Over the years, your work management processes have improved. You have established a direct alignment with ISO 55000. You are seeing improvements against your SMRP metrics and getting closer to best-in-class values. You have a robust problem-solving approach that gets to the root cause and puts actions in place to prevent reoccurrence. You have proactive processes such as your preventive maintenance optimization program and the equipment blitzes validating your FMEAs. Confidently, you find yourself submerged in a recent failure analysis of a steam exhaust fan and eager to prove to your boss with MTBI, MTTR, and MTBF a no-brainer upgrade. Damn it.
You find out that all historical work on this fan had been written to an asset higher in the asset hierarchy and not to the fan itself. You have been focused on improving your metrics but realize that the asset selected on the work orders is the easier selection versus the right one. Instead of dozens of work orders with failure modes to underwrite a capital upgrade, you must choose if you want to manually decipher thousands of work order descriptions to underwrite the upgrade.
Intent of record retention within a CMMS
A record retention program is an administrative program and set of expectations by an organization that ensures records are accurate to reflect life cycles. They are used to ensure efficiencies are tracked and evidence is robust for future objectives when needed. Within a CMMS (computerized maintenance management system), the record retention is organized with an asset hierarchy to organize future work and capture completed work. But how do you measure the effectiveness of record retention in a CMMS versus an indolent selection?
Asset Hierarchy as it relates to record retention
The asset hierarchy is a structured, tree-like system where higher levels represent a broader grouping of assets, and the lower levels represent the granularity of individual components. Typically, when writing a work order in a CMMS, the composer can walk-the-tree with a visual navigation tool to find the lowest possible asset that represents where the work is to be completed rather than a higher level. For example, work on a pump called Pump 1 at a facility could be Plant > Production Unit > Operating System > Assembly > Pump 1 (See Table 1 - Asset Hierarchy Example). In this example, work done on Pump 1 would be considered five levels into the asset hierarchy. Doing work on Pump 1 with a work order attached to the operating system would not be the lowest possible asset representing the work.
For most work orders in a CMMS, the composer could select any level higher than the fan to save themselves a few clicks and seconds. The composer may not be familiar enough with the hierarchy to know where to drill down. They may give up by selecting something higher. Inevitably, this lack of thoroughness would result in a work order not being written deep enough into the asset hierarchy. But how do you measure to improve the quality of the drill-downs to validate if you are walking-the-tree far enough?
Plant | Production Unit | Operating System | Assembly | Component |
Plant A | Unit 1 | Main Hydraulics | Pump Assembly 1 | Pump 1 |
Plant A | Unit 1 | Main Hydraulics | Pump Assembly 1 | Motor 1 |
Plant A | Unit 1 | Main Hydraulics | Pump Assembly 2 | Pump 2 |
Plant A | Unit 1 | Main Hydraulics | Pump Assembly 2 | Motor 2 |
Plant A | Unit 1 | Main Hydraulics | Pump Assembly 3 | Pump 3 |
Plant A | Unit 1 | Main Hydraulics | Pump Assembly 3 | Motor 3 |
Table 1 - Asset Hierarchy Example
How to measure if you are writing a work order to the right asset
No one has the resources to validate if every work order is written low enough in the asset hierarchy. Therefore, we need to measure trends and behaviors that would indicate how well work orders are being written to the expectations of the record retention program. This can be measured by calculating the Asset Hierarchy Depth and Asset Hierarchy Altitude.
Calculation 1 - Asset Hierarchy Depth
We can count the levels between the top asset and the specific asset selected for the work order. This depth in the asset hierarchy can represent how far we are going down to capture our work. Using Table 1 as an example, writing a work order to Main Hydraulics would be valued at three and writing a work order to Pump 1 would be a five. Measuring this metric monthly on a time-series plot to look for patterns and behaviors is recommended. There would not be a best practice or a target to achieve, but instead, trends analyzed to see the thoroughness of walking the tree. See Calculation 1 on Table 2 - Example of Work Orders.
Calculation 2 - Asset Hierarchy Altitude
The second method evaluates the amount of assets below the asset selected. In Table 2, there are six work orders written to assets. Interpreting the work order's intention with the description, you can conclude the first three would be non-compliant and the last three compliant. In the Asset Hierarchy Altitude calculation, we would trend the reciprocal of how many assets are below the asset selected. If the work order is written to the lowest asset, the value would be 1/1 or 1. But, using Table 1 as an example, if you write a work order against Main Hydraulic to indicate work to Grease Motor 2, nine unique assets are lower. This results in a calculation of 1/9 or 0.11 (See second row, Calculation 2 on Table 2 - Examples of Work Orders). Similar to Asset Hierarchy Depth, there would not be a best practice or a target to achieve in this, but trends used to evaluate the performance of walking the tree. Trending this metric would be with a time-series plot.
Work Order Description | Work Order's Asset Selected | Compliance | Calculation 1 | Calculation 2 |
Replace Pump 1 | Unit 1 | Non-compliant | 2 | 1/10 |
Grease Motor 2 | Main Hydraulics | Non-compliant | 3 | 1/9 |
Tighten base bolts on Motor 3 | Pump Assembly 3 | Non-compliant | 4 | 1/2 |
Align motor and pump on Pump Assembly 1 | Pump Assembly 1 | Compliant | 4 | 1/2 |
Replace Pump 3 | Pump 3 | Compliant | 5 | 1/1 |
Inspect Motor 3 input bearing | Motor 3 | Compliant | 5 | 1/1 |
Averages | 3.83 | 0.52 |
Table 2 - Example of Work Orders
In record retention programs, we want the work order history written to the lowest asset, and checking if every work order is compliant is impractical. Instead, these calculations can identify where improvements, laziness, and lack of training exist. When done correctly, you will have the information necessary to calculate accurate MTBI, MTBF, and MTTR. Doing this wrong, well… good luck.