absorb work that was going to overtime or contractors, which lowers cost while service improves. Breakdown work is also typically two to four times as costly as the same job done in planned mode, so every job moved from emergency to planned status adds to the saving. None of this can be verified without work order data on labor, materials, and repair duration, so the work order is the measurement point for the result.

• If a request involves safety, health, or process integrity and the asset has broken down, treat it as an emergency and bypass planning. Otherwise, queue it for planning and weekly scheduling.
• If planned-and-scheduled work is below 80-90 percent of activity, then the organization is still functionally reactive. Focus on planning capacity before adding new initiatives.
• If a cost cannot be tied to a specific work order, then it cannot be rolled up to equipment, line, department, or plant. Reject or return incomplete work orders.
• If repair duration or run-time between repairs is missing, then MTBF and MTTR are unreliable. Fix the data capture before trusting the indicators.
• If you propose a change to planning or work practice, then run a before-and-after study. Quantified results are what win management support.

• Bypassing the work order for 'quick' jobs. The tell: labor hours exist that match no work order, and cost rollups come in lower than actual spend.
• Treating most requests as emergencies. The tell: planned and scheduled work stays far below 80-90 percent and overtime and contractor use stay high.
• Planning without scheduling, or scheduling without planning. The tell: wrench time stays near 25 percent because technicians wait for parts, information, or access.
• Incomplete work order closure. The tell: blank labor, material, contractor, or duration fields, so MTBF and MTTR cannot be calculated.
• Buying a CMMS and using only part of it. The source notes many companies reach only about 50 percent CMMS usage. The tell: some work orders are in the system and others on paper or in memory.

• ▸Requester / date / equipment ID: Identify who asked, when, and which asset, so costs roll up from equipment to plant.
• ▸Emergency test: Answer yes or no: breakdown occurred and safety, health, or process integrity is at stake? Yes bypasses planning.
• ▸Planned labor (craft and hours): List the crafts and estimated hours for the job.
• ▸Planned materials: List the parts and materials, and confirm stores availability before scheduling.
• ▸Contractor (if any): Record the vendor and the planned cost.
• ▸Scheduled week and assigned crew: Enter the weekly slot; the target is 80-90 percent of work scheduled this way.
• ▸Actual labor, materials, contractor cost: Fill in at completion; this is the work-order-level cost for rollup.
• ▸Repair duration and run-time since last repair: Record both; they are the inputs for MTTR and MTBF.
• ▸Planned vs. actual variance note: Note why the plan differed, to improve the next plan.

1. 1Fix the two definitions before you measure anything. Service level is the percentage of time a part is in the storeroom when it is required. A stock out is an event where a part is required and is not available. Log every parts request as either filled from stock or a stock out. Service level is then filled requests divided by total requests.
2. 2Tie every request to a work order. The work order is the document that requests, plans, schedules, tracks and reports maintenance, so a part request with no work order cannot be counted as a stock out against a real repair. If the CMMS is only partly used (the glossary notes many sites reach about 50% utilization), your service level is built on half the data. Fix that first.
3. 3Split the inventory into critical spares and everything else. Critical spares are major, high-value, long-lead-time parts carried to prevent excessive downtime in a breakdown. They run about 10% of items but can be up to 50% of inventory value. Tag them in the CMMS and manage them as a separate class.
4. 4Set a separate service-level target for each class. Critical spares are insurance against long downtime, so a stock out there costs the most. Routine parts need high availability, but not at the price of excess stock. The definition requires high service levels without excessive inventory, so track both service level and inventory value on every review.
5. 5Cut stock outs at the cause, not the symptom. For each stock out, record whether the part was never stocked, was stocked below demand, or was ordered late. Match the fix to the cause: add the item, change the quantity, or shorten the purchasing lead time.
6. 6Separate small purchases from stocked parts. Credit cards (VISA, Mastercard, American Express) are used for small plant items, usually for maintenance. Use them for low-value, non-stocked needs so purchase orders and the purchasing department focus on stocked and critical items.
7. 7Evaluate vendors on total cost of ownership, not low bid. Life cycle costing examines cost from design to disposal and usually produces dramatic savings for companies used to buying from low-bid vendors. Apply it to repeat-purchase parts and critical spares first.
8. 8Review the stores results against the cost of the alternative. Reactive maintenance costs two to four times as much as planned work. Price each stock out as a repair that slipped from planned to reactive, and use that figure to decide how much stock is worth carrying.

