Architecture analysis to predict project risks

My PhD is about the design time risk assessment of software architectures. My motivation is to help projects avoid costly and time consuming issues. Te proposition is to investigate whether risk inducing architecture elements can be isolated into containers at design time. Doing so would enable containers to be ranked using risk indicators and mitigations to be better scoped and estimated. It is envisaged these abilities would lead to cost and time savings as fewer risks turn into implementation issues. To evaluate the proposition, container level design metrics will be compared to the implementation, to determine whether risks could have been isolated within containers at design time. To date, two industrial case studies have been used to determine which of three container types tested, is most effective at isolating implementation error-proneness risks