Resist the temptation to Use MTBF
Sure, it would be easy to use MTBF for a system reliability goal. Your organization has regularly used MTBF. Your customers are asking for MTBF. The competition all use MTBF. Even your vendors supply only MTBF.
Yet, you know it’s not the best metric to use. It’s not accurate, it’s not useful, and you rather use something else.
As you know, reliability is the probability that a system will function in a specific environment for a specified duration. We regularly define the system functions, plus attempt to understand the transportation, storage, installation, and operating environments.
It’s the probability and duration that we miss. Using just MTBF we have a poor metric for only the failure rate (probability element). We’re missing the duration.
So, state the duration and the probability directly. Say, 98% will survive 2 years. Also, we want to have 99.9% survive the first month (early life failures should be minimized). And, lets add a stretch goal, say 90% should survive 10 years.
The three couplets do not have to line up on any distribution or curve. The three points directly describe our goal in a way that is easy to understand and use by our vendors, design team, and customers.
Connect Reliability Goal to Business Objectives
The best part is we can, and should, connect the reliability goals to business objectives.
Of course, the product reliability achieved is a balance between what is possible, cost, expectations, and business objectives. For example, if we want to be the most reliable product offering compared to our competitors, that is a business decision. In practice it will mean understanding the reliability performance both short and long term of our competition to be able to set appropriate goals.
Another objective is profit. Warranty or service costs for field failures directly reduces profit. So, finding the failure rates that allow suitable profit given a selling price, directly connects the reliability goals to the business objective of profit.
Customer satisfaction is often related to the system’s reliable performance. You’re marketing folks may have studied the relationship between system failure rate and brand loyalty . Loyalty impacts word of mouth marketing, repurchasing, etc. We may be able to justify improving reliability over local objections because the improvement will provide significant return on investment. Again a business decision.
In short, make it easy. State reliability directly as a probability of surviving over a duration. I recommend three such couplets to connect with key markers for your team. Early life, warranty and design life – or what ever makes sense for your program and market.
Don (PLC Training) says
Yes, I am always the first to say, just because everyone does it that way, and/or it is ‘standard practice’, does not always mean it is the best way.
Dave Deroian says
Enjoyed your papers/tutorials at RAMS this year. I’m working on changing my company’s culture to specifying a reliability goal rather than MTBF. Thanks for all of your great work on this most misunderstood and misapplied metric.
Fred Schenkelberg says
Glad our paths crossed. Good luck with our efforts an the NoMTBF front and do keep me posted.
Thanks for the kind words.