MTBF confirmation testing is helpful to assess the reliability of products, components, or tools, especially those which are costly and difficult to replace, by assuring that they don’t stop working too early in their lives.
What is mean time between failures?
As the name of the MTBF test suggests, this is the average time between failures of repairable products or components during regular system operation.
It is usually measured in hours and will vary depending on the product/component’s usage and type. For instance, very cheap or even disposable products will require a low MTBF, whereas parts used in automobile or aerospace applications would need a very high one.
MTBF can usually be calculated by dividing the number of operational hours by the number of failures.
Related 👉 MTBF wiki
An example of MTBF confirmation testing we conducted
Here you can see an MTBF confirmation plan we applied on 25 samples of electronic devices (no customer info given for privacy). The target MTBF on samples was 70,000 hours, a very high number. So we had to test the samples for just over 3 weeks.
As with most reliability tests, testing time can be adjusted per your needs.
Here is the equation used to calculate the testing hours required for the above test:
How to use an MTBF test’s data?
On the production side, your MTBF confirmation test could be a part of your regular reliability testing schedule.
When it comes to using the product/part/tool that requires a long time between failures, this information helps you implement an adequate preventive maintenance plan. For example, you might inspect that part more often and you might make sure you have a replacement ready as you get closer to the MTBF.
Also, if such an item keeps failing too early and any measures you take to try to increase the MTBF are unsuccessful, you know that more drastic measures may be required such as redesigning the product to use a different material/component type in order to improve reliability.
Other reliability testing
Aside from an MTBF test, you may also consider this testing for your items: