Maintenance Planning and Scheduling Handbook, 2nd Edition (McGraw-Hill Handbooks)

1. 1Baseline first: run a wrench-time study on crews available the full shift, and exclude training and vacation time. The source power station found about 35% productivity: 3.5 productive hours of a 10-hour shift, with 6.5 hours nonproductive. Record how much of the 6.5 hours is break time and how much is delay for parts, instructions, or tools.
2. 2Convert the baseline to dollars: multiply the payroll for the crews studied by the nonproductive share. The station paid over $5 million in wages and benefits, so 65% nonproductive equals $3,250,000. This figure justifies the effort and sets the target.
3. 3Plan each work order before it is scheduled. Specify job scope, craft and skill level, number of persons per skill, time estimate, and anticipated parts and tools. A detailed procedure is not always required. This is what removes delays during a job.
4. 4Write the time estimate and headcount on the technician's copy of the work order. Hiding estimates for fear technicians will slow down is not recommended. Technicians are part of the team, and the estimate tells them what the job plan expects.
5. 5Schedule the crew's full capacity for the period. If the crew has 1,000 person-hours available for the week, assign 1,000 hours of estimated work, drawn from the planned backlog. This removes delays between jobs, because no one is left to find work or wait for the next assignment.
6. 6Close the loop at week's end. Compare completed work against assigned hours. For every unstarted or unfinished job, ask what blocked it (a missing item, a wrong scope, a shortage of a normally stocked item) and fix the planning or stocking cause. Also ask whether other jobs were slotted in while waiting.
7. 7Stage parts and tools only when the added effort pays. One person stages several jobs per storeroom trip, ideally the planner who specified the items. Staging also keeps technicians on the job. Treat it as optional: planning and scheduling can achieve high productivity without it, and staging can waste time.
8. 8Coordinate equipment clearance in the scheduling meeting. For example, operations clears the equipment on the night shift from the previous day's request, so technicians do not arrive and wait.
9. 9Track wrench time against stage targets: about 45% with a fundamental planning and scheduling system, about 50% once job files avoid past problems, and over 55% after tool room, inventory, or computerization aids. Do not skip to the last stage before the first is working.

The source power station achieved 93% availability while spending over $9 million on maintenance, including more than $5 million in craft wages and benefits. A study found productivity of about 35%. Mechanic Joe Stark, with a pump job and a valve job in a 10-hour day, physically worked on equipment for only 3.5 hours. The rest went to break time and to stopping for a gasket, a special wrench, or instructions, during which the job did not progress. At 35%, only $1,750,000 of the $5 million bought productive work, and $3,250,000 paid for nonproductive time. A later utility case shows the scheduling link: it had planning without scheduling, and wrench time studies showed time freed from delay areas but no gain in productive time, because no additional work was assigned. After planning and scheduling were extended to 137 personnel, the utility expected to free the equivalent of 78 technicians, who could take on backlog, outsourced work, in-house part building, or preventive maintenance, or absorb attrition without rehiring.

• If a job plan calls for one welder for 4 hours, then do not assign two people for the day. Reassign the freed labor, as a planned 20 person-hours can collapse to a 4-hour task.
• If planning exists but wrench time has not moved, then check whether scheduling assigns the full crew-hours. Planning alone freed time at the utility, but productive time stayed flat.
• If the consequence of an incorrect repair is high, then require a detailed procedure and checklist. If craft skill is high and the consequence is low, then a lighter plan suffices, and the required skills should be defined, not left to chance.
• If a crew leaves jobs unstarted, then trace each to its root cause (missed parts, scope, or a stock-out of a normal item) before blaming effort.
• If you must choose where to put effort, then put scheduling control ahead of staging. Staging is optional, and scheduling control contributes more to productivity than parts and tool delays alone.

• Planning that stops at gathering parts and tools. Tell: the planner reports kits assembled and parts ready, but wrench time and completed hours are flat.
• Planning without scheduling. Tell: individual jobs run smoothly, yet the crew's assigned hours remain well under available hours and productive time does not rise.
• Withholding time estimates from technicians. Tell: work orders carry no hours, supervisors cannot say how much work a crew should finish, and no one can answer whether 40 hours of work were done in 40 hours.
• Accepting delays as normal. Tell: the answer to 'how did the week go?' is 'we worked hard,' with no comparison to assigned work. A crew can be busy without being productive.
• Staging that costs more than it saves. Tell: staged items sit unused, or the staging person makes separate trips per job instead of batching several jobs.

• ▸Crew / period: name the crew and the week covered.
• ▸Available person-hours: the total hours the crew has for the week, excluding training and vacation.
• ▸Planned hours assigned: the sum of the job-plan estimates scheduled; the target equals available hours.
• ▸Hours completed vs. assigned: planned hours finished and the percent of the assignment.
• ▸Jobs not started / unfinished: list each job and the planned hours left over.
• ▸Delay cause per job: parts, tools, instructions, equipment not cleared, travel, or scope error. Tie each to a fix.
• ▸Fix owner and due date: who corrects the planning or stocking cause, and by when.
• ▸Replacement work slotted: whether other jobs were worked in during a blocked job, and how many hours.
• ▸Wrench-time reading: the latest measured productive percent against the stage target (45%, 50%, or 55%+).

1. 1Start each week with a goal of work, not a list of single jobs. Fill the crew's forecast of available work hours with planned, prioritized jobs from the backlog (the 100% strategy). Do not reserve a standing share of hours for expected emergencies.
2. 2Assign work through the work order system so the supervisor holds a tangible handle on the workload. Each crew member becomes responsible for an assigned number of work orders, whether on paper or in a computer listing. Document all work consistently so the supervisor can see and control it.
3. 3Use the planner's time and crew-size estimates to decide how much work to assign, not to grade single jobs. A plan that calls for one welder for 4 hours exposes a two-person, all-day assignment. The estimates show where labor can be reallocated.
4. 4Fill hours by priority, not by craft match. If 100 hours of high priority helper work and 100 hours of low priority machinist work are in the backlog and only 100 machinist hours are available, put the machinists on the helper work.
5. 5Run a daily scheduling process under the weekly goal. After the supervisor assigns a job, the technician owns it. Keep planners out of jobs in progress and do not hold technicians to the time estimate for a single job.
6. 6Break the schedule for true emergencies and urgent work, and tell supervisors up front that they may. Handle emergencies by interrupting jobs in progress, and assure supervisors they never have to wait for planning.
7. 7At week's end, measure schedule compliance and require the supervisor to explain every miss. Sort misses into high priority schedule breakers, low priority breakers, crew low productivity, job problems, technician problems, and supervision problems. Then separate true emergencies from false ones.
8. 8Read the compliance score as a plant indicator. High compliance means the maintenance force is controlling the plant. Low compliance means the plant's reactive workload is controlling the maintenance force.