The source gives a materials-budget case. A company has a $3M annual maintenance labor budget, and studies show the materials budget is a similar amount. The source says one third of maintenance spending is wasted through inefficient and ineffective use of resources. One third of $3M in materials is $1M that is not required, and it flows to profit as expense dollars. Add the equal labor saving and the combined figure approaches $2M per year. To work the stores side, take that $3M materials line and split it into critical spares and routine parts. Critical spares may be about 10% of items yet up to half the value, so up to roughly $1.5M of the $3M could sit in that small group. Reviewing that group first targets the largest dollars with the fewest line items. The source warns the savings are not immediate when improving a reactive organization, so expect them to build over time.

• If a stock out stopped or delayed a repair, then count it against the work order and price it as reactive work at two to four times planned cost.
• If an item is high value with a long lead time, then manage it as a critical spare with its own service-level target and review, not under routine reorder rules.
• If a purchase is small and not stocked, then buy it by credit card instead of routing a full purchase order.
• If two vendors differ on price, then compare life cycle cost (design to disposal) rather than the low bid.
• If service level is high but inventory value keeps growing, then the function is buying its way to availability. Review stock levels against actual demand before raising targets.

• Measuring service level without work-order discipline. The tell: reported service level looks high while technicians say parts are never there, because they fetch parts without a request being logged.
• Treating all parts alike. The tell: a small number of high-value critical spares account for a large share of inventory dollars, yet receive the same reorder logic as bolts.
• Buying on low bid. The tell: repeat failures and replacement of the same part, while the purchasing report shows unit-price savings. Life cycle costing is meant to expose this.
• Chasing service level with unlimited stock. The tell: stock outs fall but inventory value rises with no matching drop in downtime. The definition requires high service without excessive inventory.
• Running stores and purchasing apart from the CMMS. The tell: the CMMS is only partly used (the glossary cites about 50%), so stock levels, requests and purchase orders disagree.

• ▸Review period: state the start and end dates covered.
• ▸Total part requests (tied to work orders): count every request logged in the period.
• ▸Requests filled from stock: count those where the part was in the storeroom when required.
• ▸Service level %: filled requests divided by total requests, times 100.
• ▸Stock-out log: for each stock out, list the part, the work order, and the date.
• ▸Stock-out cause: mark never stocked, stocked below demand, or ordered late.
• ▸Critical-spare status: mark whether the part is high value with a long lead time.
• ▸Repair impact: note whether the repair became reactive (cost two to four times planned) and by how much it was delayed.
• ▸Corrective action and owner: record the stocking, quantity or lead-time change, who owns it, and the due date.

1. 1Run the sequence gate first. The source places PdM and RCM after the basics: an effective PM program (reactive work under 20% of total maintenance labor, where 'reactive' means less than one week of advance planning), then organized stores with a service level above 95%. If either gate fails, fix it before buying any condition-monitoring technology.
2. 2Free up maintenance and engineering time through operator, production and facilities involvement. That means inspecting equipment before start-up, writing work requests, recording breakdown data, basic lubrication and routine adjustments. The freed hours are what you redirect to predictive work.
3. 3Pick the assets by problem, not by what technology is available. Use the 80/20 idea in the source: about 20% of assets create about 80% of the problems. List the chronic failures on that 20%. The source says the aim is to buy technology that solves or mitigates chronic equipment problems, not all the technology on the market.
4. 4Match the technology to the asset type. The source gives two pairings: rotating equipment fits vibration analysis, and electrical equipment fits thermography. Where the asset is already tied to a building automation system, distributed control system or PLC, monitor its parameters in real time.
5. 5Plan and schedule predictive inspections with the same techniques you use for PM tasks. They go on the weekly schedule with estimated time and resources, not into someone's spare moments.
6. 6Record every reading in the CMMS, or interface the monitoring devices to it. RCM can only start once this data exists.
7. 7Apply RCM to the combined PM/PdM program. For each asset, ask whether it is environmentally sensitive, safety related, or extremely critical to operations. If so, decide the appropriate PM/PdM technique and use it. The source text cuts off after this, so state your own rule for the remaining non-critical assets and write it down.
8. 8Review the program periodically for cost-benefit, as the maintenance engineer's role requires. Ask whether each task is best done by operators, craft workers or contractors, and whether it needs doing at all. Keep scanning for new tools and technology, since monitoring them is a named engineering duty.

