SOFTWARE FAILURE MODE AND EFFECT ANALYSIS (SFMEA) - Software Design for Six-Sigma: A Roadmap for Excellence - page 414

TABLE 16.2 Major Software Failure Mode and Effect Analysis Research

Contributions

Year Reference Contribution

Goddard, P.L, Validating the
safety of embedded real-time
control systems using
FMEA, Proceedings Annual
Reliability and
Maintainability Symposium,
pp. 227-230, 1993.

Goddard:
Described the use of software FMEA at Hughes
Aircraft. Goddard noted that performing the
software FMEA as early as possible allows
early identification of potential failure modes.
Pointed out that a static technique like FMEA
cannot fully assess the dynamics of control
loops.

1993

Fenelon, P. & McDermid,
J.A., An Integrated Tool set
for software safety analysis,
The Journal of Systems and
Software, 21, pp. 279-290,
1993.

Fenelon and McDermid:
Pointed out that FMEA is highly labor intensive
and relies on the experience of the analysts.

Banerjee, N., Utilization of
FMEA concept in software
lifecycle management.
Proceedings of Conference
on Software Quality
Management, pp. 219-230,
1995.

Banerjee:
Provided an insightful look at how teams
should use FMEA in software development.
FMEA requires teamwork and the pooled
knowledge of all team members. Many
potential failure modes are common to a class
of software projects.
Pointed out that the corresponding
recommended actions are also common. Good
learning mechanisms in a project team or in an
organization greatly increase the effectiveness
of FMEA. FMEA can improve software quality
by identifying potential failure modes.
Stated that FMEA can improve productivity
through its prioritization of recommended
actions.

1995

Luke, S.R., Failure mode,
effects and criticality
analysis (FMECA) for
software. 5th Fleet
Maintenance Symposium,
Virginia Beach, VA (USA),
24-25 Oct 1995, pp.
731-735, 1995.

Luke:
Discussed the use of FMEA for software. He
pointed out that early identification of potential
failure modes is an excellent practice in
software development because it helps in the
design of tests to check for the presence of
failure modes. In FMEA, a software failure may
have effects on the current module, on higher
level modules, and on the system as a whole.
Suggested that a proxy such as historical failure
rate be substituted for occurrence.