What is the difference between Random Hardware Failure and Systematic Failure?

This is 'The GSEG (Global Safety Experts Group)'.

 

Today I would like to briefly explain the clear definitions of Random Hardware Failure and Systematic Failure, which are somewhat confusing.

 

1. Systematic Failure
The definition from ISO 26262 for Systematic Failure is as follows.

"failure related in a deterministic way to a certain cause, that can only be eliminated by a change of the design or of the manufacturing process, operational procedures, documentation or other relevant factors".

In other words, Systemtatic Failure is not caused by the physical limitations of the components of the system being developed (example), but rather occurs in the activities of system development and production.

 

The general points that apply to Systematic Failure are as follows.

1) Systematic Failure is a failure resulting from the system development and production activity itself, and human error of personnel participating in development and production activities is regarded as the biggest cause of Systematic Failure.

2) Systematic Failure cannot be measured quantitatively, and the relative size of Systematic Failure can be estimated  only with a qualitative method.

In other words, it can be seen that the system function that satisfies the requirements assigned to ASIL D has relatively

smaller Systematic Failure compared to the system function assigned ASIL A.

3) Systematic Failure exists in all levels of components where development and production activities occur, regardless of the level of system components.
(e.g. body controller, sensor, ASIC, uC, SW component).

4) Systematic Failure is a potential existence of a specific group in the process of development and the degree to which Systematic Faliure can be mitigated is directly related to the development capacity for a specific group.

5) It is judged that Systematic Failure can be mitigated or lowered by going through all activities of system or component development/production from more diverse angles and from more diverse viewpoints and methods.

Example: Wrong setting of parameters related to system function, Wrong setting of cut-off frequency of HW filter, Wrong setting of SW configuration for HW default status.

 

 

Do you need support or help for your project regarding functional safety?
Please let us sit down together with you and help you out in the end.

We will be there to solve your any topics on functional safety.
e-mail address: tofusiexpertask@gmail.com.

Please note that the cost of this work will be determined by each individual's decision in the form of a donation

2022.10.13 - [ISO 26262 engineering (Eng)] - Do you need support or help for your WPs?

Do you need support or help for your WPs?

 

2. Random Hardware Failure

The dictionary definition from ISO 26262 for Random Hardware Failure is as follows.


"failure that can occur unpredictably during the lifetime of a hardware element and that follows a probability distribution".


In other words, Random Hardware Failure is a term that is limited to HW parts among system components.

It means that is a failure caused by physical limitations of HW parts that may occur at any time during the period of use of HW part and the related failure rate can be reliably predicted based on specific conditions.


SN 29500 (Siemens norm) or IEC/TR 62380 is generally used for the calculation of Random Hardware Failure, and in  the case of IEC/TR 62380, it is deleted from the source for calculating Random Hardware Failure in the 2nd ISO 26262 part 5 (clause 8.4.3 ), but it is still considered one of the valid ways to calculate Random Hardware Failure.


Examples of Random Hardware Failure include 'capacitor short circuit, transistor instantaneous state change, and  op-amp input-to-output ratio error'.