Reliability Blog

Reliability in the automotive electronic systems

The automotive electronics technology based on cameras, radars, and LIDARs, offers automakers the opportunity to develop autonomous vehicle systems.

All these technologies require a specific function to minimize defects, improve the device life cycle, and allow companies to provide a reliable product or system into the market. Such a process involves an analysis of reliability for each element in the device expressed as failure rates (F.R.). This examination should be carried out during the development of the product to qualify components or systems and to reduce potential failures.


MTBF metrics

The Meantime between failure (MTBF) metrics refers to the average time that a reparable device functions before failing. It is a measure of how reliable a product or a component is. The required MTBF can be used as a quantitative target when designing a new product. 

Reliability metrics during the sales process

Before signing the purchase offer, more automotive manufacturers ask for the MTBF figure of the product they need. In today's business, this requirement has become the primary concern for electronic system providers. Therefore, the anticipation of the MTBF prediction will improve product reliability, will satisfy customer requirements, and facilitate business negotiation.


Reliability Analysis Part 1

The term reliability refers to the probability that a device will function properly and without failures for a specified period of time. Reliability is a measure of failures over time. It has an effect on repair and maintenance costs. MTBF (Mean Time Between Failures) is the most used analysis to evaluate a part or a device. The higher the MTBF number is, the higher the reliability of the product.

When should the MTBF prediction be made?

The reliability prediction may be performed during product development to ensure dependability and performance. At this stage, there could be an impact on the design determined on the changes that will be made when comparing the failure rate between the existing components and the proposed components. The MTBF prediction can be used to maintain an acceptable level of reliability of the parts or product under different environmental conditions.

Why should the MTBF prediction be made?

MTBF nan.png
  • Make the product more reliable;

  • Good engineering practices;

  • Establish the presence of a safety margin;

  • Improve the company's reputation;

  • Improves system life;

  • Customer's request;

  • Reduce costs;

  • Reduce product returns.

Reliability prediction methods

Several prediction methods over time have been developed to determine reliability. The two most used analysis when compiling reliability data are: MIL-HDBK-217,F2, “Reliability Prediction of Electronic Equipment” and Telcordia SR-332, “Reliability Prediction Procedure for Electronic Equipment”.

Several prediction methods over time have been developed to determine reliability. The two most used analysis when compiling reliability data are: MIL-HDBK-217,F2, “Reliability Prediction of Electronic Equipment” and Telcordia SR-332, “Reliability Prediction Procedure for Electronic Equipment”.


The MIL-HDBK-217 was developed by the US Department of Defense with the assistance of the military departments, federal agencies, and industry. It is the most widely known and used reliability prediction. It is used by both commercial companies and the defense industry and is accepted and known world-wide. It contains failure rate models for numerous electronic components such as integrated circuits, transistors, diodes, resistors, capacitors, relays, switches, connectors and more. There are two ways to predict reliability under MIL-HDBK-217: Parts Count Analysis Prediction and Parts Stress Analysis Prediction.

The Part Count prediction is used to predict the reliability of a product at the beginning of the development stage to obtain an estimate of reliability. A failure rate is calculated by counting similar components and grouping them into different types of parts. The number of components in each group is then multiplied by a failure rate and a quality factor. Finally, the failure rates for all the different groups of parts are added together for the final failure rate report. Part count assumes that all parts are in series.

The Part Stress prediction is typically used later in the product development stage when the product approaches to production. It is similar to the number of pieces in the way the failure rates are added together. To allocate the proper stress levels to each component, the failure rate of each piece is calculated based on the specific constraint levels the component is subjected to (voltage, environment, etc.). Due to the level of analysis required, this method takes a long time compared to the Part Count prediction.

Telcordia SR-332 was developed by Bell Laboratories for the reliability prediction of commercial electronic components. The Telcordia Reliability Prediction Procedure is now used in both the telecommunications and commercial industry. It provides models for predicting the failure rates of units and devices during the first year of operation. It also contains failure rate models for numerous electronic components such as integrated circuits, transistors, diodes, resistors, capacitors, relays, switches, connectors, fiber optic transceivers, hard drives, ferrite beads and more.

Reliability Analysis Part 2

Failure rate

Failure rates are the basis of reliability. The failure rate is noted λ (Lambda). This is the frequency in time at which the system fails. Failure rates generally increase in systems because of use and age. The failure rate value is expressed as failures per million hours (fpmh or 10^6 hours). It can also be expressed as failures per billion hours (FIT or failures in time or 10^9 hours).


MTBF (Mean Time Between Failures) is the most used analysis to evaluate a part or a device. The higher the MTBF number is, the higher the reliability of the product. MTBF is the average operating time of a product between two outages without including repair times, when the item is repaired and put back into service. It is a term of reliability used to provide the number of failures per million hours of a repairable product or system.

MTTF ''Mean Time to Failure''. This is the average time for the first failure of a component or product. It is a basic measure of reliability for non-repairable products and the expected average time until the first failure of a component or part. The calculation of MTBF is used for products that can be repaired and reused, while MTTF is used for non-repairable products.

MTTD and MTTR Mean Time to Detect (MTTD) and Mean Time to Repair (MTTR) are methods used to describe the time it takes to find a problem and the time it takes to get the product back into service. It is very important to proceed as soon as possible. Restore is composed of several steps: identify failure, delete failure and test. The lower the MTTR, the shorter the downtime and the more the product is available.


