In many plants and facilities, when a work request comes in, it is immediately dispatched to a technician who completes the request and then moves on to the next job. At best, some preventive maintenance (PM) is performed when scheduled. All of these work orders (WOs) are entered into a computerized maintenance management system (CMMS). Day after day, month after month, year after year, they seem to disappear into the system, never to be seen or reviewed again. This is how maintenance is typically operating these days (and has been for years). We need to do better. And we can.
Ask yourself these questions:
If you seem to be answering “yes,” you need to switch gears. Most organizations have some form of CMMS. You can use the data from this system to actually “drive” your maintenance.
Maintenance work-process flow
As shown in Fig. 1 below, a proper maintenance work-process flow should include eight distinct steps:
Unfortunately, most operations miss some of these steps: Follow-Up and Reporting Metrics/KPI, for example, are rarely implemented.
Proper CMMS implementation
What were your objectives for implementing your CMMS? You probably need to examine them. Based on this author’s experience, it appears that many operations use a CMMS merely to satisfy compliance requirements. If this is one of your objectives, it’s a good one: Compliance requirements must be met. If, however, compliance is your only objective, you should reconsider your strategy. A CMMS can do much more for your operations, including boosting labor efficiency and productivity. The four objectives on the next page are important when it comes to CMMS implementations:
The referenced data—all of it—comes from your CMMS. Thus, you need to properly plan for your CMMS implementation and data collection. For managers to make meaningful decisions, all data must be accurate. Inaccurate data will lead to faulty decisions.
It’s also crucial to get buy-in from all personnel involved with your CMMS. You can’t simply demand that everyone enter the right information. They will, though, provide accurate data if they’re convinced that there’s a good reason to do so.
Your objectives and the reports you wish to generate should be used to determine what type of data is input into your system. Procedures need to be in place to verify that everyone involved is entering sufficient data. For example, if you have data such as a request listed as “pump failed,” and the action taken listed as “pump fixed,” this data does not help to make meaningful decisions. Meaningful data should include problem/cause/remedy: the definition of the problem, what caused it and what actions were taken to correct the problem, etc.
Data review and follow-up
It is not enough for a WO to be completed and never seen again. Someone—usually a planner or supervisor—has to review completed WOs on a daily or weekly basis. These WOs will reveal invaluable information that can be used to drive your maintenance operation and improve efficiency.
Follow-up is basically driven by data. Consider backlog data, for example. If a maintenance operation has too little backlog, it is an indication that the operation is overstaffed. It would seem that the converse (a large backlog indicates that the operation is understaffed) would be true; however, this is a common misconception. The follow-up process, including a review of the data, will help determine the issue. A review of the data may show that the operation experiences a high percentage of emergencies. The data could also show that technicians are constantly interrupted due to additional emergency jobs and that they are assigned new jobs before they can finish the current job. The data could also indicate that a lot of time is spent looking for or waiting for parts. Some operations spend hours looking for parts: this should not be the case. These types of scenarios can make it seem as if your operation does not have enough manpower. Reviewing the data, using it to drive your operation and including follow-up will help identify these types of situations and help you improve your operations.
Scenario 1: Preventive maintenance. . .
— The planner may review the data to see how many PM jobs are completed versus total jobs. If the data reveals that PM is only a small percent of total jobs, this is an indication that not enough PM is being done. The goal of a successful maintenance operation is to increase planned maintenance to reduce breakdowns. By reducing breakdowns, you increase asset reliability and uptime, which, in turn, increases profits. Based on this author’s experience, many organizations today have a ratio of 10-20% planned maintenance to 80-90% unplanned. This ratio should be reversed.
— If you are not doing enough PM, it will eventually surface in the form of breakdowns, numerous repairs and emergencies. These problems shed light on the lack of PM that has led to these situations, and PM should be incorporated or the frequency should be increased to avoid these issues in the future. Overdoing PMs is often overlooked but leads to unnecessary expenditure. The author’s experience shows that many companies are following procedures that were entered into the system or developed 20-25 years ago. If a review of the data shows that there is no breakdown/corrective maintenance, the frequency of PM can be decreased. If weekly PM was being performed, you can try decreasing the PM to every other week. This will save 50% of your expenses on that piece of equipment. If you have multiple pieces, this will yield a significant savings. As part of follow-up, you also need to review your PM tasks. You will again find that many of these were defined a long time ago. You need to determine if each task has a value, and if it does not, the task should be eliminated to save time. With both tasks and PM frequency appropriated, it will lead you to an optimized PM program.
