CM & Lube Newsletter Article - Maintenance History

From Imrtwiki

Return to Maintenance History

Maintenance History

I recently organised a Common Interest Workgroup meeting on ‘Knowledge Management in Maintenance’. There were some great discussions and some interesting ideas about how maintenance people can share information and the challenge of handing over a maintenance role to the next person who takes over your job. After the meeting I started thinking about some of these ideas and how they might be applied to improving maintenance history recording for a plant, which is usually an area where people have problems and an area where there are usually large opportunities not being exploited.

I asked myself, what is the best practice in recording and using maintenance history? I first thought about some of the maintenance history information that I had personally put together when I was more involved with front line maintenance. Even though I got some good value out of these history systems, I ruled them out as best practice as I know that after I left those roles the systems and the information I had put together were rarely used again.

I thought about some of the great work I had seen done with failure and other coding systems in Computer Maintenance Management Systems (CMMS) and the value received from this. Even though use of good coding systems might be best practice in analysis and finding improvement opportunities, it does not necessarily mean the detailed work order history comments are good quality or useful. The hierarchy of data quality is Data, Information, Knowledge and then Wisdom. The ideal history comments are words of wisdom from what has been learnt.

I thought about some of the whinges I have experienced and heard over the years about maintenance history recorded in Computer Maintenance Management Systems.

• That the hand written comments made on work order documents don’t get typed into the CMMS
• That information that get typed into CMMS never get used (or that is the perception)
• That it is hard to find the information that is typed into the CMMS
• Once data is typed into a CMMS it becomes invisible and no one knows it is there
• That it is hard to get a sensible history report out of a CMMS for items of equipment
• History information does not go back far enough (often only to the last CMMS upgrade)
• That there are blank spots in maintenance history (usually when someone less diligent or less experienced was in control)
• That useful data is often hidden in amongst a huge number of more trivial routine work orders entries eg. Greasing of bearings or weekly inspections
• That there are too many Plant Index (Functional Location) levels and no standardisation at which level history should be recorded, making it difficult to create a simple history report
• That the data recorded is not the more important data that should have been recorded.
• There is no consistency in what data has been recorded and everyone that inputs data has a different view of what is important

I am sure if you have been involved with trying to use CMMS maintenance history you would have no problem adding to this list. My only added comment to this is that even if your maintenance history is poor; I have always found it valuable trying to use it. I have always learnt something worthwhile and received good payback for the time spent.

About 8 years ago when I worked at Bluescope Steel’s Tin Plate Rolling Mill (Port Kembla) I needed to review bearing failure history related to a recirculating oil system. When analysing the CMMS work order history I found many of the problems that are listed above. In frustration I eventually just went and talked to some of the more experienced guys that had been associated with the mill for about 15 years. The plant had a paper based maintenance history system predating computerisation in the form of cards stored in pullout trays. In comparison to computerised history the old manual card history was much easier to use and was consistent, concise and seemed to have no gaps. I was able to easily answer my bearing failure questions from this ‘old fashion’ system.

So why was the paper card history so much better than the comprised history? My thoughts on this question are:-

• The history was more complete because it was so much easier for a manager or anyone else to check that the history had been recorded. But doesn’t computerisation make data easier to access? For skilled CMMS users the answer is definitely yes but for an average or a non CMMS user, the answer is no. Some CMMS are better than others.
• It was more consistent because the person recording the history had all previous history comments recorded above and in cards behind. The person naturally followed the previous pattern of information recorded, which acted as a template. If it was not the right information being recorded someone would have been easily able to identify that.
• The history was more concise as lower importance information (like greasing and inspection activities) were never recorded and the limited space on the cards encouraged conciseness. The information often seemed to be a summary of more detailed information like reports that were probably stored elsewhere in filing cabinets. It seemed to have been written somewhat after the event, not during or immediately after, enabling a better overview.

I am not suggesting that maintenance history was better before computers but just that there were some instances of good history recording practices. Those good paper systems have some learning’s for us in the comprised age, mostly around data visibility. The best paper maintenance history system I remember seeing was in the Blast Furnace Department of the steelworks. In the very late 70’s the Blast Furnaces had a technical support agreement with one of the Japanese steel companies and they setup a history system based on the Japanese approach. The system required each maintenance supervisor who looked after a major area of plant to keep two history books. The books used were high quality hard cover with indexed replaceable pages and were almost A3 sized. Book 1 was for recording equipment related history information split into logical equipment areas similar to the rolling mill card history of events but with lots more room to store photos, diagrams, etc. Book 2 was more about the technical information and the knowledge required to maintain the equipment and run the maintenance section. The maintenance supervisor for each section took a lot of pride in the quality of the information that went into these books. The books were always available on the book shelf of the supervisor for anyone to inspect and use. I think one of the inputs into the performance assessment of the maintenance supervisors was the quality of what they were recording in their two books.

