Risk-Based Resource Allocation for Collaborative System Design in a Distributed Environment

Abstract

Risk analysis is important in system design because of its essential role in evaluating functional reliability and mitigating system failures. In this work, we aim at expanding existing risk modeling methods to collaborative system designs: specifically, to facilitate resource allocation among distributed stakeholders. Because of different perspectives and limited local information, inconsistent and/or incoherent risk assessments (such as different probability and confusing consequence evaluations) may occur among stakeholders, who are responsible for same or different risk components of a system. The discrepancies can become potential barriers in achieving consensus or acceptable disagreement for distributed resource allocation. Built upon our previous work, a risk-based distributed resource allocation methodology (R-DRAM) is developed to help a system manager allocate limited resources among collaborating stakeholders based on a cost-benefit measure of risk. Besides probability and consequence, two additional risk aspects, tolerance and hierarchy, are considered for system risk modeling in a collaborative/distributed environment. Given a total amount of resources to be allocated, the four risk aspects are combined to form the cost-benefit measure in a multiobjective optimization framework for achieving a desired risk reduction of a targeted system. An example is used to demonstrate the implementation process of the methodology. The preliminary investigation shows promise of the R-DRAM as a systematic and quantifiable approach in facilitating distributed resource allocation for collaborative system design.