A while back, Reliabilityweb.com conducted a survey about Computerized Maintenance Information Management Systems (CMMS/EAM) with overt 600 companies and found that the majority of software implementations failed to meet expectations or generate the expected Return on Investment.

I have attached a copy of the CMMS/EAM Best Practices Report with this post.

I would appreciate some comments as to:

1) Why the success rate is so low

2) What steps should be taken to improve the success rate

Any other comments are welcome as well.

Thanks,
Terry O

cmms_best_practices.pdf (103 Kb, 78 downloads) CMMS/EAM Best Practices Report

Actually your report already answers your questions. Anyhow, to reinforce some of the points, here are my views:

1) The maint organization should have a deep knowledge of maint mgmt areas. Without this knowledge, doing it with computer will not do any good because the level of knowledge remains the same. In fact, doing it with CMMS now may make it more difficult to do because of theseemingly extra, unnecessary work to do it using computer. Eg does the organization track MTBF? If tracking MTBF is not important to them, then it's quite useless & lack of motivation to use CMMS. Thus, some standards for maint mgmt should exist beforehand.

2) The CMMS should come with a standard set of KPIS with a flexibility for user-defined analysis. Without visible & immediate KPI results from CMMS, it appears useless & meaningless to use a CMMS. Don't forget the KPis should comprise lagging/reactive & leading/proactive KPIs. Again does the organization appreciate KPIs, not to say lagging & leading ones? If no, useless to use CMMS because to carry out maint works, they can just use hardcopy forms. However, if KPIs are important & in order to ease/speed up data analysis, there will be motivation to capture data in softcopy in CMMS and avoid manual tracking.

3) Eqpt Master data must be populated in CMMS before or just prior to start of using CMMS so that the users can see the benefits of using CMMS eg automatic triggering of WOs for PM with its tasklist, materials & tools required which have been planned in advanced.

Three main areas to highlight:

3a) Equipment master data must be populated with a proper criticality, class, type, manufacturer, model no, age, characteristics for data sheet, catalog profile (or failure codes), bill of materials, etc. Some of these already have standards eg ISO14224. Spare list for eqpt is absolutely essential for WO use. Asset Hierarchy is another pre-requisite for meaningful data analysis subsequently.

3b)Prev maint master data include tasklist/job plan. To set up this, a lot of works need to be done. Using a team is a good approach to get it done fast without distracting existing manpower but may lack buyin from discipline maint engineers/users after project handover. So some involvement of users/customers should be ensured during the project execution phase. Finally KPIs for eqpt master data should be available for monitoring completeness & improvement of this area.

3c) Material master data must be available to enable easy & readiness of CMMS use. Again use of project team can be considered to code materials with proper criticality, min/max stocks with different modes of replenishment, nonstock, capitalised materials etc. Maint contracts where can be identified in advance should be set up in the CMMS for easy use by WO users. KPIs for material masters & contract masters should be available for monitoring this area.

4) Enforcement of the use of CMMS should be exerted in the CMMS itself (full-proof system) or intervention by authorities outside. Eg a) minimum mandatory data entries into work oders for proper data analysis and KPI reporting. b) no work permit released by operations without a proper work order by maint and work permit request is done via the CMMS itself. c) no works can proceed without a work notification.

5) All depts & disciplines shall use CMMS ie not only maint but also operatiods for operator-based PM, inspection, projects on existing assets etc or not only mechanical, E&I but also civil, marine etc. This to capture the total picture/amount of works in the facilities for lifecycle cost & reliability monitoring purposes. All eqpt shall be also included in the CMMS ie not only main eqpt but also piping loops, instrument loops, electrical loops etc. Sets of KPIs should be available to monitor each area's performance.

Hope I add something by the above comment. TQ for reading.

This message has been edited. Last edited by: Josh,

I would agree that proper policing of the newly implemented system is essential. As Josh says one way of ensuring that all unplanned maintenance information is captured is to reject any work requests that are not submitted properly through the system and as defined. (Bear in mind that if someone reports smoke in the explosives cabinet you may have to make an exception

I would also agree that populating the system with most of the meaningful data before launch is very important. If you don't do this beforehand the chances are that it just will not get done. I also recommend that where possible the system should be piloted in a small department before plant wide rollout. Any errors that have been made can be rectified at this stage because it is sometimes too late to change things after the system is fully in place.

However, I think that by far the main reason for CMMS system failure is that the full cast of characters are not often invited to the rehearsals. Quite often the decision to implement a CMMS is made by one or two people who then go on to put things in place without first bringing the users (both maintenace and non-maintenance personnel) on board. When implementation starts the users are turned off immediately by what they may see as just another new fangled tool that makes their life more difficult. Proper planning is essential. Meetings should be held and opinions sought long before any software is purchased. If the participants are involved in the process from day one they are far more likely to embrace it when it actually happens.

Terrence, I may have asked you this before but how many companies contibuted to your survey and were they from a cross section of industries and sizes? I think that the survey results may be different if applied exclusively to companies of a similar size.

[edited for typo]

This message has been edited. Last edited by: Bryan Weir,

even for that explosive cabinet, a work request should be submitted & may be preceded by a phone call. Raising a work request shouldn't take more than 1 minute & don't use the PC next to smoking explosive cabinet though. TQ

Bryan,

Over 650 companies of all sizes from around the world.

I have attached a pdf of the power point as well which explains the demographics of the survey respondents a little more.

We wanted to take a high level snapshot of CMMS use on a global basis - then depending on the findings - decide to drill down deeper in a subsequent survey.

In my opinion, after we crunched the survey results - we reached the conclusion that we need to start to talking about some of the soft side issues (that you mention above) like planning, participation, people, culture as well as the underlying maintenance process. This was reinforced at CMMS-2005 over and over.

It cannot be stated enough that a CMMS will automate a maintenance process - if that maintenance process includes dysfunction - the CMMS will automate that dysfunction and make it more effective and effective dysfunction is not what anyone is aiming for.

My latest area of interest is integrated Enterprise Asset Management (SAP, Oracle, IFS etc...) versus Best of Breed (MRO, Datastream, Indus etc...)

Questions include: Once basic functionality is covered - do the bells, buzzers and whistles that best of breed sells - get used? What are all the advantages for finance integration? Can finance and maintenance work togther successfully? I would welcome suggestions for more questions and topic areas to research.

Josh - great posts - I am becoming a big fan - please keep it going.

Bryan - thanks for adding to this project.

Terry O

CMMS_Best_Practices.pdf (312 Kb, 55 downloads) CMMS Best Practice slides with demographics