Enhancing State-of-the-art Safety Case Patterns to Support Change Impact Analysis

abstract

The new generation of safety-critical systems will be interconnected, having other systems as collaborating partners for achieving common goals (e.g., interconnected cyber-physical systems such as connected cars, or collaborative embedded systems such as an advanced driver assistance system connected with different sensors). Frequent new business goals of such systems are to enable new collaborations with new types of technical systems, thus changing their operating context, which triggers the need for agile development in automotive. In safety-critical domains, a change in the operating context triggers the need for impact analysis on the artefacts generated during the safety lifecycle. Impact analyses are time and resource consuming, hindering agile development. Hence, the need for automation. Safety cases comprise safety arguments explicitly specifying the traces among the artefacts generated during the safety lifecycle. Our longer term goal is to support the automated identification of the artefacts affected by changes in the system's operating context, while proposing an automated change impact analysis executed on the system's safety case. To ensure completeness of the results of such analysis, in this work, we enhance state-of-the-art safety case patterns by referencing all artefacts generated during the safety lifecycle. Further, we enable the explicit specification of the properties of the operating context for which we foresee certain changes. We evaluate our patterns by using them for the construction of the safety case of a simplified airbag system.