System risk analysis enhanced with system graph properties

Classical failure modes and effects analyses are known to be tedious activities and perceived by design teams as reengineering efforts. This research investigates a new manner to cut down side effects of classical risks analyses. It proposes to use the complexity measure of the underlying graph of the system being designed, to infer information about risks. To do so, the research method is three-folded: (1) a literature review is made to select relevant risk analyses and graph methodologies employed during design to ground our proposal (2) the proposal is built (3) the feasibility of this concept is tested over a real product design. The main result of this paper is to unveil a new manner to consider risks for improving their control during design stages.