3.2.3 Risk documentation
As noted before, the Identify-Assess-Control cycle in Figure 3.2 may need to be repeated
as countermeasures can introduce new risks in their turn; risk management is an iterative
process. This process should be document and to support requirements evolution – changes
rationale for countermeasure requirement and the need for different countermeasures. The documentation is also needed for risk monitoring at system runtime and dynamic selection of more appropriate countermeasures.
Risk documentation should ideally include, for each identified risk:
• The conditions or events characterizing is occurrence.
• Its estimated likelihood of occurrence.
• Its possible causes and consequences.
• The estimated likelihood and seventy of each possible consequence
• The countermeasure that were identified together with respective risk reduction leverage.
• The Select subset of countermeasures.
This documentation can be organized around risk trees . We will come back to this in Chapter 9
3.2.4 Integrating risk management in the requirements lifecycle
NASA’s Defect Detection Prevention approach (DDP) is a notable effort to systematize the
Identify – Assess – Control cycle and integrate risk management in the RE process (Feather and Cornford, 2003, Feather et al., 2005). The approach is supported by a quantitative reasoning
risk consequences as loss of attainment of the corresponding objective.
In DDP , object, risk and countermeasures are called requirement, failure modes and
PACTs, respectively. There is a coarser counterpart of the notion of risk-reduction leverage,
Called effectiveness, defined as the proportion by which a countermeasure reduces a risk.
The likelihood of risk, the severity of consequences and the effectiveness of countermeasures must be estimated quantitatively by elicitation from expert or from accumulated measurements.
As Figure 3.5 show, the DDP approach consists of there steps.
Step 1: Elaborate the impact matrix
We first need to build a risk-consequence table with domain experts. This table captures
The estimated severity of the consequences of each risk. For each pair of objective and
associated risk r, it specifies an estimated loss of proportion of attainment of the objective if
the risk occurs – from 0 (no loss) to 1 (total loss). This last line of the table yields the overall
criticality of each risk, obtained by summing down the corresponding column in the following
weighted way:
The most risk – driving objectives are thereby highlighted.
Table 3.3 show an impact matrix for our library management case study (the value in
the last line/column were rounded to two significant digits). Note again that the numbers
taken individually have no accurate meaning; they are to be considered in a relative sense
for pairwise comparison. Also note that we might play with such tables in a spreadsheet –
like fashion to see how criticalities and losses are affected by modifications of weights,
likelihoods and loss of proportion of attainment. From this table the DDP tool produces
a bar chart visualizing the critical impact, by decreasing order, of the various risk on all
objectives.