This article presents a gap analysis for a maintenance planning and scheduling program of a typical plant. Many of the plants Doc Palmer has visited have common opportunities to improve. This gap analysis presents these areas and recommends how they should be handled. The discussion helps clarify the purpose and role of a maintenance planning and scheduling program. It presents actionable strategies for improvement. This gap analysis would be presented to company management.
To represent our fictitious typical company, we will use the name ACME. We will look at a brief review of what a maintenance planning and scheduling program is all about. Then we will look at what ACME does well. It is critical to identify what ACME is already doing well and why. Many companies already have in place certain things that support good planning programs. They are in danger of unwittingly discontinuing these best practices if they don’t know how important they are. Then we will look at opportunities for improvement. Many of the opportunities are recognized from a better understanding of what maintenance planning and scheduling is all about. Next, we will examine good KPIs to manage the planning and scheduling program. Finally, we will conclude with four guiding principles to ensure continued success of the planning and scheduling program.
The Onsite Visit
The gap analysis begins with training. I led a four-day onsite effort at ACME, including a two-day maintenance planning and scheduling program best-practices workshop. This class was followed by two additional days doing one-on-one mentoring, primarily with the planners and change agent. I also spent some time with some of the supervisors and technicians. Throughout the site visit, I developed the gap analysis in my mind. I finalized it by the end of the week. In the class, I had the change agent, planners, supervisors and even some of the technicians. I led the class before the one-on-one mentoring so that during the mentoring I would not have to constantly explain why I was asking certain questions.
ACME requested that I do some one-on-one work before the class so the class could be more specifically tailored to the actual plant conditions, but I preferred to do the class first to get everyone on board with a common language. Doing the class before the “inspection” also gave me the freedom to teach “common best practices” where the students had the freedom to listen with an open mind, knowing that they would get an opportunity to show me the “real world” afterward. Developing a common language first allowed us to communicate better later and reach a common understanding. I developed the formal gap analysis for presentation with the ACME change agent. It was “our” report.
The Essence of a Maintenance Planning and Scheduling Program
Most planning systems fail because they think planning is to provide job plans so there are no problems during job execution. This thinking sounds good, but in practice, management tells craftspersons that they “will no longer have to hunt for parts” and because of that, the maintenance planning and scheduling program fails. Planners cannot plan the perfect plan, and by inferring that they can, craftspersons complain about the planners for every mistake. Planners then spend too much time on each plan trying to be better, but to the detriment of getting all the work planned.
In addition, planners feel obligated to help jobs that have already started in execution and have run into problems. This help on jobs already in progress further keeps the planner from planning new work. The result of such programs is extreme frustration. Instead, management should position planners as “craft historians.” While they faithfully give at least a head-start on most jobs, they primarily function to save feedback that the craftspersons give them. The planners “real job” is to save and then apply the feedback to future work on the same asset. Planners should be running Dr. Edwards Deming’s plan, do, check, act cycle. The intent of this cycle is to put a system in place to continually improve maintenance work over the long run.
But planning also supports scheduling because scheduling requires the skill and labor estimates. Similar to planning, the practice of scheduling leads most companies into frustration because they miss the whole point of scheduling. The purpose of scheduling is not “to complete the schedule” as it would appear many companies think. If that were the purpose of scheduling, a company would only schedule the work it knew it would complete and then complete that work. This apparently logical statement does seem to make sense. But companies take it to the extreme of only scheduling, say, the three jobs they know they would complete anyway. So they can share a bonus or meet a company mandate to have high schedule compliance.
A better company might instead insist that a full schedule be established with much detail to promote better coordination on the hopes of being productive, but even this approach is off the mark. Extremely detailed schedules lead to frustration because of the very uncertain nature of time estimate accuracy. In real life, many jobs run significantly over or under their original time estimates, regardless of a planner’s best attempt to make accurate estimates.
