Why focusing on CMMS due dates can lead to bad maintenance behavior
Measuring weekly schedule compliance simply by satisfying the work order due dates that are set in the computerized maintenance management system (CMMS) drives some bad behaviors. People end up not reporting little problems that could head off big problems. And they only report big problems when they start to work on them. Instead, best maintenance practice is to encourage the reporting and remedy of all the little defects we can find to avoid reactive work altogether.
To begin with, I favor priorities based on due dates rather than multiple adjectives. (With adjectives, where do you stop?) Consider the following very simplistic priority system:
Priority by Due Dates
0. Emergency - Start Now
1. Urgent - Complete 1 week
2. Routine High - Complete 2 weeks
3. Routine Normal - Complete 1 month
4. Routine Low - Complete >1 month
Priorities simply help us put work in the best order for completion acceptable to both Operations and Maintenance. Operations needs something done. Maintenance needs to know how long it can wait. Quick coordination must not be overly complicated by a slew of adjectives.
Most operators and maintainers can use their professional judgment (which is also somewhat checked by gatekeepers and morning staff meetings). The resulting priority of each work order pins down the agreement of the coordinating conversation:
Operator: “I need this thing done.”
Maintainer: “Does it have to be done right away or can it wait?”
O: “Well, it’s pretty important. It doesn’t have to be today, but it can’t wait too long.”
M: “Can it wait until next week?”
O: “That’s sounds okay. Let’s make it a Priority 2.”
Afterward, the maintenance force generally works in order of priority (with some bundling of less urgent work with more urgent work in the same area or asset). Maintenance completes the 0 emergencies, then the 1 urgent work, then the 2's, then 3’s, and then 4’s. In the short term, the finite-size maintenance force does not expand or shrink based on the workload. While we would like the workforce to be exactly sized for the work, sometimes the load of 1’s and 2’s exceeds what can be done this week or next.
Furthermore, I like scoring (or looking at) the median completions times for the different priorities. Are we generally completing Priority 1’s within a week, 2’s within a couple of weeks, and 3’s within a month? How long for a 4? It helps if we can show Operations that we generally complete a Priority 3 within a month. That encourages them not making everything a 1 or 2 so that it will not die in the “black hole of maintenance.”
Some CMMSs automatically set “due dates” as soon as work requests are entered with a priority. For example, when someone enters a new Priority 1 work request on a Wednesday, the CMMS sets the due date one week later as next Wednesday. This is not a bad practice in itself. The problem comes when a company measures “schedule compliance” as the percentage of all the work orders with a due date within that week that were completed within that week. So, if the company completed 60 of the 80 work orders that had CMMS-set due dates in that week, schedule compliance would be 75%. Yet, if only a single work order had a due date that week and it was completed, schedule compliance would be 100%. Therein lies the rub.
Meeting these CMMS-set due dates should not be used for measuring weekly “schedule compliance.” We should measure schedule compliance by the percentage of work orders that we scheduled (regardless of their priority or “due dates”) and that we completed. If we scheduled 80 work orders and then completed 60 of those work orders, our schedule compliance would be 75%. This measure tells whether schedulers properly loaded the schedule and supervisors properly used the schedule, as well as an insight to a host of other issues including operations, maintenance, engineering, and stores practices. The score tells us how well we are in control.
Using CMMS-set due dates for schedule compliance creates a bad situation. Low schedule compliance encourages Operations and others not to report Priority 3 and 4 issues, the little proactive items such as loose or missing bolts and tiny cracks or leaks. They only report the items that they know maintenance will do right away. They do not report anything where they do not know when maintenance will do it. Generally, this situation means people will report 0’s and 1’s as they begin to work on them this week. They will not report anything else because they cannot guarantee its completion time. They achieve high schedule compliance. But this score does not tell them how well they are in control.
These practices are totally the opposite of the behavior we want to drive. We want a huge backlog of proactive Priority 3’s and 4’s to prevent future reactive 0’s and 1’s. Then we want to fully load schedules to drive high productivity to complete all the 3’s and 4’s with room to spare for completing 0’s, 1’s, and maybe 2’s that pop up along the way. Incidentally, we will then have good schedule compliance as measured by CMMS-set due dates. But using that as a driving metric is too dangerous. It is not the way we want to measure schedule compliance.
We want a huge backlog of little things without anything breaking down. Surprisingly the schedule compliance based on CMMS-set due dates is met when we do not focus on it. Don’t settle for looking good. Be really great!
