One of the key things a reliability professional is asked to do is look at ways their organization should be measuring their performance against known metrics. One of the best lists I've found is SMRP's Metrics & KPI's list.1 This is a list of recognized metrics and KPI's an organization can use to monitor the performance of their maintenance group. One of the most common metrics organizations can use is the PM to CM ratio. The goal for the PM to CM Ratio metric is to monitor how much Preventative(PM) and/or Predictive (PdM) maintenance work is being completed as compared to Corrective maintenance (CM). If the work a maintenance department is getting done is generated from breakdown or ad-hoc work, this ratio will be low (typically less than 1.0). If the work a maintenance department is getting done is generated from PM/PdM work, thus addressing issues before the equipment breaks down, the ratio will be higher (typically between 2.0 and 3.0).2 A common task most organizations don't complete is validating the metrics and information that is being collected. A reliability engineer, or maintenance manager, will start measuring a key piece of information based on industry standards, run metric for a while, but then never take any steps to see if the data is being manipulated or is still accurate from the initial assessment.
Take a look at the following chart3 showing the PM to CM work order ratio for a various plant sites:
From this chart, it would appear that Site #5 is doing a pretty good job at moving to a highly proactive PM/PdM environment. So how can we validate their efforts? One way is how many man hours are being put into PM/PdM work.
Huh.
While Site #5 has a 2.23:1 PM to CM work order ratio, it only has a 0.46:1 PM to CM ratio for man hours charged. Digging deeper into the root cause of the discrepancy, it was found Site #5 had started to shift from routed PM work orders to individual PM work orders. The site understood how this metric worked and found a way to alter the results of the metric without making any significant changes to their work processes.4
Maintenance Rule #5:
Validate your metrics or you don't know what's really happening.
SQL Query for Maximo KPI to monitor PM to CM Labor Hours:
Listed below is the SQL query to use in Maximo's KPI module to generate a PM to CM labor hour ratio. The entire query should be placed in the 'Select' box and the 'Where' box will remain empty. Make sure to leave the number type to 'Decimal'.5
select (select sum(labtrans.regularhrs) from labtrans join workorder on labtrans.refwo = workorder.wonum and labtrans.siteid = workorder.siteid where labtrans.siteid='BEDFORD' and labtrans.startdate >= dateadd(day, -14, convert(varchar, getdate(), 101)) and labtrans.startdate < dateadd(day, -7, convert(varchar, getdate(), 101)) and workorder.worktype in ('PM', 'PDM')) / (select sum(labtrans.regularhrs) from labtrans join workorder on labtrans.refwo = workorder.wonum and labtrans.siteid = workorder.siteid where labtrans.siteid='BEDFORD' and labtrans.startdate >= dateadd(day, -14, convert(varchar, getdate(), 101)) and labtrans.startdate < dateadd(day, -7, convert(varchar, getdate(), 101)) and workorder.worktype not in ('PM', 'PDM')) from dummy_table
-
Note: The PDF from SMRP has to be purchased for full details. ↩
-
Look for a future post on the philosophy of Proactive vs Reactive maintenance. ↩
-
Charts are from a test/development database and in no way reflect how I would operate a maintenance department. ↩
-
While the charts are fake, one of my plants thought this was a good way to make their numbers look better. ↩
-
The SQL statement for the KPI query were written for SQL Server. I'll update the post and give credit to anyone who wants to translate the statement for Oracle or DB2. ↩