Fault tree analysis is performed on software to determine the
areas in the product which could cause a potential failure and to determine
the risk and severity of any such potential failure. The timing of this
analysis is important and should start during the design phase to identify
top-level hazards. The analysis can continue through code development and
testing to identify paths for testing and verify that safety related hazards
will not occur.
The steps for performing a software fault tree are:
Determine failure modes for software starting from top
level product and working downward.
||Make these failure modes the top nodes of
the fault tree. Assume that these failure modes have already occurred
and refer to them as events.|
||When tree completed for top level failure
modes, determine risk and severity for each of the bottom nodes on the