Palmer describes a utility that had planning without scheduling. Wrench time studies showed planning had freed time from earlier delay areas, but overall productive time did not rise, because the maintenance group never assigned more work. The fix is the one the passages prescribe. Take a forecast of crew hours, say 100 machinist hours, and load it with the planned backlog by priority. If the backlog holds 100 hours of high priority helper-only work and 100 hours of low priority machinist work, the machinists take the helper work. The supervisor hands out work orders against the full week. At week's end, management does not dock pay for a miss. It asks why the schedule broke. A true emergency, a false emergency, or a crew problem each leads to a different fix, and the crew learns whether it won or lost the week.

• If a supervisor is told to hit the weekly schedule and is paid on it, then replace that with accountability for starting each week with a goal of work, running daily scheduling, and explaining misses. Tying pay to compliance creates stress, not control.
• If urgent work arrives, then break the schedule and interrupt jobs in progress. Record it as a schedule breaker, and do not make the supervisor wait for planning.
• If the backlog's highest priority work needs a craft you lack, then assign available people from other crafts to it, since machinists can do helper work.
• If a job runs over its time estimate, then do not hold the technician accountable on that one job. Judge accountability on the whole allocation of work for the period.
• If outage time limits or a credibility problem with operations apply, then an 80% arrangement can be justified. Publicize the planned work and report results regularly. Otherwise default to 100%.

• Planning without scheduling: planning improves job quality but the crew is not given more work. The tell is that wrench time shows delays reduced while overall productive time stays flat.
• Pay-linked schedule pressure: supervisors are held responsible for meeting the schedule instead of for knowing why they missed it. The tell is stress and resistance to the weekly schedule.
• Holding technicians to single-job time estimates: the estimate becomes a stick. The tell is that technicians slow down or distrust the plans.
• Reserving hours for emergencies (the 80% habit): leaving labor unscheduled each week produces less completed work, leaving the preventive work that would head off emergencies undone. The tell is emergencies that then justify the unscheduled time, a self-fulfilling prophecy.
• Treating all schedule breakers as emergencies: low compliance gets blamed on urgent work without checking. The tell is that false emergencies, crew productivity, and supervision problems are never separated out of the miss analysis.

• ▸Crew / week of: identify the crew and the week.
• ▸Forecast available work hours by craft: list hours after leave and known absences.
• ▸Scheduled hours by priority: fill the forecast to 100% from the planned backlog, highest priority first.
• ▸Cross-craft assignments: note where a craft is working a different craft's high priority job.
• ▸Work orders assigned per technician: list the work order numbers each person is responsible for.
• ▸Urgent jobs that broke the schedule: list each with its hours and whether true or false emergency.
• ▸Scheduled work not completed: list each job and the reason (job problem, technician, supervision, crew productivity, other work done, worked less).
• ▸Schedule compliance %: completed scheduled hours divided by scheduled hours.
• ▸Supervisor explanation and next-week action: state why the schedule was missed and one fix.

1. 1Fix the denominator before you measure. Count only available-to-work time under the control of supervisors. Exclude administrative time such as training, meetings, and vacation. Planning does not address that time, and counting it hides what planning does.
2. 2Define the categories in writing before observation begins. Category 1 is productive work on the job site. Every other category is delay: waiting for assignment, clearance, parts, tools, instructions, travel, coordination with other crafts, or equipment information. The source's appendices show sample category definitions under Working, Waiting, Other, and Unaccountable. Use them as the template.
3. 3Measure with a statistically valid work sampling study, not with opinion. Surveys of self-reported wrench time offer choices from 30% to 90%, which leads respondents toward high numbers. Reality is far lower. Use a documented procedure and calculations for the observations, as in the sample studies (Procedure for Measuring Work Force Productivity by Work Sampling, and Work Sampling Calculations).
4. 4Report the full time distribution, not just the headline percentage. A figure showing only 35% on the job is useful only when the other 65% is broken into named delay categories.
5. 5Analyze the nonproductive time. For each delay category, ask whether the delay is avoidable. How much time is spent waiting for parts, for tools, for instruction? Can each delay be avoided? The source calls this one of the most valuable parts of a work sampling study.
6. 6Match each delay to the planning or scheduling action that removes it. Planning before assignment reduces delays during jobs (parts, tools, instructions, clearance). Scheduling reduces delays between jobs (waiting for assignment).
7. 7Repeat the study over time and compare category by category. Ask: has waiting for parts gone down? Waiting for tools? Waiting for instruction? Separate studies show whether planning is getting better or worse.
8. 8Read the result as a verdict on the planning function, not on the technicians. Do not use it to rate individuals. Wrench time is not strictly under the technicians' control (People Rule 12).

The source's illustration (Figure 2.10) shows a company whose time on the job is only 35%. Category 1, productive work at the job site, is the only productive slice. Every other category counts as delay. The source also cites the common perception that a mechanic on a 10-hour shift is productive most of the day, and notes that getting only 3 1/2 hours of productive work from such a mechanic seems hard to believe. Yet 3.5 of 10 hours is exactly 35%. The gap between perception and reality is the opportunity. The expert reads the study as follows. First, a mechanic fetching a part is busy but in a delay that might have been avoided, so it counts against wrench time. Second, the minutes spent getting tools, traveling, and waiting add up to significant delays. Third, the next step is to break the 65% into categories such as parts, tools, instructions, and travel. The planner then targets the largest avoidable category first. A later study shows whether that category's share fell.

