Hardware Reference
In-Depth Information
Table 3.
Organising QATs into Different Risk Management Categories
Risk Category
Safety Techniques
Performance Techniques
Risk Identification
HAZOP, Checklist, ETBA
EG, UCM, Interactive Tree
Algorithm, Performance An-
tipatterns
Risk Analysis
FMEA, FMECA, FTA, ETA
Software Architecture Analy-
sis Method (SAAM), PASA;
Layered queuing network
(LQN), Stochastic Petri Nets,
Markov Chains
Risk Prioritisation
FMEA,FMECA,Critica ity
Analysis, Consequence Analy-
sis
Layered queueing network
(LQN), Markov Chains
Risk
Treatment
Hazard Reduction Design (e.g.
Simplification and decoupling)
Principles (e.g. Locality, Par-
allel Processing)
-
Risk
Reduction
Design
Risk
Treatment
-
Error Recovery (e.g. feedback,
checking procedures)
Performance Tuning, Perfor-
mance
Corrective Actions
Patterns
(e.g.
Fast
Path Speed-Up)
Risk Conformance -
Verification
Sneak Circuit Analysis, Con-
trol Flow Analysis and Bound-
ary Value Analysis
Load Testing, Stress Testing,
Instrumentation
Risk Conformance -
Validation
SMORT, Safety Review, Fa-
gan Inspection
Benchmark, Profilers
Development teams can use the QATF to help them compare FMEA with other
safety QATs such as FTA and ETA, to determine a sequence of using QATs by
referring to the process information captured by the QATF. The most appropri-
ate QATs can be selected to execute specific safety tasks. For example, FTA and
ETA can be used when the design is completed. FTA begins with all hazards
identified from other QATs such as FMEA and HAZOP and works backwards
to determine their possible causes until reaching a base event. ETA uses inputs
from QATs such as FTA to analyse all possible consequences and determine the
percentage of consequences which lead to the desired result.
As with safety, there are various performance techniques available to iden-
tify and address performance problems throught development processes. These
QATs include performance estimation techniques, performance modelling tech-
niques, performance evaluation techniques to ensure that the implementation
meets performance objectives and also some principles and patterns for perfor-
mance design. Table 2 organises some performance QATs into different devel-
opment phases. For example, there are a set of performance-oriented principles
to identify design alternatives that help to meet specific performance objectives.
Design engineers can use the QATF to help choose the most suitable principle
by referring to the definition and examples of applying these principles.
Table 3 organises some of the safety and performance QATs into different qual-
ity risk management types (refer to Section 3.3). QATs are categorised based on
