You may already know my position on MTBF. If not, in short, do not use MTBF at all, ever, in any form. So what should we use instead.
Accendo Reliability Webinar Series The MTBF Sucks. Here's Why
May 28, 2025
Explore the pitfalls of the Mean Time Between Failures (MTBF) metric, revealing how larger MTBFs can still lead to more failures. Discover the unpredictable reliability of mechanical components like ball bearings, influenced by microscopic factors. Dive into the failure patterns of small satellites, shedding light on the challenges of infant mortality and how statistics can mislead interpretations. The discussion critiques traditional reliability metrics, advocating for a shift towards data-driven assessment methods that enhance product reliability.
AI Snips
Chapters
Transcript
Episode notes
Ball Bearing Failure Is Random
- Failure is a random process causing seemingly identical ball bearings to fail at different times.
- Microscopic differences make them indistinguishable but not truly identical to reliability engineers.
Failure Patterns Revealed Visually
- The probability density curve shows likelihood of failures over time, revealing failure patterns.
- More data smooths the curve, making failure patterns clearer for reliability analysis.
Understanding Mean Time to Failure
- The mean of a failure time distribution is the balance point of its probability density curve.
- It represents the mean time to failure but may not correspond to the most likely failure time.