• If a delay category (parts, tools, instructions, travel) is large and avoidable, then assign it to planning for prevention. Treat it as a delay to be avoided, not as part of the job.
• If the delay occurs between jobs (waiting for assignment), then fix scheduling. If it occurs during jobs, then fix planning.
• If the question is whether technicians work hard once at the job, then do not answer it with wrench time. Planning aims to put people on their tools in front of a job. It does not measure effort once there, apart from the planner's informal time standard in the estimate.
• If wrench time is low, then analyze the delays before blaming the crew. The measurement reflects the planning function, and wrench time is not strictly under technician control.
• If management's vision is only that planning somehow contributes to effectiveness, then restate the mission as moving technicians out of delay areas and onto job sites. Otherwise the measure has no owner.

• Relying on self-estimates or consultant surveys. The tell: reported wrench time of 50-70% with no statistical study behind it, against a measured result near 35%.
• Treating parts, tools, and travel as part of the job. The tell: nobody can name the avoidable delays, and busy-looking technicians are counted as productive.
• Reporting only the headline percentage. The tell: wrench time is quoted but nobody can say what fraction is waiting for parts, tools, or instruction.
• Including administrative time such as training, meetings, and vacation in the denominator. The tell: results are depressed by time planning cannot affect.
• Using the study to judge individual technicians. The tell: results are discussed by name rather than as a verdict on planning, which damages trust and distorts later observations.

• ▸Study period and crew: dates and the craft group observed.
• ▸Available-to-work time basis: confirm that training, meetings, and vacation are excluded.
• ▸Category 1 (productive work at job site) share: percent of observations.
• ▸Delay share by category: percent for assignment, clearance, parts, tools, instructions, travel, coordination with other crafts, equipment information.
• ▸Avoidable? (Y/N) per delay category: the reason and the evidence.
• ▸Owning function per delay: planning (during jobs) or scheduling (between jobs).
• ▸Corrective action and owner: one specific action per top delay.
• ▸Prior study comparison: change in each category since the last study.
• ▸Verdict: is planning getting better or worse? Note the next study date.

1. 1Measure completion against an assigned baseline, not as a raw job count. The source's owner hears "we turned out 50 jobs, more than ever" and still cannot say whether that is good. A count has no denominator. Your primary metric is: estimated hours of work orders completed ÷ estimated hours of work orders assigned for the period. Keep the job count as a secondary figure.
2. 2Plan every work order before it can be scheduled. The planner specifies job scope, craft and skill level, time estimate, and anticipated parts and tools. Without an hours estimate and a skill level, you can neither size the assignment nor score the result.
3. 3Size the weekly assignment from available hours. In the source, ten technicians at 40 hours each give 400 hours, so the owner selects 400 hours of estimated work from the backlog. Schedule for every work hour available and hand the list over before the week starts (the source does it on Friday).
4. 4Plan each job for the lowest skill level that can do it, and recompute the hours. The source plans the same five jobs at 64 labor hours with two mechanics or two welders on several jobs, and at 44 labor hours with a welder or mechanic plus a helper. If you overstate skill, you overstate the hours and misstate both the allocation and the completion rate.
5. 5At the end of the period, pull back every job not started and sum its estimated hours. The owner's Friday check shows about 100 hours of jobs not started. Then ask for each job why it did not happen, and record the reason on the work order, not just in conversation.
6. 6Close the loop on each job. The lead technician or supervisor gives the planner feedback after every job: problems, plan changes, anything that improves future plans. The planner files it by equipment number. Over time, this is how your estimates and next week's completion rate improve.
7. 7Protect the pipeline that feeds completion. Keep at least one week of planned, approved, ready-to-execute backlog. Keep planners out of crew supervisors' control so they are not pulled into field work. Have crew supervisors handle the current day's work and problems.
8. 8Make parts shortages visible. Planners and technicians reserve parts even when they plan around a missing item. Otherwise the storeroom never records a stockout, and the 'no parts' cause of incomplete jobs stays hidden.

The source's one-person-company story shows the method. An owner who had grown to ten salaried technicians asked how the week went. The answer was 'we did a lot, 50 jobs, more than ever.' The owner had no way to judge that. So he estimated every job in the backlog and handed the crew 400 hours of work: 10 people × 40 hours. The next Friday he asked how much of that list was done. The reply was 'most of it.' He took back the unstarted jobs and found about 100 hours' worth. That means roughly 300 of the 400 assigned hours (about 75%) were started, and the completion rate can be no higher than that. The question then shifts from 'how many jobs' to 'why did these 100 hours not happen.' The technician's first answer begins with jobs for which something was missing, and the owner can now chase each cause. The count of 50 never could have led him there.

• If you can only report a job count, then do not call it completion. Add an estimated-hours baseline first, because 50 jobs means nothing without knowing how many hours should have been done.
• If the assigned hours fall well below the available hours (for example, under 400 for ten people at 40 hours), then fix the allocation before judging the crew. A low completion rate is not the issue; under-assignment is.
• If the same job type shows large hour swings between skill-level plans (such as 64 versus 44 labor hours for the same five jobs), then re-plan at the lowest required skill level before you schedule or score it.
• If unstarted jobs trace to missing parts, then check whether stockouts were ever reserved and recorded. If the storeroom shows few stockouts but jobs keep slipping, then the reservations are not being made.
• If planners are being lent to crews for field work, then refuse. Extend the job's schedule instead, as Smith does in the source, so the planned backlog stays at one week or more.

• Celebrating volume. You report 'we did a lot' or a record job count. The tell: nobody can state how many hours the crew should have completed.
• Estimating without skill level. Hours are listed but craft level is blank, or everything is planned at the top skill level. The tell: the supervisor has to 'second-guess the planner' and shift apprentices onto mechanic work, and welder work always exceeds the welders available.
• Scoring against a thin backlog. Planners are pulled into field work, so fewer than a week of ready work exists. The tell: crews work mostly unplanned jobs and completion cannot be compared to any plan.
• Hidden stockouts. Planners change the plan, for example replacing the whole valve because bonnets are out, and never request the part. The tell: the stockout rate looks low while incomplete jobs mention parts, and the storeroom never gets the signal to stock more.
• No feedback loop. Incomplete or changed jobs are discussed but not filed against the equipment. The tell: the same job slips for the same reason a few months later, and estimates do not improve.

