Why 80% of the failure codes reported into CMMS get wrongly coded with the words 'Other', 'N/A' or 'Failed to Function'
Do not be surprised if the ratio of useless failure code reporting on maintenance work history is the same in your company. It is completely understandable and explainable. You will need to investigate and find out why poor failure history reporting happens in your operation if you are to correct the underlying causes.
I am working on the rationalisation of equipment failure codes for a large oil company as part of my project dissertation. A survey done by the company recently covering a period of six months uncovered that over 80% of the failure code reporting into their CMMS's worldwide was "Other", N/A or "Failed to Function". Obviously this kind of information is not helpful to reliability engineers in their quest for continuous improvement.
Part of my work is to try to unearth some of the reasons for this. My initial research suggests this is a common problem all over. I would be most grateful if you could furnish me with some ideas as how to go about this
What you say about work order reporting is true in many companies. The feedback received on the maintenance work done is terribly inadequate for reliability failure analysis.
Since you asked ‘how to go about this’ and not ‘what to do about this’ I will tell how I would tackle the task.
Very importantly you should speak to the people writing the work order reports, and find-out why they think it is okay to feedback details in a job report with only the words ‘Other’, ‘N/A’ and ‘Failed to Function’? (What does N/A mean?) What are the thoughts going through their heads that justifies to themselves that it is acceptable to put so little of value down about what they did, and about what they found? Do they know why a detailed report with excellent content is important and what trouble they cause people elsewhere in the operation when the close-out content is poor? Also ask the tradespeople what should be done about helping them to complete job report-backs with the necessary details.
I would also ask the same thing of their supervisors and their managers—why do these senior people accept such poor content for work order history? What can they actively do about it? It might be worth your while to develop a properly structured questionnaire (not with leading questions) to ask your questions of the whole workforce; as well as doing face-to-face interviews with selected people.
I would also be looking at the culture and corporate norms of the operation. What are the values and beliefs of the crew, their supervisors and their managers on job feedback, information management and reliability improvement? (Plus whatever else you sense is worth asking about the latent factors brewing under the surface in the operation.)
I’d also be curious to know how literate are the people who work in the maintenance area, and how well they can spell and write (and the same query for the crew supervisors). A lot of tradespeople are not well educated and cannot explain their ideas very well. They would be especially worried that they would look silly if they know their reports will be read by university qualified people.
Have a look at the job feedback process requirements and any forms or screens to be completed. Is the close-out report form very user friendly and easy for people to navigate and use? Is the close-out process intuitive and easy, or does it drain energy and time because it is confusing or convoluted? Does the report ask leading questions that people have to answer with specific details, or are the fields left blank for free text to be created during job close-off? How much time are people given to complete a work order report in sufficient detail? Do people know how much time they are allowed for closing-out work orders properly? Are people overloaded with jobs and rushing to close-off work orders as fast as they can?
Are the fault codes used during job close-out self-explanatory and clear in what they mean? Are the fault codes easy to differentiate in meaning and not interchangeable? Are there few fault codes or so many fault codes that it is hard to choose one from another—too many codes just confuse people? Is the choice of fault codes available actually useful. Using 'Failed to Function' on job close-out looks right to tradespeople because the equipment did fail, but that isn't what the reliability engineers need to know. They want to know about the failure mode and the failure mechanism. Do the people closing-out a work order know what the words 'failure mode' and 'failure mechanism' even mean?
What training are people given in writing good feedback from the jobs they do? How do people learn what they should write when they close-off their work orders? Who is responsible to teach people what to write about when closing jobs off? Do these ‘trainers’ do job history reporting very well themselves?
One more area to look into is whether there are documented standards of what excellent work order history reports contain for each trade discipline (electrical, mechanical, etc). Are there many reference examples of what a great work order report looks like? If there are good examples, are they right at the computer terminals or writing desks where the tradespeople can see them and mimic them? Can people see a top quality job report when they need to refresh their memory of what to include? Do the supervisors and managers know what top-class work order feedback looks like and can they show you an example quickly, without asking where it is kept?
Are there any role models of people that write top quality work order history reports? Find-out what makes these people different. Do these role models teach others, or do they ‘keep it to themselves’?
The only people that really know what content they need from maintenance history are the reliability engineers using the information for failure analysis. Can the reliability engineers define and specify what content is needed and how the report should be written? How many of these engineers are out with the crews teaching the maintenance tradespeople how to write great job history content in their reports?
How much of the problem is actually business system/process related and how much is peoples’ misunderstandings, ignorance and beliefs related?
You should also read this FAQ response on Problems Tracking Downtime Causes With Equipment Failure Codes and Fault Code.
Good luck, my friend. You will learn a lot about human frailties and poor business processes in doing your dissertation.
My best regards to you,
Lifetime Reliability Solutions HQ