Instead, the larger purpose of scheduling is simply to promote better productivity. So, in practice, experience has shown that providing a crew supervisor with a simple list of work orders that matches the anticipated labor available for the next week greatly improves productivity. The scheduler performs the role of “backlog researcher” coming up with the appropriate batch of work for the anticipated labor capacity considering priorities and opportunities for common lock out tag out. To emphasize this point, let me repeat: We should start off every crew with a fully loaded schedule, but not dictate when during the week to do the work if possible.
This list focuses the supervisor and discourages simply waiting on operations and otherwise keeping crafts “busy” with backlog work. Instead, the simple list focuses the crew on trying to complete a certain amount of work. Such use of scheduling as goal setting does increase productivity beyond what a crew would normally complete. And that accomplishes the objective of a maintenance planning and scheduling program.
Interestingly, the important caveat to conducting scheduling as goal setting is that it must be OK to break the schedule. Then it works. Some supervisors insist you should give them the amount of work they would normally accomplish. This is the argument to win. No, we want to give them more so they will complete more than normal. But they won’t get it all done. That is OK. It is OK to break the schedule. That’s how goal setting works.
Companies need this extra productivity to eradicate normal levels of reactive maintenance. Companies can have good plants with significant levels of reactive maintenance. To become great plants, companies must go beyond what is normal and increase proactive work at the same time reactive work is present. This reduction of reactive work can be done if plants use planning and scheduling to “increase their work force without hiring.” Usually when the productivity picks up, plants also find they are not very good at identifying proactive work.
Another significant benefit of a maintenance planning and scheduling program is that it frees up supervisors to be in the field with their crews. Rather than dictate rigid procedures and ram schedules down the throats of supervisors, proper planning and scheduling does work that would otherwise keep the supervisor tied up. Planners do the legwork before work starts to try and get things ready to go. Later, planners take care of the feedback and put it away for future application. Scheduling (many times by the planner) performs backlog research by combing through the ugly “black hole” of maintenance looking for appropriate work. All of this is done so the supervisor won’t have to do it. It is a service. It is a help.
A better maintenance planning and scheduling program has a host of good results for the plant. More proactive work heads off not only reactive work, but the collateral damage of reactive work. Increased productivity snowballs into making the black hole of maintenance more visible and encouraging operators to tell us more about little symptoms instead of waiting to tell us about breakdowns. If we can do more work, we can also do more projects in-house saving money and accelerating completion dates. And perhaps most importantly, we can institutionalize knowledge through the planners capturing feedback and applying it to make future work more effective.
ACME Current Best Practices
Organizationally, ACME already has a number of things going well to support a maintenance planning and scheduling program. Management clearly supports the program and has set a specific change agent in place. These two items often make the major difference in the success of a maintenance planning and scheduling program. The maintenance structure is “clean” without overlapping areas of authority at the crew level which might otherwise frustrate planning. Planners are assigned full-time and do not share any crew supervising assignments.
The planners are also separate enough physically in their seating to avoid excessive interruptions. At ACME, they are also positioned at the first level of supervision which makes planning an attractive position. The planners also have appropriate rights to manage job plans in the CMMS. A maintenance clerk is available to handle some of the typing and data entry which might otherwise bog down planners or supervisors. ACME is also using an operations person as a gatekeeper to help control workflow and to be available to advise planners from an operations’ perspective.
Technically, ACME also already has a number of good features in its maintenance planning and scheduling program. ACME has a good CMMS for planning. I like five levels of priorities, but ACME’s four levels are adequate. Even in this age of electronics, I do like ACME having paper work orders in the field to help craftspersons record feedback as they think of it and not having to wait until later to enter computer feedback. The crafts are already giving some feedback on job plans, which is a good start. Currently, the planners are focused on building parts lists (bill of materials), which is an essential part of planning. Finally, PMs come out already in a planned status with warehouse support. (It would be a waste of planner time to continually verify that parts are ready for routine PMs.)
All of the above current best practices are worthy of mention because many companies do not get them right. Good job, ACME.