• ▸Period and crew: write the week and crew name.
• ▸Available hours: headcount × hours per person (for example, 10 × 40 = 400).
• ▸Assigned estimated hours: the sum of planner estimates for the jobs handed to the crew. Planned at the lowest required skill level? Yes/No.
• ▸Jobs assigned (count): the number of work orders on the list.
• ▸Jobs completed (count) and completed estimated hours: sum the planner estimates for finished jobs only.
• ▸Unstarted jobs and their estimated hours: list the work order numbers and total the hours.
• ▸Completion rate: completed estimated hours ÷ assigned estimated hours. Also record assigned ÷ available hours.
• ▸Reason for each incomplete job: write one reason per work order (for example parts, skill shortage, access, or other). Note whether the part was reserved.
• ▸Feedback sent to planner: Yes/No, date, and the equipment file number it was filed under.

1. 1Define capacity in the unit your business sells. The source's example is equivalent availability, a utility measure of generating capacity: 93% in 1993 at a large electric power station, well above industry average. Pick your equivalent and record the baseline before any change.
2. 2Link capacity to revenue in your own business model. The source flags companies that show reliable capacity in the revenue section and maintenance in the fixed-cost section. State in writing that reliable capacity is an investment in maintenance, because capacity is not reliable by itself.
3. 3Map how the maintenance system actually works before you change the budget. Cutting or adding maintenance budget without understanding the details of the system is the error the source describes. Record annual maintenance spend and the labor share (the station: over $9 million, of which more than $5 million was wages and benefits for mechanical, electrical, and I&C crafts).
4. 4Measure productive time by work sampling. At the station, productivity was about 35%, so a person on a 10-hour shift made job progress for only 3.5 hours. Sample only people available for the full shift, and list the delays: waiting for parts, instructions, and tools.
5. 5Remove the delays with planning. Planning identifies parts and tools and reserves or stages them, builds minifile support for field technicians, and gives management the information to schedule enough work. Planners do not put first priority on extensive research into possible parts and tool needs.
6. 6Convert the freed labor into capacity work. Redirect the extra labor power to added-value activities. Calculate the gain: the station's mechanical craft of about 30 persons delivered the effort of 47 in 1994, with no new hires.
7. 7Check the result against the outcome metrics: availability held or raised, breakdowns reduced, and lower maintenance and production costs. Decide upstream spending with the 1-10-100 Rule: $1 spent in engineering to reduce maintenance eliminates $10 of later proper maintenance or $100 of breakdown maintenance.

A large electric power station achieved 93% equivalent availability in 1993, well above industry average. It could not have sold electricity without that capacity. In the same 12 months it spent over $9 million on maintenance, with more than $5 million in craft wages and benefits. A study found productivity of about 35%: on a 10-hour shift, mechanic Joe Stark, with a pump job and a valve job, physically worked on the equipment for only 3.5 hours. The other 6.5 hours went to necessary breaks and to delays for gaskets, special wrenches, and instructions. Management redirected its existing planning group and implemented planning for the mechanical craft of about 30 persons. Less than a year later, in 1994, those 30 yielded the effort of 47. No one was hired, so the 17 extra persons' worth of work cost nothing. Sustained availability like this lets a company defer new construction even as sales grow, which lowers capital cost.

• If maintenance sits in the fixed-cost section of your business model and capacity in the revenue section, then restate maintenance as the investment that makes the capacity reliable before you debate the budget.
• If management proposes a budget change (cut, or no increase after capital additions) and nobody can describe how maintenance work flows, then study the system first. Budget pressure alone will not make crews 'work harder'.
• If measured productivity is low (the station's was about 35%), then look for delays in parts, tools, and instructions before adding headcount.
• If you are deciding on upstream engineering spend, then apply the 1-10-100 Rule: $1 up front avoids $10 of proper maintenance or $100 of breakdown maintenance later.
• If planning has run for years with no visible improvement, then check whether it follows the actual principles and practices. Knowing what planning is does not make it happen.

• Treating capacity as free: the business model carries it as revenue and maintenance as a fixed cost. The tell is maintenance budgets cut while sales targets rise.
• Changing the maintenance budget without understanding the system. The tell is hoping crews will 'do what it takes' while jobs still stall for parts and tools.
• Equating busy with productive. The tell is full shifts with only about 3.5 hours of actual job progress out of 10, and delays treated as normal.
• Adopting the common perception of planning (gather parts, write instructions) without the specific results, principles, and practices. The tell is more than 10 years of planning with no significant improvement in maintenance.
• Skipping a baseline, so the gain cannot be shown. The tell is that you cannot state availability, productivity, or the labor-equivalent gain (30 persons delivering the work of 47).

• ▸Capacity measure: name the unit sold or generated (e.g., equivalent availability) and its current value.
• ▸Industry or internal benchmark: record the comparison figure.
• ▸Annual maintenance spend: total cost, with the wages-and-benefits portion shown separately.
• ▸Craft headcount in scope: number of persons in the group being studied.
• ▸Measured productive time: share of the shift with actual job progress, from work sampling of full-shift persons.
• ▸Top delay causes: list the delays for parts, tools, instructions, and breaks, ranked by time lost.
• ▸Planning actions taken: parts and tools identified, reserved, or staged; minifiles built.
• ▸Effort-equivalent after planning: persons' worth of work delivered by the same headcount (e.g., 30 delivering 47).
• ▸Outcome check: availability, breakdown count, and maintenance and production cost versus baseline.

