Validation activities that occurred in generating VARIES deliverables [D3.3], [D3.5] and [D3.6] resulted in the stepwise creation of a methodology that helps companies make informed decisions on variability use.
To deploy the service, the following artefacts are relevant:
- The variability decision conceptual model (deliverable [D3.3]).
- Variability driver descriptions (deliverable [D3.6], section 7)
- Question framework for identifying and analysing variability in the product portfolio (deliverable [D3.5], sections 3.2 and 3.3.2)
The proposed variability driver identification service consists of a number of workshops with the company and will require an expert to reflect on the material collected. A typical scenario is as follows:
- Presentation of the service, the material and methods used, and expected results for the company
- Workshop to collect information on variability decisions, based on:
- the iterative workshop format described in deliverable [D3.2], section 7.2.3.
- the question framework described in deliverable [D3.5], section 3.2 and Appendix IV.
- The expert analyses the material collected, and suggests a first set of variability drivers for the variability decisions documented in the previous step. The following intermediate deliverables will be generated for validation with the company in the next step:
- variability decision timeline (deliverable [D3.5], section 220.127.116.11)
- variability driver genome (deliverable [D3.5], section 18.104.22.168)
- initialised variability decision impact fill-in sheet
- Workshop to validate the deliverables from the previous step and to enrich the data collected:
- validation of the variability decision timeline (decision causality, early variability driver relationships)
- confirm the relevance of identified variability drivers by rating their contribution to / impact on each variability decision
- collect additional input on variability decisions: stakeholders, impact factors…
- The expert analyses the additional material collected, and creates the following final deliverables:
- prioritized variability decision timelines (deliverable [D3.6], section 9.3)
- prioritized variability driver genomes (deliverable [D3.6], section 9.4)
- variability decision impact matrix (deliverable [D3.6], section 9.5)
- analysis of patterns & recommendations
- Workshop to transfer the results.
The variability driver descriptions from deliverable [D3.6] (chapter 7) will be handed to the company, as well as some material to perform a simplified version of the service, e.g. by producing a “variability decision canvas” to help identify key factors that influence the decision process.
1.1 Prioritized variability decision timeline
The prioritized variability decision timeline (deliverable [D3.6], section 9.3) represents the relations between variability decisions and variability drivers over time. An example is given in Figure 1. The priorities relate to the impact of variability drivers on variability decisions: high impact variability drivers are rendered with a thicker arrow than medium or low impact variability drivers. This representation allows visual identification of key variability drivers, e.g.:
- High impact variability drivers can be identified by the number of wide arrows departing. See e.g. [M-07-M] and [T-03] in Figure 1.
- The most relevant variability drivers affecting a variability decision can be identified by the number of wide arrows arriving in the decison. See e.g. decision  most significantly affected by [M-07-M] and [T-08] in Figure 1.
1.2 Prioritized variability driver genome
The prioritized variability driver genome (Figure 2) represents the incidence of variability drivers over time in a matrix, where rows represent a time period (Figure 2, top) or a lifecycle phase (Figure 2, bottom), and columns represent the incidence of a variability driver in that period. The impact priority of a variability driver is expressed as sub rows labelled H (high), M (medium) and L (low) with decreasing text size for improved visualization.
The prioritized variability driver genome can be seen as a variability driver incidence heat map. The following interpretation can be proposed:
- From the timeline view (Figure 2, top):
- Technology-related variability drivers were the main initiators of variability decisions in Y1 to Y3.
- Compliance-related variability drivers were important in Y2 and Y3.
- From the lifecycle view (Figure 2, bottom):
- Technology-related variability drivers tend to appear across the entire product lifecycle.
1.3 Variability decision impact matrix
The variability decision impact matrix facilitates assessing the impact of a variability decision. It contains the following impact factors:
- Disciplines affected
- Product portfolio
- Lifecycle phase
- Stakeholders affected
- Impact factors: resources, cost, market share…
1.4 Applying the service
To help organizations make informed variability decisions, the variability driver identification service has been developed by Sirris in the VARIES project and can help organizations in their variability decisions. The methodology has been validated in 9 cases during the VARIES project, and can now be applied to other organizations.
For more information, please contact the CoIE:
|D3.2||ARTEMIS Call 2011, ARTEMIS-2011-1, 295397 VARIES: VARiability In safety-critical Embedded Systems. Deliverable D3.2: Methods to capture product variability: drivers and variants|
|D3.3||ARTEMIS Call 2011, ARTEMIS-2011-1, 295397 VARIES: VARiability In safety-critical Embedded Systems. Deliverable D3.3: Variability drivers|
|D3.5||ARTEMIS Call 2011, ARTEMIS-2011-1, 295397 VARIES: VARiability In safety-critical Embedded Systems. Deliverable D3.5: Identifying and shaping variability in the product portfolio|
|D3.6||ARTEMIS Call 2011, ARTEMIS-2011-1, 295397 VARIES: VARiability In safety-critical Embedded Systems. Deliverable D3.6: Updated methods to capture product variability: drivers and variants|