This is an illustrative walk-through. The thresholds come from the source, but the plant and asset are hypothetical. A plant reviews its labor records and finds more than 20% of maintenance hours go to work started with less than a week of planning. By the source's rule the PM program is not effective, so the plant does not buy vibration sensors yet. It tightens PM and then confirms stores deliver parts at least 95% of the time. Next it lists its chronic problem assets, the roughly 20% that cause most of the trouble. One is a rotating pump that fails repeatedly. Because it is rotating equipment, vibration analysis is the natural fit. The vibration route is added to the weekly schedule like any PM task, and each reading goes into the CMMS. After the data accumulates, RCM is applied. The pump supports a safety-related process, so it keeps its PdM and PM tasks. A low-criticality asset with no failure history gets its tasks reviewed for cost-benefit instead of receiving the same monitoring.

• If reactive work exceeds 20% of total maintenance labor, then fix the PM program before investing in PdM; the source says progress in the later areas is difficult until the 80/20 rule is met.
• If spares are available less than 95% of the time, then fix stores and purchasing first, because a predicted failure you cannot repair is wasted.
• If an asset is environmentally sensitive, safety related or extremely critical, then decide its PM/PdM techniques deliberately through RCM and apply them.
• If a candidate technology does not address a chronic equipment problem you can name, then do not buy it.
• If a predictive reading is not captured in the CMMS, then treat the inspection as incomplete, since RCM and cost-benefit review depend on that data.

• Buying technology before the basics: sensors are installed while reactive work is still above 20%. The tell is that the team is pulled off inspection routes to fight breakdowns.
• Technology-first purchasing: equipment is bought because it exists, not because it solves a chronic problem. The tell is that no one can name the failure the tool targets.
• Predictive inspections run outside the schedule: routes get done when someone has time. The tell is that readings are missing or irregular.
• Data stranded outside the CMMS: readings sit in standalone files or on the analyst's laptop. The tell is that RCM or cost-benefit reviews cannot be run from the records.
• Applying RCM uniformly: every asset gets the same tasks regardless of safety, environmental or operational criticality. The tell is that critical and trivial assets carry identical PM/PdM workloads.

• ▸Reactive labor share: record reactive hours as a % of total maintenance labor (target under 20%; reactive = under one week of planning).
• ▸Stores service level: % of time a requested part is in stock (target above 95%).
• ▸Asset and chronic problem: name the asset in the roughly 20% that causes most problems, and its recurring failure.
• ▸Asset type and technique: rotating equipment suggests vibration analysis, electrical suggests thermography; note any real-time link (BAS, DCS, PLC).
• ▸Criticality screen: mark yes/no for environmentally sensitive, safety related, extremely critical to operations.
• ▸Scheduled frequency: how the inspection is planned and scheduled, using the PM scheduling method.
• ▸CMMS record: where readings are recorded or which interface feeds the CMMS.
• ▸RCM decision: the PM/PdM task chosen given the criticality screen, and who performs it (operator, craft, contractor).
• ▸Cost-benefit review date: when the task and technology are next reviewed for value.