Scenario 2: Pump repair. . .
— Consider the following: A pump breaks down and the maintenance technician determines that a bearing is the problem and needs to be replaced, but he does not know the exact details of the bearing, such as the part number. The technician may spend hours opening the equipment to determine this information. Once he determines the needed bearing and goes to the storeroom, he may find that this bearing is not on hand and needs to be ordered. The technician may spend several additional hours locating a vendor to supply the needed bearing. The equipment is down during this entire process, costing the facility thousands of dollars or more per hour.
— Now consider the data-driven scenario: For each piece of equipment, a list of spare parts should be entered into your inventory database, along with the part vendors. When a pump breaks down and the maintenance technician determines that a bearing is the problem and needs to be replaced, it will take only a few clicks to determine the exact details of the bearing and a couple of clicks to determine if the spare bearing is on hand. If so, the repair can be completed. If the bearing needs to be ordered, a few clicks can provide a list of vendors who carry that bearing. The entire process takes minutes and can save the facility hours or days of downtime and thousands of dollars.
— You need to review the data for failure-code analysis, e.g., problem, cause and remedy (PCR). In this scenario, the problem was the pump, the cause was the bearing, and the remedy was the replacement of the bearing. That is just one example. There are hundreds of PCR codes. By utilizing these codes, managers can review this data and easily identify which problems occur often and on which equipment. Corrective action can then be taken, which will lead to higher asset reliability.
Scenario 3: Route-based preventive maintenance. . .
— Facilities have thousands of pieces of equipment—which translates into thousands of PM jobs. This includes equipment such as smoke detectors, heat detectors, fire extinguishers, fire doors, exit signs, pumps, motors, etc.
— Consider a facility with 600 fire extinguishers that need to be inspected every month. Each inspection only takes a few minutes (with equipment such as exit signs, inspection may only take a matter of seconds). Still, you have created 600 different WOs in the system, one for each fire extinguisher.
It takes more time to generate and complete each WO than to inspect the extinguisher.
— The data will clearly show that this process is counterproductive. The solution is to implement a route-based PM system. Instead of generating one WO for each piece of equipment, one or a few WOs are generated for the entire group of equipment. This could include one WO for a group of equipment per floor or building. This will save a tremendous amount of time and money.
Scenario 4: Recurring AC-unit breakdown. . .
— Consider the following: A facility has 20 AC units, all of the same model, in different areas of the facility. One unit has more issues and requires more repairs than any of the other units. If no one is reviewing the data, repairs may be made without noticing these repeat issues.
— The key is to use the data to determine the solution. Repeated repairs on the same unit may signal that the technician may be new and requires more training. If that is not the case, the unit may be older than the other units. The data will show that too much money is being spent on repairs, and it would be cost-effective to replace the unit.
In each case, the data is driving the maintenance, cutting cost and improving efficiency and productivity.
Scenario 5: Hospital tube-system repair. . .
— The tube system in a hospital is crucial and cannot be down for an extended period of time because the hospital relies on the system for important operations, such as transferring blood to different areas of the building. In this case study, when the tube system went down, the technician would go to inspect the system and spend eight to ten hours diagnosing and repairing the system, during which time the equipment was down. This process was repeated each time the system went down.
— After analyzing the data, it was discovered that the majority of the time, the problem was due to a failure of a swift motor. The technician would remove the motor, bring it to the shop, and sometimes find that a replacement motor was not in stock. The replacement motor would be ordered, and the damaged motor would need to be repaired. Meanwhile, the tube system would be down for hours until the motor could be reinstalled.
After the data analysis showed that this motor was causing the problem in the majority of cases, the hospital was able to use this data to improve the maintenance process. The hospital started stocking a couple of extra swift motors at all times so that a replacement would always be on hand. If the tube system went down and the information suggested that the motor could be the issue, the technician would take the replacement motor with him. The technician could immediately swap the damaged motor with the replacement and take the damaged motor with him to be repaired, reducing the downtime of the system to around half an hour.
These days, almost all organizations have some form of CMMS. In the majority of cases (over 90%), however, the system is vastly underutilized. Many organizations, in fact, use this valuable tool merely for recordkeeping and creating WOs. If that’s all that needed to be done, the organization wouldn’t need a CMMS. To fully leverage the power of your CMMS, it’s necessary from the outset to define the objectives for using it and enter accurate and meaningful data accordingly. As a follow-up process, the data must be analyzed to make meaningful decisions, which will improve both efficiency and productivity. MT