Capturing the Right Data Supports both Failure Analysis and Weekly Scheduling

Gatekeeper Role/Position -- Make This Your Next Hire

Everyone complains about the accuracy of the CMMS. The next complaint is about value, i.e. are we just making work orders and closing them out? And then we get into failure coding and reporting. Plus there is weekly scheduling. If you had a gatekeeper you could solve all of these problems (my opinion of course).

There are Two Advanced Processes that are Most Significant

These are (1) chronic failure analysis and a (2) resource-leveled weekly maintenance schedule. Historically almost everyone struggles to perform/generate either. The reason is the lack of good data -- and perhaps some programming. The only real way to get accurate data is to have a 2nd check reviewer, i.e. the gatekeeper. This role would verify data during WO initiation and at WO completion. And, this role would enter data that only he knows how to enter, and that is the rough estimate. If we know the lead craft, then we can programmatically read the actuals (work order labor history by craft) and come up with an average. This value would be defaulted onto the WO but still allow the gatekeeper to override.

Imagine a Moment when 100% of the Backlog has a Craft Estimate

Now, we can do real size/growth trending -- and not just by work order counts. But even more importantly we can perform automatic resource leveling. Note: the key word here is automatic. But before we start explaining the leveling program, I need to know WHICH work order to process first.

But Your Work Order Prioritization Scheme is Hogwash

Sorry. If you have a 700 work order backlog and a new job arrives today, what should be the WO priority? Is this new job more important than the existing backlog? Static prioritization methods are flawed. Plant conditions change. Project improvements affect existing work. Therein the backlog needs to be periodically reviewed. But you and I know this isn't happening -- nor would it be feasible. Lets assume your priority range is 1,2,3,4,5 whereby 1 is emergency and 2 is urgent (meaning they both get dispatched immediately). That leaves you 3,4,5 to differentiate the 700 work orders. Sounds like FIFO will rule the day when ties exist.

How to Create a Numerical Ranking of the Backlog

The objective is to create a numerical ranking using a pre-built WO matrix. This would be applied nightly.

The above matrix reflects what the core team (and functional user groups) want as to left-to-right. For example, some may want INPRG work to precede PM work. Plus you could insert as many vertical columns as necessary for your site.

The asset criticality is a very basic design, being 1 through 5. But, you could implement BRE.

Then, with an order of fire (using the ranking value) in place and a total estimated backlog (and craft availabilities) we can create a set of work for the planners to begin planning. A new automation script for leveling work inside the CMMS would be installed.

The Chronic Failure Analysis (CFA) Process

There may not be another initiative that offers greater opportunity to reduce O&M costs. A reliability action team would run the failure analytic (selecting functional failures) to extract bad actors using a sort-metric called Average Annual Maintenance Cost divided by Replacement Cost. This is the #1 SMRP economic indicator. With the bad actors identified, the reliability team then dynamically drills down on failure mode arriving at cause.