MTBF a common misconception

A misconception of MTBF is that it is equivalent to the number of hours of operation before a product fails, or takes the MTBF figure as the life of the product. It is not uncommon to see an MTBF figure of 1 million hours or more and to think that the product could operate without failure for over 100 years.

These numbers are, on occasion, so high because they are based on the failure rate of the product during its useful life. Another misconception is to think that it will continue to fail at this rate indefinitely. In real life, the extended mode of use of the product may limit its life much earlier than its MTBF figure based on variables such as the environment the product has been installed. It is quite possible to have a product with an extremely high MTBF but an average or more realistic expected service life. So, there should be no relationship between the life of a product and its failure rate.

Reliability Test Plan

The reliability test plan is a technical document that describes systematic procedures for testing a specific device or machine. It contains particular information such as product design and assembly and is used to formulate the accelerated failures test.

Why do accelerated life tests?

In the reliability estimation, an accelerated life analysis can be shaped to validate product functions and potential failures. Various methods can be used involving product quantification failures under typical conditions of use. It can be a good idea to conduct accelerated life tests at development levels before the product accesses the market.

ALT and HALT Tests

Accelerated Life Test and Highly Accelerated Life Tests are the most common tests to accelerate product failures. Both methods outline the process of exposing a product to severe conditions beyond normal environmental requirements, to identify the product endurance limit and failures in a short time.



When the ALT completed, it is reasonable to perform an MTBF estimation with the results from the test. With the bill of materials and specific information, we can estimate the MTBF using the appropriate standard and environment.

RMA - Reliability, Maintainability, and Availability.

Reliability (R) is the probability that a component or system will perform its intended function without failure during a given period when it is under a specific operating environment.
Maintainability (M) is the possibility that a defective item is replaced or repaired within a given time. Maintainability is an integrated parameter that must be done when designing the equipment.
Availability (A) is the likelihood that a repairable system will perform its intended function for a specified time when used in an assigned manner. Therefore, availability is a function of reliability and maintainability.

If "R" is satisfactory, then no need for "M" and "A"; if "R" is not enough, then "R" and "M" are needed.
Availability is simply the total uptime/total time (uptime+downtime). This is basically the hours of operation of a device.

There are two rules to estimate reliability
By calculating the Mean Time Between Failures, which is the Total Time in Operation/Number of Failures.
By computing the Failure Rate. The failure rate is how often a device is failing in time. It is the number of failures/total time in operation.

A high failure rate is indicative of low MTBF
The main achievement of reliability is to assess the success of a product to establish potential
failures, as early in the product's lifecycle, and then take appropriate action to make the necessary improvements. It is never too late to enhance the reliability of a product. Remember: modifications are less costly at the start of design than later when the product is manufactured.

Difference Between Quality and Reliability
In manufacturing, quality is defined as a measure of excellence of being free from defects, deficiencies, and significant variations.
A product may have a reliable design, but its reliability is not satisfactory when it is operational. Perhaps the reason is the result of a weak manufacturing process. The failure did not occur due to incorrect design, but its quality is unacceptable due to the assembly method.

Installation mère

Reliability of electronic components

A point of concern for engineers when developing a new product is the failure rate figure of components. In most datasheets, such information is not always available. Thus, the designers turn to specialists to estimate the value and get a review of the MTBF before launching the new product.


It is appropriate to distinguish two failure conditions into components:

  1. Defective components are faulty from the moment they are born.

  2. Failed components are manufactured according to their rules, but they deteriorate itself over time or by exposure to high stress.


Three factors may cause failures:

(1) Hidden internal factors that settled in the piece from the beginning;

(2) External stress caused by the environment, such as temperature and humidity;

(3) Physical degradation over time.


Periods of failures

To explain the states of failure of a component, we use a graph named the bathtub curve, which shows three correlation periods of failures over time. They are infant mortality, accidental failure, and wear-out period.


Infant mortality: failures occur abruptly after the product start-up and gradually weaken over time. The principal cause is hidden internal defects.

Accidental failure: during this period, failures occur by accident at a constant rate and are not time-related, for example, lightning or electromagnetic forces.


Wear-out: at this point, the failure rate increases gradually over time. It is due to the wear of the product when it enters the end of life.


Overview of reliability tests for electronic components

LED Life Expectancy

LEDs are reliable light sources, but their performance decreases gradually over time. In lighting applications, where the light level must be maintained, the lifespan of LEDs remains a decisive point.

Environment, vibrations, shocks, power supply, and driver performances are factors that affect life expectancy.

Lumen maintenance factor:

The light from a source is referred to as the lamp lumen maintenance factor.

People usually do not observe a gradual reduction in LED light until it drops 30% of its initial brightness. LED life expectancy is shown then, as an L70 format that it equates to a reduction of 30% in the light output.

New technologies provide LEDs for lighting applications increasing the L70 figure from 50,000 to 100,000 hours.

Life expectancy prediction:

After submitting samples to environmental tests, the Weibull distribution method is used to estimate LED's life expectancy and expected behavior.

LED life expectancy

© 2021 by Les services-conseils Conformitec Inc.  All rights reserved. Tous droits reservés.