1. 1Make the rule absolute: no work starts without a work order. The source names the failure directly: supervisors 'accept work requested without work orders' and 'back door' jobs bypass the schedule. Tell requesters and supervisors that undocumented work is not planned, scheduled or counted.
2. 2Standardize the work order form and its coding before anything else (the handbook devotes separate sections to Work Order Form and Coding Work Orders). Fix one set of fields and codes, including work type, so every order can later be sorted and measured.
3. 3Give the system one intake and one authorization point. Every request enters as a work order, and someone with authority approves it before it reaches the planner. Record requester, equipment and symptom at this stage.
4. 4Have the planner develop each approved order to the planned level of detail: scope the job, troubleshoot if needed, set craft skill level, estimate work hours and duration, identify parts, special tools and job safety items, and estimate job cost. Flag work to contract out. Record all of it on the order or its attachments.
5. 5Use the order as the sorting and allocation unit for scheduling. Total the backlog work orders and hours, forecast available work hours, sort orders, and allocate them to the weekly schedule. The weekly schedule is a list of work orders, never a list of verbal requests.
6. 6At the daily level, the supervisor assigns names and hands out the work orders. The crew leader and technicians own the job once it is assigned. Give the technicians the order with its parts, tools and attachments already staged.
7. 7Capture feedback at closure through an easy input process. Technicians record actual hours, what was found, parts and part numbers actually used, and any follow-up work. Each piece of follow-up work gets its own work order. Do not let it ride along undocumented.
8. 8Close and file the order into the component-level file (minifile) or equipment history. Specific part numbers and lessons from past jobs are what technicians need to avoid repeating delays; generic procedures do not carry them.
9. 9Read the system through its indicators: backlog work orders, work orders completed, backlog work hours, and schedule compliance. If these cannot be pulled from your orders, the form or coding is inadequate.

The source describes one plant's CMMS rollout that shows how a work order system gets adopted. The plant first installed the CMMS only to track paper work orders and to supply equipment data to planners. Management then gave everyone the option to write work orders directly into the system. The implementer expected only 10% to 20% of new work orders to be entered directly. The surprise was that over 95% were (the passage cuts off here, so the exact wording after '95%' is not available). The lesson is that an easy-to-use intake makes people route all requests through the system. The same source warns that the cost is paid if work bypasses the order: technicians who find follow-up work and pursue it without documenting it break the schedule, and management cannot see that productivity suffered. The fix is to open a new work order for the follow-up, plan it, and then close the original with feedback filed to the minifile.

• If a request arrives without a work order, then have it written up before work begins, unless it is a true emergency, and then document it as soon as the job is done.
• If a technician finds follow-up work during a job, then if it fits within the current assignment without altering the schedule, do it and record it. If not, write a new work order and let it enter the backlog.
• If an order lacks a scope, hours estimate, parts or tools, then it does not go onto the weekly schedule. Return it to the planner.
• If feedback entry is hard, then expect incomplete closeouts. Simplify the input before demanding more detail.
• If you are considering a CMMS, then first make sure the paper process is sound. The source lists 'faulty processes' as a caution, and a computer will not repair a flawed order flow.

• Back-door work: supervisors accept jobs by word of mouth. The tell is crews busy on tasks with no order number and a schedule that is routinely broken.
• Undocumented follow-up work: technicians extend jobs without recording them. The tell is low schedule compliance with no explanation and repeat failures missing from the history.
• Orders closed without feedback: no actual hours, parts or findings are filed. The tell is that the minifile does not grow and the same part-number delays recur.
• Inconsistent coding: work type and equipment are entered differently by each person. The tell is that backlog and reactive-work reports cannot be trusted or sorted.
• Over-reliance on generic CMMS templates: the system suggests generic procedures instead of equipment-specific history. The tell is job plans lacking part numbers and previous-job lessons.

• ▸Work order number / date requested / requester: unique ID and who asked.
• ▸Equipment ID and symptom: the specific equipment and the problem as observed.
• ▸Work type code and priority: use your standard coding so the order can be sorted and measured.
• ▸Authorization: name and date of the person approving the work.
• ▸Scope and job plan: what is to be done, any troubleshooting done, and attachments (sketches, procedures).
• ▸Craft skill level, estimated work hours and duration: who is needed and for how long.
• ▸Parts, special tools, safety items: part numbers, reservations or staging status, and required permits or hazards.
• ▸Estimated cost / contract-out flag: planned cost and whether outside labor is needed.
• ▸Scheduled week and assigned names: weekly schedule slot and the supervisor's daily assignment.
• ▸Closeout feedback: actual hours, parts used, findings, follow-up work order number, and filing location (minifile/history).

1. 1Name the sponsor and the reason first. The source says planning sits 'down in the maintenance group' and so 'does not command the plant manager's attention.' Put a plant-level sponsor on paper before any planner is hired. Otherwise the three standing objections win by default: 'beneath the plant manager,' 'unnecessary to the maintenance manager,' and 'uninteresting to the plant engineer.'
2. 2Brief management on what planning costs them. The techniques need more organization, coordination, and accountability, and they take some control away from supervisors. Say this to the maintenance manager and supervisors up front, so the loss of control is agreed in advance instead of resisted later.
3. 3Establish the quality focus before planning and scheduling start. Tell technicians in plain terms that they are not expected to hit a limited time estimate at the expense of good work. Planning pushes hard on productivity, so the workforce must be assured that quality still comes first.
4. 4Introduce planning properly, especially on parts. Management should not advertise that planning will identify all future parts. The source says planning cannot gather all parts information before all jobs, and that overpromising gives planning a poor reputation that hinders later technician cooperation. State the real role instead: planning saves and retrieves the parts information technicians have already gathered, and ensures availability before the job begins.
5. 5Fund the foundation in the order the source gives it. A work order system comes first, then planning as the biggest next improvement. Give the planning group a defined organization and interfaces, a qualified and trained planner, a workspace, and effective files. Provide equipment tags, timely purchasing of nonstock parts, and a helpful CMMS as the aids management has to supply.
6. 6Set the first target as 'plan and schedule enough jobs for one week.' The source lists this as the planning and scheduling milestone for a new group. Then require execution of the scheduled jobs and feedback, and keep planning and scheduling ongoing.
7. 7Hold the discipline through indicators, not through single jobs. Review planned coverage, proactive versus reactive work, schedule compliance, wrench time, and backlog on a regular cadence. The People Rules say management cannot hold technicians accountable to time estimates for single jobs, and that everyone should go home each day knowing whether they have won or lost. Use group-level measures for the first and daily schedule compliance for the second.
8. 8Be visible where the work is staged and fed back. A supervisor who receives staged items with the crew present sees directly how staging is working. Management follows the same logic: show up where planning is delivering, and ask the questions the book lists for managers (Appendix P).

