<<< Back to Index
This Month's Tip >>>

Bellcore/Telcordia Reliability Prediction in Lambda Predict

 

 

Standards based reliability prediction is a methodology for predicting the reliability of a product by using failure rate estimates published in globally recognized standards. It is usually conducted during the concept and design phase of product development, where test data are not yet available. With the release of Version 9, Lambda Predict now supports four reliability prediction standards: MIL-HDBK-217F, Bellcore/Telcordia (SR-332), NSWC-07 and FIDES. This article focuses on the Bellcore/Telcordia standard. We will present an overview of the four versions of the standard that are supported by the software, and then give an example that uses the latest version of the standard, Telcordia SSR-332 Issue 3.

Bellcore/Telcordia Standards

The Telcordia reliability prediction procedure has a long and distinguished history of use in the telecommunications and electronics industry. During the past fifteen years, four versions of the standard have been published:

Each new version modifies the previous edition based on new data. Telcordia Issue 1 made minor changes and updated the device failure rates published in Bellcore TR-332. In Telcordia Issue 2, several new devices were introduced and the generic device failure rates were updated. In addition, the standard deviation of the generic failure rate for each device was provided so calculations for the upper confidence bound of the failure rate were made possible. Telcordia Issue 3, the latest version of the standard, introduced several major changes to Issue 2, including:

With the release of Version 9, ReliaSoft’s Lambda Predict software allows you to choose any of these versions for your standards based reliability predictions.

Three Methods to Calculate Failure Rates

Depending on the amount of data available, the Bellcore/Telcordia standards provide three methods to predict failure rates:

In Lambda Predict, Methods II and III can be applied at both the device and unit levels; while Method I can be applied at the device, unit and/or system level.

Example

Super Electronic Inc. has redesigned one of its LED lighting controller products. Reliability engineer Amy is given the task of conducting the reliability prediction for the new prototype. The goal is to estimate the following metrics:

Amy will use ReliaSoft's Lambda Predict software with the Telcordia Issue 3 prediction standard to perform the analysis.

First, she studies the electronic circuit, and then groups the devices according to their functions: dimmer detection, dimmer control and pulse width modulation (PWM) generation.

Next, she uses Lambda Predict to build the system configuration. The following picture shows the resulting system hierarchy of the LED lighting controller system.

Lambda Predict System Hierarchy

Next, Amy defines the properties of each block and component in the system, using information she obtained from specification sheets. For example, the following picture shows the properties for the Dimmer Detection block. It shows that the Method I approach is used to calculate the failure rate of the unit and its devices (and because Method I assumes that no reliability data are available for the unit or for its devices, the Lab Test Data and Field Data fields become unavailable).

Dimmer Detection Block Properties

When Amy has entered all of the required inputs for the Telcordia prediction, the system hierarchy shows that the overall failure rate of the system is 139.0415 failures in a billion operating hours (FITs), the MTBF is 7,192,100 hours and the upper confidence bound for the failure rate is 156.6884 FITs, as shown next.

Lambda Predict System Hierarchy - Calculated

Amy knows that she could incorporate any available lab testing data and field data to modify the generic failure rates and make the prediction more accurate.

She checks the previous design of the product and finds that the daughter board, which contains the PWM generation assembly, is exactly the same assembly that was used in the previous version of the product. The warranty data of the previous design show that 3,000 products were used in the field for a period of one year, and 5 of the failures were due to a malfunctioning daughter board.

To incorporate this data into the analysis, Amy changes the method used for the PMW Generation block to Method III (a) - Unit. She then enters the following information under the Field Data heading: Field Time = 26,280,000 hours (i.e., 3,000 products * 8,760 hours per year) and Number of Failures = 5, as shown next.

PMW Generation Block Properties

When she has entered the field data, the system hierarchy shows that the predicted failure rate of the PWM generation block increased from 63.5237 FITs to 121.1820 FITs. The overall failure rate of the system also increased from 139.0415 FITs to 196.6999 FITs.

Lambda Predict System Hierarchy - Calculated

Conclusion

In this article, we briefly discussed the differences between the four different versions of the Bellcore/Telcordia standards. The latest version of the standard, Telcordia SSR-332 Issue 3, reflects the current electronic technologies and is therefore the recommended version to use when performing new Telcordia predictions.

The Bellcore/Telcordia prediction standard allows you to incorporate test data or field data into the prediction in order to obtain more accurate results. Depending on the available data, the modified failure rate could either be larger or smaller than the result obtained from a generic failure rate.

 
ReliaSoft