Tip #126: Work Order Categorization
A work order filing system does not support data grouping, selection or sorting (although it does have a "paper sort", e.g. WO #, or Report date, or asset, or system). It's amazing to me that for so many years the file cabinets existed in the work place when computer systems were most prevalent.
What started my interest in this subject was an older posting from Terrance O’Hanlon regarding ideal work type design, and, …. “what is a PM”. Where most users struggle is with the bigger picture - which I call work order categorization. It is the first step to proper identification of the job. There can be different types of work, and categories of work, and even different spellings therein, e.g. preventive or preventative. But I believe the real key for this discussion is that for whatever design you come up with, there should be a reason for collecting all data, and that is, to (1) make more informed decisions, (2) manage by exception, and (3) create a true knowledge base. I call this "line of sight". Basic data capture starts as WO initiation; but this is also a critical moment for gathering data which will be used for KPIs, failure analysis and other analytical reports. Lets call the latter data type ... critical data.
In this case we are assuming an operations staff member creates the initial document.
First, some notes & understandings:
- Equipment Operators may perform regular "preventive maintenance type tasks" as part of normal start-up & operation of the equipment in support of operator driven reliability (ODR).
- Operators performing plant rounds do take readings (e.g. temperature, pressure, levels) as well as sensory observations, and in some cases initiate a work request if they discover a "need for maintenance".
- Operators also record plant significant events in a control room log book.
If a work order is created, then it needs to be properly coded and categorized. Validated fields allow slice-and-dice using SQL statements.
Here are some key processes which benefit from WO categorization:
- Differentiate between "repair work" and all other work. Let’s call this work type a corrective maintenance ticket (or CM type). This may be emergency, urgent or routine.
- Identify "shutdown" work either by asset, system, or unit; and if it is roughly a 24, 48, 72 hour shutdown or longer. If longer then it might be tied to a future planned shutdown.
- Identify work that alters or modifies the plant configuration (e.g. MOD). Sometimes, this is a “capital project”. Alterations to the plant should be routed to CAD coordinator for incorporation at job completion.
- Identify work that performs condition monitoring (or PdM) e.g. vibration analysis, ultrasound, infrared. This is scheduled work. And the Technician/analyst may discover potential failure which needs review.
- Identify preventive maintenance (PM) of equipment as to prolong life expectancy and discover undesired conditions. Scheduled PM activities might include lubrication, cleaning, adjusting, painting, and minor component replacement. A PM might also be called an “inspection”. Sometimes meter readings are captured.
- Identify “discovery work” (CBM work type) resulting from the PM or PdM activity. This might have a 45 day expected completion as follow-up work order.
- Sub-note to PM: there should be a “business rule” by the leadership that states…. additional work time allowed that can be spent under this PM-workorder to perform other repairs is 30 minutes, otherwise a CBM work order needs to be created (which can be planned & scheduled).
- WO PRIORITY: a very simplistic design is Operations chooses 1,2,3 where 1 is emergency, 2 is urgent, and 3 is routine. And then, Maintenance staff sub-divides the 3 into 3,4,5 for purposes of ranking the open backlog.
- Distinguish “call-out” work as it's own Y/N field. This differentiates Priority 1 from call-outs.
- Routine work should “be considered” for the weekly maintenance schedule. But depending on craft availabilities only a subset will “be selected”.
- Distinguish repair work Type Failure as to Full, Partial, Potential or Defect classification. The Full and Partial would feed the MTBF calculations whereas a Potential or Defect would not. A Potential work order may also be coded in terms of “cost avoidance”. Staff might receive an attaboys in this case.
- Minor maintenance might be a Y/N field and can be treated as "fill-in" work.
- Action verb is merely a one-word descriptor saying what was done, e.g. Inspect, Train, Assemble, Replace-Asset, Replace-Part, Modify, Repair, Adjust, Paint, Fabricate, Rebuild.
- Downtime flag: this might be auto-tripped by Full failure and “possibly” Partial. Note: this would also populate a Downdate field.
- Safety, Environmental, Regulatory issues (usually Yes/No flags)
- High Risk work: [a] Safety officer must be present, [b] Supervisor present, [c] No significant risk. An automatic email could be sent to HSE for [a] category.
- Failure data [entered under different moments in WO life]
- Asset problem code (excessive vibration, noisy, odors, smoke, overheating, function/fault)
- Failed component
- Component problem code
Other WO basic data include requestor name/date; location; asset; glaccount code; problem description (e.g. symptom); plant system; and asset criticality.
Output capabilities from critical data:
- Percent Repair work vs PM vs PdM.
- The amount of discovery work (CBM) from a PM or PdM
- Identify the amount of call-out work
- Identify percent REACTIVE MAINTENANCE (by craft, or unit, or plant/site).
- Calculate MTBF by asset for repair work with Full/Partial failure
- Tally total down time by asset
- Identify safety, environmental, and regulatory work
- Identify high risk work
- Calculate “cost avoidance” from potential failure
- Extract all replace (like-for-like) work
- Run the Asset Offender Report
When it comes to work order categorization there can be many successful strategies. And, rather than have 20 different work types I believe there is more flexibility in having multiple fields to choose from which in turn provides more combinations. *** Design with the end in mind ***
Either way, I certainly welcome your thoughts.
The Gatekeeper Role is the next tip (#127).
Book author, CRL, CMM and CMMS champion.
9 年the role of Planner/Scheduler is described in Tip #128. .... thanks Bill M.
Manager, Asset Management & Reliability
9 年I'd like to get everyone's input on how the maintenance planner fits into the key processes above. Looking forward to seeing the comments and hope this does not hijack the post, John Reeve!
Book author, CRL, CMM and CMMS champion.
9 年g'day Nuno, WO prioritization can sometimes be over-designed. If the O&M Supervisors show up at the weekly schedule meeting with a pen-on-paper list, then obviously they do not care about prioritization schemes. But staying positive..... the ideal design is a Calculated Priority which is normalized by combining the dynamic WO priority of the moment with the static asset/location priority. This can be used for backlog management and work scheduling. Keeping it simple, I normally propose Emergency (1), Urgent (2), and Routine (3) for the person initializing the work order - normally Operations (or customer). And then, the Maintenance staff (or Daily Review meeting) can further sub-divide the 3 into 3, 4, or 5.... or even 3-10. When the resource leveling algorithm starts processing, this is a common field to utilize.
General Manager, Assets Management (Jedco, Jeddah)
9 年John, regarding priority field, is in your method a value defined by the technician or supervisor, or based on the risk of the asset? We have another field called 'urgency' that is set at the incident detection, which is variable and set by the technician/supervisor, but the criticality is fixed based on the risk of the asset.
Book author, CRL, CMM and CMMS champion.
9 年Gday Hamid, Defect(s) are discussed/defined here. https://www.dhirubhai.net/pulse/tip-day-102-john-reeve Defect elimination is a significant value-add process.