A systemic approach for stochastic reliability management in human-machine systems

In today's complex engineered systems, comprising a multitude of interacting components, preserving system performance is of utmost importance. The challenge often lies in effectively prioritizing components with the highest potential to compromise system reliability, mainly when human interaction with technical artefacts is not negligible. This study proposes a systemic methodology for pragmatic reliability management within human-machine systems. The proposed approach combines a rule-based adaptation of the well-established Failure Mode, Effects, and Criticality Analysis (FMECA) with a probabilistic Fault Tree Analysis (FTA). Furthermore, the technical considerations are seamlessly integrated into a human-centric analysis, utilizing the Standardized Plant Analysis Risk - Human Reliability Analysis (SPAR-H). The proposed decision-support methodology is instantiated through Monte Carlo simulations to account for stochastic phenomena and uncertain operating conditions. The effectiveness and practicality of the proposed approach are elucidated through a case study involving a high-reliability system, specifically a high-mobility multi-wheeled vehicle. This study demonstrates the step-by-step application of the proposed approach and its implications in challenging operating scenarios, reaffirming its potential to enhance reliability management within human-machine systems.