An Integrated Checklist for Architecture Design of Critical Software Systems

Abstract

With the advancement of digitalization, critical information infrastructures, such as intelligent energy distribution, transportation, or healthcare, have opened themselves towards intelligent technological opportunities, including automation of previously manual decision making. As a side effect, the digitalization of these infrastructures gives rise to new challenges, especially linked to the complexity of architecture design of these infrastructures, to later support necessary software quality and safeguard the systems against attacks and other harm. To support software architects in the design of these critical software systems, well structure architectural knowledge would be of great help to prevent the architects from missing some of the crucial concerns that need to be reflected with built-in architectural mechanisms, early during architecture design.Given the narrow scope of existing guidelines, with the need of browsing and combining multiple sources, this paper proposes an integrated checklist to cover the breath of architectural concerns for the design of critical software systems, covering the need for built-in mechanisms to prevent, detect, stop, recover from and analyse intentional as well as unintentional threats to system dependability. Contrary to existing guidelines that typically focus on runtime incident handling, our checklist is to be used during architecture design to ensure that the system has built-in mechanisms to either handle the incidents automatically or include the right mechanisms to support the runtime incident handling.