ACME Opportunities
ACME does not have any “problems” in its current implementation of planning and scheduling, but it does have some “opportunities.”
As far as scheduling goes, ACME needs to develop a methodology to provide crews each week with a batch of work that matches 100 percent of their labor capacity. A short weekly scheduling meeting should also be used to legitimize and sustain the process. ACME also has a number of standing, blanket work orders to charge for small jobs, but should try to eliminate as many of them as possible. Blanket work makes fully loaded scheduling difficult, and blanket work orders often get abused and used for larger jobs.
Currently, many jobs are completed or worked without status updates in the CMMS. This makes scheduling difficult because of the uncertainty of which work is still untouched. In particular, ACME should do a better job changing competed work to “comp” and changing other work that will at least be started in the week to “inprg” no later than near the end of the week shortly before the scheduler starts on the next week. ACME should also encourage more originators to avoid using priority one or two for work that could wait. (Here is one case for making the priority system five levels instead of four; maybe more persons would pick a one or four if there were a level five.)
To emphasize some role clarity, the planners should create the weekly schedule and not get involved in the daily schedule including not going to any daily schedule meetings. The maintenance planning and scheduling program mission of focusing on future work often gets lost in daily schedule meetings and the planners should not participate in them.
The primary opportunity in planning actually is the creation of reusable job plans. Planners simply must plan in the specific planning module provided by the CMMS. Currently ACME only uses this module for PM job plans (as do most companies). The planners should focus their efforts on continually improving these plans over the life of the plant and getting good feedback from crafts for this purpose. One idea that really encourages feedback is when the planner adds a procedure history section at the bottom of the printed job plan. The craftspersons really appreciate seeing a running list of how the planner has updated the plan based on their feedback in the past.
Another major opportunity for a maintenance planning and scheduling program is to plan the urgent work, but only after checking with supervisors to see if the work has not yet started. Without ever telling a supervisor to wait, a planner can quickly plan even urgent work if it has not yet begun. Because plans do not have to be perfect, planners can quickly inspect the job site, look at the file, and provide a simplistic plan that runs the plan, do, check, act cycle. The quick plan also allows that work to be scheduled for the next week if the job never ends up starting the week it was identified.
There are some maintenance planning and scheduling program opportunities due to the ratio of planners to craftspersons at ACME. In industry, planners can normally plan for 10-20 persons. Because ACME has a better ratio, ACME planners can provide more help on jobs in progress than is normally possible. ACME also has some ambiguity of who is supposed to write follow-up work orders and the planners could provide this service if the crafts would clearly describe the requests. This would keep crafts off the computers a bit more.
Planners frequently have to multi-task occasionally helping a job-in-progress while preparing or editing other job plans. Simply providing planners a second computer monitor greatly helps them switch back and forth without getting lost in what they are doing.
A few opportunities are available with regard to estimating hours. PMs should be reviewed overall especially with respect to hour estimates. They should also not be priority 2 (urgent schedule breakers). Not just in PMs, currently planners seem to be biased a bit too much toward making 1, 4, or 8 hour estimates. There should be more 2 hour jobs, 5 hours jobs, etc.
An opportunity also exists for operations to write better work requests. ACME should always encourage the operators to give as much detail as possible including when the problem occurred and even suggested solutions. Because one thing that really helps planning is to have the equipment numbers on the work requests, the plant could have better field tagging and may also want to consider deficiency tags.
Finally, under the category of miscellaneous, and somewhat out of the area of planning, the warehouse seems a bit small for this ACME facility. Dr. Deming really makes a point of the necessity of the purchasing function supporting the operation of the plant. Many plants simply do not carry enough stock to support effective maintenance. On the other hand, many plants never record stockouts because craftspersons are so good at “making do,” often with alternate parts that may or not be adequate. Management is fully justified in cutting stock if we never record any stockouts. Encourage craftspersons to record in their feedback whenever the warehouse did not have something it should have had.