The author describes a plant where management revamped its existing planning organization. The result was a total clearance of a large backlog, with some work orders as old as 2 years, in less than 3 months. Clearing it freed in-house labor, so a scheduled major overhaul could start without costly contractor assistance. The author also says the plant had to unlearn as many false notions about planning as it had to learn principles. That points to management support as active correction of wrong ideas, not just approval. Use this as your template. Define planning correctly, rebuild the group, point the freed labor at a visible target such as the backlog or a planned overhaul, and report that result back to the plant. The source gives no other numbers for the effort, so measure your own baseline before you start.

• If the plant manager does not see planning as their concern, then assign a named plant-level sponsor and tie planning to a result they care about, such as backlog or overhaul labor, before hiring planners.
• If supervisors resist the loss of control, then name the shift openly (more organization, coordination, and accountability) and give them a defined role in following the P&S system, rather than pretending nothing changes.
• If a manager wants to hold a technician to the estimate on one job, then stop it. Judge compliance through the schedule and group indicators, because planners cannot make the perfect time estimate and single-job accountability is off the table.
• If planning is being sold to technicians as 'we will find all your parts,' then correct the message now. Restate it as capturing and retrieving parts information plus ensuring availability, and keep any staging promises limited to what the group can deliver.
• If the group is being pulled into jobs already in progress or urgent breakdowns, then protect planner time. The People Rules say planners do not add value helping jobs-in-progress, and the aids and barriers list treats reactive environments and technician interruptions as problems for management to solve.

• Sponsorship in name only. The plan is announced, but no resources, planners, tags, or files follow. The tell is a planner who starts with no file system, no work order foundation, and no way to buy nonstock parts.
• Overselling parts help. Planning is promised as complete parts identification, and technicians judge it by the gaps. The tell is technicians citing missed parts as the reason to ignore plans.
• Time estimates used as a stick. Estimates are enforced job by job and technicians feel they must trade quality for speed. The tell is rushed work, padded estimates, or technicians who stop giving job feedback.
• Planning treated as a clerical side function. Management does not look at the group again after setup. The tell is no indicator reviews, planners absorbed into breakdown work, and no feedback loop from execution.
• No quality focus before launch. Productivity pressure arrives first and the workforce reads planning as speed-up. The tell is technician distrust and weak cooperation with the planning group.

• ▸Executive sponsor (name/role): Who at plant level owns planning and has agreed to it in writing?
• ▸Why it matters to them: Which result (backlog, overhaul labor, other) is the sponsor accountable for?
• ▸Control-loss briefing done: Date and audience told that planning means more organization, coordination, and accountability and less supervisor control.
• ▸Quality-first message: Date and wording of the statement that time estimates do not override good work.
• ▸Planning scope statement: Written description of what planning will do with parts, including that it saves and retrieves prior parts information and does not promise all future parts.
• ▸Foundation resources committed: Work order system, planner position, workspace, files, equipment tags, purchasing path for nonstock parts, CMMS. Mark each Yes/No/Owner.
• ▸First-week target: Number of jobs to be planned and scheduled for one week, and the start date.
• ▸Indicators and review cadence: Which of planned coverage, proactive vs reactive, schedule compliance, wrench time, and backlog you review, who reviews them, and how often.
• ▸Visible presence plan: Where and when management will observe staging, crew start-up, and feedback in person.

1. 1Split the two roles before you staff either. Craft technicians execute work that has started. Planners prepare work that has not started. Write this into the job descriptions so nobody can read a planner as spare field labor.
2. 2Select planners from your top technicians. The passages say planners usually come directly from the crafts or craft supervision, and sometimes from engineering or construction technology. Pick the people whose craft skill you would most want on a hard job, because the plan is only as good as the planner's field judgment.
3. 3Move planners out of the crew supervisors' chain of command. Put them in their own crew with their own supervisor. With a small group, have them report to the manager above the crew supervisors, and name a lead planner for direction and consistency. A supervisor with direct access to a planner who is a qualified welder will always pull that planner onto the job in progress, because to the supervisor the present is always more urgent than the future.
4. 4Pay planners at supervisor level. Engineers earn more than crew supervisors because of market demand, and the market can also pull your best technicians away if they have no room to grow. Supervisor-level pay keeps technical talent in the company. Do not leave an easy way to send planners back to the crew if planning stalls. That signals half-hearted support.
5. 5Take time estimates from the skilled technician-planner's judgment, not from file averages or built-up standards. History gives only general guidelines for a current job. Before using a past time, ask why the jobs differed: the inexperienced technician, no competing work, no schedule pressure.
6. 6Feed actual outcomes back to the planner. The oil burner gun history could not be interpreted because feedback was minimal. Without feedback the planner cannot tell a bad estimate from an unusual job.
7. 7Before approving new craft headcount, test whether converting existing craftspeople to planners solves the problem. One manager with 38 craftpersons planned to hire 13 more. The source's advice was to hire no one and make two existing craftpersons planners, avoiding over $500,000 a year in cost for the productivity of about 20 extra people.
8. 8Select and train planners deliberately. The source treats selection and training as the basis of control (Chapter 11) and points to Appendix M for selection guidance. Treat both as a defined program, not an informal promotion.

The source gives three actual histories for cleaning an oil burner gun: one person for 20 hours; two persons at 4 hours each; two persons at 6 hours each. A planner working from the file might average these and plan two persons at about 7 hours each. A skilled technician-planner asks why the jobs differed. The 20-hour job may have gone to an inexperienced technician. Or the person may have had no other work or schedule pressure and simply stretched the job across two 10-hour days. If so, the two shorter jobs are the realistic base, and the planner plans two persons at about 5 hours each. The file shows only the hours. The planner's craft experience explains them, and that is why the source prefers the technician-planner's opinion over built-up estimates.