One discussion point in the ‘Knowledge Management in Maintenance’ meeting was on succession planning techniques and one approach called a ‘Knowledge Handover Sheet’, which was being effectively used by Defence and other organisations. This sheet is used to pass on information about the knowledge required to carry out a role. The current person performing a role produces the sheet, which is regularly reviewed as a part of their KPI’s. The sheet is not used to define what tasks to do in a job but the knowledge you need to do the job and where to get it or who to get it from. It specified the knowledge flow process for the role, including how the role supports others with knowledge. The maintenance history Book 2 as used in the Blast Furnaces in the 70’s and 80’s used exactly the same logic as the ‘Knowledge Handover Sheet’ but with a lot more detail. I think the Book 2 concept is a fantastic tool to help with getting new maintenance people up to speed with the complexities of a plant they have to learn to look after.

I have two ideas that might help improve maintenance history systems by simulating the paper based Book 1 and Book 2 systems and making information more visible. The first is to once every month print work-order reports with all relevant history information, broken up into standardised plant area groupings. These printouts should then be posted on a pin board in an accessible area to make the history information recorded in the last month more visible. This would also create a deadline to ‘Get the History Data Right’. Anyone who has made crappy half-assed history comments will have their work on display for a month. Good history comments will standout and act as something for people to emulate. At the end of the month those sheets could then be added to an A4 Book 1. The Book 2 could be a MS Word document broken up into different chapters for the plant area or technical groupings, with any updates made printed out and displayed. There are various systems to convert A4 pages into books. It is then easy for anyone to review the history. I suggest making history information more visible such as by using history pin boards will significantly improve the quantity, quality and consistency of your maintenance history.

My second idea might seem a little radical, especially to your CMMS experts. Computer Maintenance Management Systems (CMMS) are mostly structured as Work Management Systems. The information you need to best manage your maintenance work is significantly different to the information you want to see as maintenance history. Work management is about procedures, specifications, coordination, resources, etc and maintenance history is about what was done, why it was done, problems experienced, what was found, operational effects, possible causes and most importantly what was learnt. It is difficult to provide a structure for recording maintenance history and so history comments are often just made into long text fields in the CMMS.

The words you are currently reading (unless it has been printed out or copied) are recorded in a software system called MediaWiki, which uses the same software used by Wikipedia – now the worlds most widely used Encyclopaedia. The wonder of Wikipedia is that it has been created online by 100,000’s of different people like you and me who know something about a specific topic. It would be extremely easy to create a Book 1 and a Book 2 in MediaWiki and the idea of everyone inputting their little bit on higher level knowledge and wisdom into a common information storage system is a great model for maintenance. A CMMS should still be the system for storing all the detailed data but it would be nice to have a system that creates an overview of our maintenance history and knowledge that can be created and maintained by a maintenance team. The most relevant history information in your CMMS system could be easily cut and pasted into a MediaWiki system and additional overview information added. There can be little argument about the cost of setting up a MediaWiki system (or the large range of competing systems) as it costs nothing to download from the web, with lots of instructions available for a more computer literate person to set it up on your organisations computer network.

A MediaWiki History System has the advantage of having an easy to use word search system and is accessible to anyone with a computer that is linked to your network. MediaWiki is also very powerful in being able to add pictures, documents and links to other web information. There is probably an advantage to initially having updates printed out every month to create a data deadline and give extra visibility but once the system develops some momentum and ownership, that will probably not be necessary. There should be enough people using the system to spot any gaps, inconsistencies, etc. To set up a MediaWiki Book 1 you would select the level at which you want to record your history and create a hierarchy of pages down to that level. Each page would have the Plant Index (Functional Location) code and description text as the title with links to lower level Plant Indexes at the top of the page. Finding a page would be either by just following the page links or by typing the Plant Index code into the search field that is on all pages.

I have been using MediaWiki for a bit over three years and like any computer software there is a learning curve but there are 4 easy learning steps to reach my existing level of user skill.

1. Using the MediaWiki system, which only requires being able to use a mouse, follow links and the ability to read (covers everyone that has used a computer for something)
2. Adding or changing text, which only involves clicking on the edit tab at the top of the page, typing and then most importantly going to the bottom of the page and pressing ‘Save’.
3. Formatting pages, which is mostly using the simple menu bar at the top of the edit mode screen.
4. Inputting documents and pictures and adding them into pages.

Learning to use MediaWiki is mostly by looking at the formatting others have used and copying what they have done.

The above ideas are only two of the many possible approaches you could take to improve your maintenance history. My suggestion is to find ways to make your maintenance history more visible and accessible to the people in your plant.

Article written by

Peter Todd

Facilitator for the NSW Industrial Maintenance Roundtable