Typical KPI Reporting
The KPIs preferred for scheduling are schedule compliance based on fully loaded schedules, work orders completed and backlog trend. Schedule compliance should be called schedule success to avoid blame on the supervisor. Schedule success is one of the most dangerous indicators in maintenance. Most companies insist on 90 percent or above, which leads to under-scheduling or inflating plan estimates. Best-in-class companies aim for between 40 and 90 percent. Schedule success is more like a bowling score than a school grade. The score is only meaningful for a fully loaded schedule of near 100 percent of available labor hours. A 50 percent schedule success score for a fully loaded crew means 55 percent wrench time, whereas a 100 percent schedule success for a half-loaded crew means 35 percent wrench time. It is that simple. Mind the store instead of minding the score.
I do like wrench time studies. Because they can cause workforce suspicion, I would rather watch work order completion rates. Normally, when we start honest weekly scheduling, the work order completion rate rises: success. The backlog mysteriously drops, which means we are not very good at generating proactive work. That tells us to get busy in that area.
The KPIs preferred for planning are planned coverage, the creation and revision of job plans, and how many emergency and urgent work orders are interfering with calmly planning the work. Planned coverage means how many jobs actually completed in the field had the benefit of a planner first looking at the job site and then looking in the CMMS to see if there was anything we learned in the past.
I would prefer at least 80 percent of our work be covered by a job plan. To get credit for a job plan, the work order must reference a job plan module number (hence, we are running a Deming cycle) and have estimated hours (hence, we are supporting scheduling). (By the way, we do not have a special KPI for job plan quality. All notions of job plan quality were resolved when ACME first picked its planners. Job plan quality is not addressed in KPIs.)
If the planners are making reusable job plans, the number of plans in the job plan module should rise. I would also expect the number of revisions to also rise as we update job plans with feedback. Urgent and emergency work orders also matter to the planning department. Planners must drop everything and help emergency work in progress. They must jump through hoops to quickly plan reactive work. Finally, planners should help the warehouse by noticing when the warehouse does not have active purchase orders or has zero quantity on hand for items that have levels below the CMMS minimum. Tracking these inventory items and taking action is the appropriate proactive method to avoid stockouts in the first place. But also track stockouts.
A final mention of schedule success is warranted. Tracking schedule success is a great way to encourage supporting behavior. ACME can only measure schedule success if it has a schedule. It can only have a schedule if it has enough planned work. The company can only have enough planned work if the planners manage their time to plan all the work and do not become bogged down in trying to make a few jobs “perfect” at the expense of not planning others. The planners can only plan if ACME protects them from excessive other duties. Finally, most plants that successfully implement goal-driven scheduling without over-worrying about the score, find they can do more proactive work than they are used to generating — a great problem to realize.
There are really only four things ACME has to do to excel with their maintenance planning and scheduling program: 1) Conduct planning as a cycle of improvement and do not insist on perfect job plans. 2) Conduct scheduling as a weekly goal of work based on 100 percent of crew labor capacities. 3) Let everyone know what we are doing. Providing imperfect job plans and fully loaded schedules that are OK to break is not “normal,” but we do not want to be a normal company. We want to be a great company. 4) This takes persistence.
Be a great company. Excel with a great maintenance planning and scheduling program.
Author
Doc Palmer, PE, MBA, CMRP is the author of McGraw-Hill’s Maintenance Planning and Scheduling Handbook. Palmer has over three decades of experience primarily as a practitioner within the maintenance department of the Jacksonville Electric Authority. From 1990 through 1994, Palmer was responsible for overhauling the existing maintenance planning and scheduling program. Retiring from JEA in 2007, Palmer now guides, mentors, and trains companies internationally for maintenance planning success as managing partner of Richard Palmer & Associates, Inc. For more information visit www.palmerplanning.com or contact Doc Palmer at docpalmer@palmerplanning.com or (904) 228-5700.