• If a supervisor asks to borrow the planner for field work, then refuse unless it is truly necessary. The structural fix is for planners not to report to crew supervisors at all, so the request has no authority behind it.
• If your history for a job shows widely varying times, then find out why before averaging. Plan from the jobs that reflect normal conditions, not the mean of all of them.
• If you are about to add craft headcount, then first ask whether converting two or so of your existing craftspeople to planners delivers the same output at lower cost.
• If you are choosing between candidates for planner, then take the stronger craft technician, not the one the crew can most easily spare.
• If a company wants planners to report as hourly workers or keep a way to send them back to the floor, then recognize that as under-supporting planning, and fix the pay and reporting line.

• The planner doubles as a crew resource. Tell: planners show up on toolboxes, and the reason given is that they are qualified welders and the repair is urgent.
• Planners reporting to crew supervisors. Tell: the planner's day is set by whatever job is behind today, and the backlog of unprepared work grows.
• Estimates by file average. Tell: planned hours match history averages while the actual hours scatter wildly and nobody can explain why.
• Staffing planners from whoever is available or weakest. Tell: technicians ignore plans because the planner lacks credibility, and the planner's estimates miss.
• Treating planning as a hedge. Tell: planners are paid below supervisors, can be reverted to the crew, and there is no selection or training program, which leaves you unable to replace those who do poorly.

• ▸Candidate name and current craft/role: record the person's trade and whether they come from the crafts, craft supervision, or engineering.
• ▸Craft skill evidence: list the jobs that show top-level technician ability, since planners must be able to judge field work.
• ▸Reporting line: state who the planner reports to, and confirm it is not a crew supervisor.
• ▸Pay level versus crew supervisor: record both and confirm the planner is paid at supervisor level.
• ▸Return-to-crew provision: state whether the company keeps an easy way to send the planner back, and note it as a support risk if so.
• ▸Training plan: list the training the planner receives and who delivers it, drawing on the selection and training guidance.
• ▸Estimate review practice: note how the planner uses history, including the check on why past job times differed.
• ▸Actual-hours feedback loop: state how completed-job feedback returns to the planner and who is responsible.
• ▸Field-borrowing log: record each time the planner is pulled to field work, who asked, and why.

1. 1Set the planners up as a separate group that reports outside the crew supervisors' daily firefighting. The handbook lists Principle 1 as a separate department. The station in the source redirected its existing planning group rather than hiring, so start by redirecting people. Planning pays off in any organization with more than ten maintenance persons.
2. 2Point the group at future work only. Planners do not help jobs already in progress, because the handbook says that adds no value. Give the current day's work to the crew leader and supervisor. Planners prepare jobs that will be scheduled later.
3. 3Build a component-level file (minifile) for each piece of equipment that receives work. Do not start with all the plant's equipment. The source's example is 300 of 10,000 pieces needing continuous attention, so file the equipment that gets work first. The 50% rule says a piece of equipment that needed work has a 50% chance of needing similar work within a year. The 80% rule gives an 80% chance within 5 years.
4. 4Scope each job before you plan it. Decide what needs to be done, whether troubleshooting or engineering help is needed first, what craft and skill level applies, and what the job's intent is. The plan states the intent of the work. It does not dictate technique, because the supervisor, not the planner, promotes good work practices.
5. 5Estimate work hours and job duration from your own expertise (Principle 4). Treat each estimate as a planning input, not a performance contract. The people rules say planners cannot make the perfect estimate and management cannot hold technicians accountable to the estimate for a single job.
6. 6Write the job plan with scope, craft, and time. Add a 'why' line whenever file history drove the strategy, for example 'This valve is being replaced since patching it in the past has not worked well.' Add any legal or regulatory requirement that technicians do not commonly know.
7. 7Do not research a full parts and tool list on the first job when identification is not readily available. Let the technicians identify what was needed and report it through feedback. Accumulate that feedback into the file, and the lists build up over time.
8. 8Close the loop on every completed job. Read the returned work order, file it in the component file, and scrutinize it for repeat problems. Jobs soon after other work often fail because of 'infant mortality' (faulty materials or practices from the first job), so feedback is where the next plan improves.

The source's power station had achieved reliable plant capacity and a year of superior availability. Studies still showed room to improve workforce productivity. Management redirected its existing planning group and implemented planning according to the handbook's principles for about 30 mechanical craft persons. Planning had existed since 1982, but planning by these principles began in 1994. Less than a year later, in 1994, 30 maintenance persons were yielding the effort of 47. No one was hired, so it was as if 17 extra people had started helping at no cost. The practices behind that result were a separate planning group, future-work focus, component files, planner estimates, and plans that respect craft skill. A smaller failure in the source shows the plan-writing rule. A planner reviewed the history file and called for a valve replacement, but left the history reason off the job plan. The technician repaired the valve instead and wrote back 'I saved the company money by repairing the valve instead of replacing it.' One sentence of 'why' on the plan would have prevented that field decision.

• If a piece of equipment has had work once, then build or update its component file now, because it has a 50% chance of needing similar work within a year and an 80% chance within 5 years.
• If the file history drove the job strategy, then state the 'why' on the plan so the technician does not make an unwise field decision.
• If parts or tool identification is not available on the first job, then do not stall on research. Let the technicians report what they used and build the list from feedback.
• If a job is already in progress, then leave it to the crew leader and supervisor. Planners add value on future work.
• If someone wants to use a time estimate to judge a technician on a single job, then refuse. Estimates are planner expertise, not accountability standards for one job.

• The planner helps with jobs in progress, which pulls the planner into daily firefighting. The tell is a planner who is always at the job site and has a backlog of unplanned work orders.
• The plan prescribes every technique and treats the crafts as unskilled. The tell is technicians resenting the plans and doing the job differently, and the supervisor's role in promoting good practice being taken over by the planner.
• The plan omits why a strategy was chosen. The tell is the technician overriding the decision, as in the valve repaired instead of replaced.
• Management treats estimates as per-job accountability measures. The tell is padded or defensive estimates and conflict over single jobs, when the source says planners cannot make the perfect estimate.
• Feedback from completed jobs is not filed or read. The tell is the same delays and repeat failures recurring, and component files that never gain parts or tool lists.

