Latent-fault metric

The Latent-Fault-Metric (LFM) is defined as follows ISO 26262, Part 5, Annex C ... 

This metric reflects the robustness of the item to latent faults either by coverage of faults in safety mechanisms or by the driver recognizing that the fault exists before the violation of the safety goal, or by design (primarily safe faults). A high latent-fault metric implies that the proportion of latent faults in the hardware is low. 

C3.1 This requirement applies to ASIL (B), (C), and D of the safety goal. The definition given by the following equation shall be used when calculating the latent-fault metric ... 

Constraining architectural metrics, to cover faults, are defined as single point fault metrics and latent fault metrics, their definition being ... 

The latent fault metrics can be seen as the capacity of a system to control latent or hidden failures. 

The same principle can be applied to the latent fault metric. It is possible to allocate the various metrics to the different components of a system. 

The random hardware failures addressed by these metrics are limited to some of the item s safety-related electrical and electronic hardware parts, namely those that can significantly contribute to the violation or the achievement of the safety goal, and to the single-point, residual and latent faults of those parts. For electromechanical hardware parts, only the electrical failure modes and failure rates are considered. 

ISO 26262 only recommends the architecture metrics for ASIL B and requires the latent failure metric only for ASIL D functions. If the quantitative approach from lEC 61508, often called FMEDA(forms maMy in MS-Excel and based on part-count method with the fault-distribution as described by Alexandre Birolini, see also ISO 26262, Part 5, Aimex E), is recommended or a more deductive approach should be apphed could depend on the apphcation. A deductive approach could provide also insights related to systematic faults and non-functional failure, by a pure part-could approach, this could be questioned. 

The architecture metrics (single-point fault metrics (SPFM) and latent failure metric (LFM)) would result from the safety architecture and would be a mathematical function of the failure rates (MFxx) and the implemented safety mechanisms (DCxx). 

Therefore, it is important to understand that the application of the standard will define the constraints for the processing of random hardware faults (with the metric in terms of probabilistic targets and coverage rate of dangerous or latent failures) and the proeessing of systematic faults (with rigorous and adapted design and validation processes). 

---