Building BIOS Architecture

Fault Management controls detection, classification, containment, response, clearance, and documentation of abnormal conditions.

A Fault Record shall contain:

Fault ID;

affected entity;

fault category;

detection source;

first occurrence;

most recent occurrence;

severity;

persistence;

affected capabilities;

dependencies affected;

immediate response;

permitted operating state;

required inspection or repair;

clearance authority;

evidence.

Fault categories shall include:

identity fault;

configuration fault;

Interface fault;

component fault;

resource fault;

communication fault;

control fault;

sensor fault;

software fault;

security fault;

safety fault;

environmental fault;

lifecycle fault.

Fault response may include:

record only;

warning;

capability reduction;

service restart;

transition to standby;

selective isolation;

controlled shutdown;

activation of redundancy;

emergency action;

Repeated faults shall not be cleared automatically in a manner that hides recurrence. Fault counters, duration, history, and environmental context shall be retained.

Safety-critical faults shall remain latched until an approved clearance process confirms that the cause has been removed and required functionality restored.

A fault shall be closed only when:

corrective action is complete;

verification has succeeded;

affected systems are recommissioned where required;

resulting state is recorded;

responsible authority approves closure.