1. 1Fix the definitions before measuring. Use the source's terms: a breakdown is 'an unexpected interruption to the service of a particular asset'. Breakdown maintenance is 'maintenance performed in response to a breakdown'. Planned and scheduled work has 'resources determined in advance and time estimated', built into a weekly schedule. Write these three definitions down so every work order can be classified without argument.
2. 2Add a mandatory work-type field to every work order: breakdown (reactive), PM/PdM inspection or service, or planned and scheduled corrective. A work order is 'planned and scheduled' only if resources and time estimates existed before the work started and it sat on the weekly schedule. The passages give no formula for proactiveness, so build it from labor hours: planned and scheduled hours divided by total maintenance hours, with breakdown hours reported alongside.
3. 3Check that the data can be trusted before reading the ratio. Reconcile CMMS/EAM labor, material and contractor costs against accounting (the source's indicators 1-3). If the percentage is below 100%, work is going unrecorded, and that work is probably the unplanned, informal kind. Trace the gaps to the equipment item. Do not open a blanket or standing work order to force the numbers to balance.
4. 4Check equipment coverage in the CMMS/EAM history (the source's indicator 4) and CMMS usage. The glossary notes that many companies reach only about 50% utilization. A proactiveness ratio computed on half the plant describes only that half.
5. 5Run a Pareto of breakdowns by equipment. The glossary defines it as the 20% of assets creating 80% of the problems. Start proactive effort on those assets, because that is where breakdown hours concentrate.
6. 6Build the proactive work: PM and PdM inspections and services on the Pareto assets, since they 'increase availability by reducing the number of unexpected breakdowns'. Assign a planner to estimate resources and time. Load the estimates into a weekly schedule for the maintenance staff.
7. 7Identify critical spares: major parts with high value and long lead times, carried in stock to prevent excessive downtime. A job cannot be planned if the part is not available. Missing parts convert planned work back into emergencies.
8. 8Bring operations into the loop, following the NASCAR analogy: the driver's report on how the car is running reaches the pit crew. Operations reports equipment performance to maintenance, and maintenance decides the work and timing. Neither side dictates to the other.
9. 9Review monthly: the planned-hours ratio, breakdown hours, and the repeat Pareto. Use the trend to strengthen practices that are working and drop those that are not.

The passages give no plant-level proactiveness data, so this example is illustrative and built on the glossary's one quantified claim: breakdown maintenance 'typically costs two-to-four times as much' as the same work in a planned mode. Take a pump bearing replacement that costs 100 units when planned. Done after a breakdown, it costs roughly 200 to 400 units for the identical job, before counting lost production capacity, which the source calls the 'impact cost' most companies never measure. A planner finds the pump on the breakdown Pareto and adds a PdM oil analysis (analysis of wear particles shows which part is wearing). The bearing job is then estimated, the part is pulled from stock, and the work is placed on next week's schedule. Each such conversion moves hours from the breakdown column to the planned column. It also removes the unexpected service interruption, which is where the cost multiplier sits. The reconciliation checks keep the ratio honest: if CMMS labor is below 100% of accounting labor, some hours are missing and the ratio is overstated or understated, so you trace the gap to the equipment item and correct it.

• If CMMS labor, material or contractor cost is below 100% of the accounting figure, then fix recording before you trust any proactiveness ratio. Trace the unrecorded work to the equipment item.
• If a few assets drive most breakdowns (the Pareto 20/80), then put PM/PdM and planning effort there first, before spreading effort evenly across the plant.
• If a job is on the weekly schedule but resources and time were not determined in advance, then classify it as unplanned. Scheduling alone does not make work proactive.
• If planned jobs repeatedly stall for missing parts, then review critical spares (high value, long lead time) before adding more PM tasks.
• If operations and maintenance disagree on how equipment is performing, then route operator input into the work-request process. Maintenance decides the work and its timing.

• Calling any scheduled-looking job 'planned'. The tell: the ratio looks high, yet no planner estimates exist and breakdowns have not fallen.
• Balancing CMMS-to-accounting differences with blanket or standing work orders. The tell: large generic work orders absorb hours, and labor shows no equipment-level history.
• Measuring only the CMMS-covered equipment. The tell: low equipment coverage or about 50% CMMS usage, and the plant still feels like fire-fighting despite a good ratio.
• Spreading PM evenly instead of targeting the Pareto assets. The tell: PM completion is high but the same few machines still account for most breakdowns.
• Treating maintenance as a cost to minimize, not a capacity lever. The tell: only expense is tracked, with no view of lost production or availability, so reactive costs stay invisible.

• ▸Period and scope: dates and plant areas covered, including the CMMS equipment coverage % for that scope.
• ▸Data check: CMMS labor ÷ accounting labor (%), and the same for materials and contractors. Note any gap to trace.
• ▸Total maintenance hours: all recorded labor hours in the period.
• ▸Breakdown hours: hours on work responding to an unexpected interruption.
• ▸PM/PdM hours: inspection and service hours.
• ▸Planned and scheduled corrective hours: only work with estimates made in advance and on the weekly schedule.
• ▸Proactiveness ratio: (PM/PdM + planned corrective hours) ÷ total hours, with breakdown hours shown beside it.
• ▸Top Pareto assets: the equipment causing most breakdowns, each with a named PM/PdM action and owner.
• ▸Critical spares status: for each Pareto asset, whether high-value, long-lead parts are stocked (yes/no, with action).