Planners should be running a Deming Cycle in maintenance by slowly improving jobs as they have more time and as they receive actual job feedback over the years. This third principle tells us how to save this information as we grow “living” job plans. Planners should save job information in files (paper or computer) that are associated with the individual component being maintained.
Consider a work order written to do some work within a manufacturing plant. This plant has an extensive amount of equipment in a hierarchy. Under the plant level, the hierarchy has several product lines as well as facility systems, such as for utilities. For each product line, there are a number of subsystems, such as conveyors, chemical reactors, heat exchangers, and holding tanks. Within each subsystem, there are individual pieces of equipment, such as valves, pumps, piping, agitators, level indicators, and other control devices.
The work order is for a specific valve, say, “anion regeneration valve #2 on the anion tank.” The plant needs to keep maintenance information in a separate file for this specific valve. If the plant keeps information at a higher level – say, the entire anion tank that has several valves, pumps, and other associated components – it simply takes too much time for planners to find the information and plan all the work for the specific pump in question.
Think of a dentist’s office. The dentist does not keep all the information massed together in a single file for all the patients on your entire street or for your entire neighborhood, and certainly not for your entire city. Instead, the dentist can quickly go to the file for a specific patient and treat the current reason for the patient being there, knowing all the past results from previous visits for that same patient.
Similarly, plants should generally not keep one single file for the same model of a device. The strategy of having a single job plan for all 2-inch valves of the same model is tempting, but each valve might have different histories and unique failure modes. One valve might have a recurring problem with insulation; another might have a recurring problem because of the chemical it handles; another might be accessible only by scaffolding; and yet another might have a unique operating context where the operator has to make certain provisions to allow for maintenance.
Recurring problems do happen, and they are often unique to specific devices, not to the model. Other circumstances make for unique situations even for the same model. Planners should err on the side of having specific files for specific devices. The equipment number is critically important for keeping information for component level files.
Names generally are not good enough for proper filing for two reasons. For one thing, many people call the same equipment by different names. For another, there are usually so many different pieces of equipment in the plant that even proper names become lengthy.
Numbers, whether they have intelligence or not, are usually better for filing. (For our earlier example, a full component number might be N01-CP-AR5.) For paper files, this number must be on the work orders and the paper files. For paper files, I generally like using open shelves with side labels on file folders rather than closed filing cabinets. The little bit of time it takes planners and craftspersons to open and pull open cabinet drawers seems to discourage them from using them. For computers, we must associate electronic work orders and data with the component number. A CMMS should allow utilization of equipment information just as easily, if not better.
Finally, take the time to tag the equipment in the field with the component number. This helps operators use the right number and not pick the wrong number by drilling down in a computer hierarchy. It also helps the planner and craftspersons find the right equipment in the field.
Keeping equipment information in component-level files helps planners run the Deming Cycle of learning in maintenance by allowing planners to quickly save and later access helpful data to improve the next job.