Decision-Support Methodology to Assess Risk in End-of-Life Management of Complex Systems

Abstract

End-of-life management of complex systems is increasingly important for industry because of growing environmental concerns and associated regulations. In many areas, lack of hindsight and significant statistical information restricts the efficiency of end-of-life management processes and additional expert knowledge is required. In this context and to promote the reuse of secondhand components, a methodology supported by risk assessment tools is proposed. The proposal consists of an approach to combine expert and statistical knowledge to improve risk assessment. The theory of belief functions provides a common framework to facilitate fusion of multisource knowledge, and a directed evidential network is used to compute a measure of the risk level. An additional indicator is proposed to determine the result quality. Finally, the approach is applied to a scenario in aircraft deconstruction. In order to support the scientific contribution, a software prototype has been developed and used to illustrate the processing of directed evidential networks.

Development of a digital twin for collaborative decision‐making, based on a multi‐agent system: application to prescriptive maintenance

“…The efficiency of cognitive tasks, and more precisely decision is of strategic importance in industry (Villeneuve et al, 2016). As the company can be seen as a system of systems (Popper et al, 2016), a MAS representation can allow the integration of the company's knowledge (Merlo, 2003).…”

A risk-based approach applied to system engineering projects: A new learning based multi-criteria decision support tool based on an Ant Colony Algorithm

“…In Aqlan and Ali (2014), the uncertainty inherent in risks is performed by using lean principles and fuzzy bow-tie analysis to improve the risk management process in the chemical industry. In Villeneuve et al (2016), the authors proposed to improve the risk assessment by using the theory of belief functions and statistical knowledge combined with the expert knowledge for aircraft deconstruction. In the case of supplier selection problem, the authors in Kaya and Karhaman (2010) developed a decision making tool that evaluates risks by using fuzzy logic models.…”

Assessment method of the multicomponent systems future ability to achieve productive tasks from local prognoses

“…The generalized Bayes theorem, developed in [43], generalizes the transferable belief model which is a development of the DST [44]. It makes the handling of epistemic uncertainty possible in belief networks binding hypotheses featured by bbas [45]. Using this ability, Simon et al in [46,47] propose an interesting approach enabling to implement the DST by the use of BN inferences for reliability analysis of complex systems.…”