Let's continue our discussion on the principles that guide world-class planning and scheduling. This article covers the application of schedule compliance.
First, we use the term "schedule success" to discuss "schedule compliance." Why does this make a difference? "Compliance" makes it sound like supervisors are responsible for whether or not they comply with the schedule. It makes it sound like they could increase schedule compliance simply by sticking to their guns and doing the scheduled work. Companies go so far as to tie supervisor pay to schedule compliance to show their strong support. Instead of "compliance," this leads to "stress" and supervisors with ulcers.
Place yourself into the role of a supervisor whose pay is tied to schedule compliance. You are working away through the week when a plant emergency occurs. You need to take your entire crew to head this one off and bring the plant back on line. Let's go. No! Wait just a minute. You now realize that if you drop the scheduled work to meet this emergency, it's going to cost you in the pocket at year end when they figure raises or bonuses. Helping the plant get back on its feet will cost your family money. What a choice. A good supervisor will honor the needs of the plant but eventually despise the management.
Measuring schedule performance with "schedule success" places an entirely different connotation on the whole situation. We all want the schedule to succeed, but if a plant emergency occurs, obviously that takes immediate priority at the expense of the schedule. It's not necessarily the fault of the supervisors if they don't comply with the schedule. It's necessary for a supervisor to break the schedule for work that can't wait. We rely on the skill of the supervisors in correctly judging when to break the schedule, not on their skill of ignoring urgent plant conditions.
Second, what is a good schedule success score? This is a trick question because if you aren't careful, it will lead you back to thinking of compliance and not success. John Crosson (a Clorox leader who spoke at the 1997 SMRP conference) says that schedule success is the "ultimate measure of proactivity". In other words, is the plant telling you what to do or are you telling the plant what to do? Low schedule success usually means that things are breaking more than it tells you that supervisors aren't working hard enough. High schedule success usually means that you are working on the correct things to keep breakdowns from occurring in the first place. The schedule success score is more of a measure of how decrepit your equipment is and if you are doing the right things to improve them or keep them in shape rather than getting enough work done.
With this said, I'd be perfectly happy in many plants with a schedule success of only 60 to 70 percent, or even much lower. Regardless of the score, the practice of scheduling leads to more work completion than otherwise, even while reacting to emergencies. We can't dictate that emergencies not occur, but we can make the best use of the remaining time and not encourage sitting back on laurels after fixing emergencies. Now, don't go and say "Doc said we should have 60 to 70 percent schedule success and tie supervisor pay to it". Your schedule success is determined by the current condition of your plant equipment, not the willingness of supervisors to forego pay raises. Your plant is what it is.
Third, and at most a minor issue, how should you measure schedule success? I've seen it measured many ways. I have little preference. I like to measure it in such a way to give supervisors the benefit of the doubt. I like to measure job starts and give supervisors credit for all estimated hours for any job on the schedule that they even start. For example, we give a crew 1,000 hours worth of work and they start jobs with estimated hours totaling 850. The schedule success is 85 percent. It doesn't matter that several of the jobs started weren't yet completed. You see, we know that many jobs stay in the backlog forever, but maintenance crews complete them as soon as possible after they finally start them. Maintenance typically doesn't leave jobs lying around half completed. So, our concern is to get jobs started.
I've seen equally good results simply using jobs instead of hours (e.g. I scheduled 30 jobs and you started 20 for a success of 67 percent). I've also seen success using job completions; but beware, this can lead to problems. It's frustrating to supervisors who complete most of a large job, only to get dinged if the job still has an hour or so left for the following week.
Also, if the plant has a complicated CMMS step that could delay recording the job's completion, this could lead to disparaging the scheduling measurement effort. Be especially wary of using paperwork completion as a condition. I've seen a few plants that have considerable computer entry requirements involving other groups getting around to the paperwork.
Finally, I really don't like giving partial credit for jobs. Take, for example, a job planned for 20 hours where the crew only completes 12 hours by the end of the week. The partial credit concept gives the crew credit for only the 12 hours. In practice, this partial credit idea quickly becomes very complicated. It requires checking on the partial status of every job (time consuming) and also presumes that there are only eight hours left anyway. In reality, the uncertainty of individual job time estimates might mean there are actually 30 hours or two hours left on the job planned for 20 hours. In the case of 30 hours left, why is it appropriate to give the crew credit for completing over half the work? In the case of only two hours left, the crew really should be given credit for much more work on the job. These are academic exercises at best and don't support our real purpose.
What is our real purpose in measuring schedule performance anyway? It isn't to see if the work got done. From a maintenance planning and scheduling standpoint, we measure schedule success to insure other good behavior to achieve our real purpose - namely, completing more work. Measuring success drives the fundamental practices of planning and scheduling, which is what gets more work done. You can't measure schedule success if you aren't doing weekly scheduling. And, you can't do weekly scheduling if you aren't planning work.
From a maintenance and reliability management standpoint, we measure schedule success to gain an indication of the reliability of the plant equipment.
Can we get through the week reliably without emergencies? With all the different aspects of proper maintenance working together, we can.