Towards Enabling Reuse in the Context of Safety-Critical Product Lines

Abstract

In the context of safety-critical product lines, industries have to face an urgent challenge: reduction of time and cost for the creation of a safety case. A safety case is a contextualized structured argument constituted of process and product-based sub-arguments to show that a system is acceptably safe. Its creation is an extremely time-consuming and costly activity. To reduce time and cost, reuse capabilities are being investigated from different perspectives however currently no satisfying approach is available. In this paper, we propose a new methodological framework called Anti-Sisyphus. Anti-Sisyphus is aimed at enabling reuse by combining process lines, product lines and safety case lines. The systematization of what varies and what remains in common with respect to process as well as product elements in turns enables the systematization of what varies and remains in common in terms of process-based as well as productbased arguments within a safety case line. Anti-Sisyphus thus has the potential to enable a 3D reuse.

Opportunities for software reuse in an uncertain world: From past to emerging trends

“…As known, Ariane 5 contained a piece of code that was appropriate for its predecessor Ariane 4 but that turned out to be catastrophic for Ariane 5. As observed by Gallina, as a consequence of the Ariane 5 accident, asset reuse in safety‐critical systems engineering has been considered a taboo . On the one hand, the “to‐reinvent or not‐to‐reinvent” dilemma has been often solved by a tacit acceptance that starting from scratch is safer.…”

Constructing product-line safety cases from contract-based specifications

“…The first line of research has focused on capturing safety case patterns that have been successfully used to argue for different properties [10,15] while the second line of research has focused on generating safety case arguments from various types of artifacts [14,26,37]. In the area of PLE, most notable contributions [14,18,19,21,26,36,37] come from the latter group.…”

“…As argued in Section 1, the present paper considers development and assurance of arbitrary PLs and consequently relies on CBS paradigm where the constructed assurance case argues for system level properties instead of only SEooC. The line of research in [18,19,21] introduces approaches for constructing safety case fragments that target particular requirements from various domain-specific safety standards. A notable aspect of these approaches is the attention to processes and their assurance with respect to standardized practices.…”

Quantitative evaluation of tailoring within SPICE‐compliant security‐informed safety‐oriented process lines

“…In critical‐product line engineering, as discussed by Gallina, changes in the criticality of the products have an impact on the stringency of the processes used to plan/develop them, and as a consequence, these product/process changes have an impact on the corresponding subarguments of the assurance cases used to argue about process compliance and product's dependability. Within the AMASS project, BVR Tool has been integrated not only with EPF Composer for managing the variability at process level but also with CHESS Toolset to manage the variability at product level and with OpenCert to manage the variability at argumentation level.…”

“…Within the AMASS project, BVR Tool has been integrated not only with EPF Composer for managing the variability at process level but also with CHESS Toolset to manage the variability at product level and with OpenCert to manage the variability at argumentation level. Thus, the seminal ideas, presented by Gallina, were made concrete within the AMASS platform. Figure depicts on the left the three‐dimension–oriented conceptual variability management and on the right the tooling infrastructure available within the AMASS platform to support it.…”