The End of MTBF Debating
Endless debate on merits of MTBF is meaningless as more than one has told me. My neighbor has a bumper sticker that says Endless War with the ‘less’ crossed out and written above it ‘this’. Instead of endless debate, how about we just end this now. Stop using MTBF and all related grand averages. Use the statistics, distributions and knowledge you have to provide accurate estimates and summaries.
My favorite argument has been MTBF has been in use for 50 years, therefore it must be useful. Buggy whips were useful at one time too and are relatively hard to find outside the house carriage enthusiasts community. At one time MTBF may have been useful, we do not live in those times. An irony in my world is as I teach a CRE refresher course, many of the sample problems use MTBF. So, in order to pass the exam and become certified one has to master the use of MTBF. In my graduate level class as well as the working professional taking the CRE refresher, the confusion around MTBF is amazing. If we, the reliability professionals continue to have trouble understanding MTBF it is difficult for us to teach and guide others. Another recent argument to use MTBF was the customer expects we use it. I look at that as a teaching moment. There are plenty of other arguments on both sides summarized in
part 1 and part 2 posts. Another argument has been it is a too big of a change and just won’t happen. If we don’t start, then I’m afraid this argument is right. So start.
Awareness and Action Use failure rate along with time period over which it applied. Use probability of success and a duration. Use clear and easy to understand language and check that all parties understand the terms and metrics. This applies outside of reliability, too. There are about 250 visits to the MTBF site a week (higher numbers when someone else posts than me). There are about 400 members of the NoMTBF Linkedin group. The paper by Author Rowland “
MTBF: What’s it good for?” posted to Slideshare 3 months ago has over 2,000 views. We are generating awareness and we’re being seen as opposing the use of MTBF. If you search for MTBF, the NoMTBF site is in the short list. This started as a dare borne from frustration. Now it is a mission. Make it a mission in your world too. Let me and others know of your success and challenges.
Barry Snider says
I stopped using MTBF some 20 years ago. I know I have influenced dozens to not use the metric. MTBF is not only useless in most applications, it is misleading and leads to some dangerous and costly conclusions. I like this post, JUST DON’T USE IT….PERIOD. As another note, the exam for CRMP certification also stresses the importance and the use of MTBF. Oh, brother, how stupid can these exam creators be?
Fred Schenkelberg says
Thanks for the comment and keep up the work – I agree that the use of MTBF in the CRE, CQE, and CRMP is pointless and damages our profession. Same for standards, although the IEC and IEEE teams are slowing working the use of MTBF out of the standards.