Reliability engineers have known for decades that realistic data must ultimately come from one source, quality field failure data for devices operating in a similar application.
Perhaps that is why IEC 61511:2016, Clause 11.9.3 states
“The reliability data used when quantifying the effect of random failures shall be credible, traceable, documented, justified and shall be based on field feedback from similar devices used in a similar operating environment.”
This strong language reinforces what reliability engineers have long understood.
The goal of high quality field failure data has been somewhat idealistic but good progress is being made. By combining the data from various sources a very realistic picture is emerging. At exida we combine data from manufacturers, from operating plants, and from OREDA, perhaps the best source of well analyzed data in the world. Constant comparison of this data while checking the data gathering processes has shown consistency. This data is then used to calibrate the FMEDA predictive failure rate method used for specific devices. It is a lot of work but that is how FMEDA predictions based on field failure data can be credible, traceable, documented, and justified.
Any failure rates published by any organization should explain the derivation and document how they meet the requirements of IEC61511:2016. If you think you might have data that seems unrealistic, check it against www.SILSafeData.com.