• ▸Component / equipment ID: name the specific piece of equipment whose component file this plan belongs in.
• ▸Job scope (intent): state what needs to be done, not the technique.
• ▸Craft and skill level: list the craft and the skill level the job requires.
• ▸Estimated work hours and job duration: record your own expert estimate for each.
• ▸Why this strategy: give the file history reason behind the chosen approach, e.g. past repairs failed.
• ▸Legal or regulatory requirements: list any procedure requirements technicians may not commonly know.
• ▸Safety notes: record job safety items such as confined space or MSDS needs.
• ▸Parts and special tools known: list only what is already identified, and mark the rest 'to be reported by technician'.
• ▸Feedback after execution: record what the technician reports about actual parts, tools, hours, and problems, then file it in the component file.

1. 1Start from the allocation principle: advance scheduling is an allocation of planned work to forecast labor hours, not a list of the jobs that feel important. The handbook's Chapter 3 opens with 'Why Maintenance Does Not Assign Enough Work', so assume the default failure is under-assigning and design the process against it.
2. 2Forecast the hours for each crew for the coming week. Chapter 6 lists 'Forecasting work hours' as the first weekly step. Build the forecast from the highest skills available (Principle 3), so the number reflects who will actually be there. Record it per crew, because the 100% target applies per crew.
3. 3Sort the planned work orders by job priority (Principle 2: schedules and job priorities are important). Only planned work, meaning work with a job plan and estimated hours, enters the sort. This follows the 'Sorting work orders' step before 'Allocating work orders'.
4. 4Allocate sorted work orders to each crew until the estimated hours equal 100% of that crew's forecast hours (Principle 4: schedule for every work hour available). Match jobs to skill. Plan each job for the lowest skill level that can do it (Principle 1), then allocate it against the skills forecast.
5. 5Hold the formal weekly schedule meeting (Chapter 6). Use it to confirm the allocation with the people who execute it and to settle conflicts before the week starts. The passages give the meeting's name and place in the sequence, not its agenda.
6. 6Stage parts and tools ahead of the work. Chapter 6 breaks staging into what, where, who and the process, so decide each of the four before the week begins. This removes the delays that keep productive time low.
7. 7Leave current-day decisions to the crew leader (Principle 5). The advance schedule fixes the week's work. The supervisor assigns names, coordinates with operations and hands out work orders each day (Chapter 7). Do not rewrite the weekly allocation daily.
8. 8Measure with schedule compliance (Principle 6) and track the schedule forecast indicator in Chapter 11. Remember Appendix B Rule 13: compliance is not strictly under the supervisors' control. Read misses as a signal about the process, not as a verdict on the crew.

The handbook's power station supplies the real benchmark. After 1993 it redirected its planning group and, following the handbook's guidelines, implemented planning for its mechanical craft of about 30 persons. Less than a year later, in 1994, those 30 people yielded the effort of 47, with no new hires. Scheduling is where that gain gets consumed, because freed-up hours produce nothing unless work is assigned to them. The handbook's productivity figures show how much room there is. Typical wrench time is 25% to 35%. On a 10-hour shift at 35%, a technician makes productive progress for only 3.5 hours, and the other 6.5 go to breaks and delays such as getting parts, instructions or tools. A fundamental planning and scheduling system moves wrench time to about 45%. File history lifts it to about 50%, and special aids to over 55%. In practice, forecast each crew's hours for the week, fill 100% of them with prioritized planned jobs, stage the parts and tools, and let the crew leader assign names each day. A crew scheduled at less than 100% leaves the freed hours unused.

• If a crew's allocated estimated hours are below 100% of its forecast hours, then keep pulling planned work from the sorted list by priority until they match. Do not stop at the jobs that feel urgent.
• If a job could be done by a lower skill level than the person you want to assign, then plan and allocate it at the lowest required skill. Reserve the highest skills for the work that needs them.
• If the forecast skills and hours change (absence, reassignment), then re-forecast from the skills actually available before allocating. Do not schedule from the roster.
• If a job is not yet planned (no plan, no parts identified), then keep it out of the advance schedule. Schedule it only after planning, or treat it as the crew leader's current-day matter.
• If schedule compliance is low, then look at the allocation process and the staging before blaming the crew. Appendix B Rule 13 says compliance is not strictly under the supervisor's control.

• Under-assigning: the week is scheduled to an informal sense of 'enough' rather than to the forecast hours. The tell is that crews finish early or fill the gap with unplanned, self-selected work while the backlog persists.
• Scheduling from the roster instead of the forecast of skills actually available. The tell is that jobs are assigned to people who are absent or whose skill does not match, and compliance drops.
• Letting urgency replace priority. Only the loudest jobs get scheduled, so job priorities lose meaning. The tell is high-priority planned work aging in the backlog.
• Weak staging, so scheduled jobs stall on parts, tools or instructions. The tell is wrench time staying near the 35% baseline despite a full schedule.
• Using the schedule to hold individuals to single-job time estimates (Appendix B Rule 6 warns against this). The tell is that technicians pad estimates or hide delays in feedback, and the planner's estimates stop being useful.

• ▸Crew / week of: identify the crew and the week being scheduled.
• ▸Forecast labor hours (by skill level): list each person's forecast hours, using the highest skills actually available that week.
• ▸Total forecast hours: sum of the forecast hours for the crew; this is the 100% target.
• ▸Planned work orders selected (priority order): list each work order number with its job priority, taken from the sorted planned work.
• ▸Estimated hours and lowest required skill per job: enter the planner's estimate and the lowest skill level that can do the job.
• ▸Total allocated hours: sum of the estimated hours on the selected work orders.
• ▸Percent of forecast filled: allocated hours divided by forecast hours; keep adding planned jobs until this reaches 100%.
• ▸Staging status (parts / tools): for each job, record what is staged, where, and who staged it.
• ▸Weekly meeting date and open issues: record the meeting held, conflicts raised, and changes agreed.
• ▸Schedule compliance at week end: record the result and the reasons for misses, treating them as process signals.