INTRODUCTION TO SOFTWARE DESIGN FOR SIX SIGMA (DFSS) - Software Design for Six-Sigma: A Roadmap for Excellence - page 203

Availability

Stimuli Parameters Responses

Hardware Redundancy

Availability

Exact/Analytic
Degree
Failure Rate
Repair Rate
Failure Detect Time
Failure Detect Accuracy

Source Type

Reliability

Value

Hardware
fault

Levels of service
Mean time
to failure

Timing
Stopping

Software
fault

Software Redundancy

Exact/Analytic
Degree
Failure Rate
Repair Rate
Failure Detect Time
Failure Detect Accuracy

Voting
Retry

Failover

FIGURE 8.A.9 ATAM—availability characterization.

8.A.5 DIDOVM PHASE: OPTIMIZE

Optimizing the design typically involves one or more of the following:

Statistical analysis of variance drivers
Robustness
Error proofing

One way to address robustness from a coding standpoint discussed in the DVD
player case study is to treat this as a CTQ, determine the X factors, and look at
effective methods to address the risks associated with such causes.

Robustness = f (Null pointers , Memory leaks , CPU load , Exceptions , Coding errors)

Error-proofing aspects typically manifest as opportunities originating from the
FMEA study, performed as part of the design. There are six mistake-proofing princi-
ples 25 or methods that can be applied to the software design. Table 8.A.1 shows the
details of the error-proofing methods.

25 Crow, K. @ http://www.npd-solutions.com/mistake.html—Error Proofing and Design.