An Analysis of Early-Stage Functional Safety Analysis Methods and Their Integration into Model-Based Systems Engineering

TL;DR

This paper compares early-stage safety analysis methods like FMEA, FHA, and FFIP, highlighting FFIP's strengths and reviewing their integration into Model-Based Systems Engineering, revealing a nascent but promising research area.

Contribution

It provides a comparative analysis of safety analysis techniques and reviews their integration into MBSE, emphasizing FFIP's advantages and identifying gaps in current integration efforts.

Findings

FFIP effectively identifies emergent behaviors and fault propagation.

MBSE integration efforts mainly focus on FMEA, with limited work on FHA and FFIP.

No universal framework exists for integrating safety analysis methods into MBSE.

Abstract

As systems become increasingly complex, conducting effective safety analysis in the earlier phases of a system's lifecycle is essential to identify and mitigate risks before they escalate. To that end, this paper investigates the capabilities of key safety analysis techniques, namely: Failure Mode and Effects Analysis (FMEA), Functional Hazard Analysis (FHA), and Functional Failure Identification and Propagation (FFIP), along with the current state of the literature in terms of their integration into Model-Based Systems Engineering (MBSE). A two-phase approach is adopted. The first phase is focused on contrasting FMEA, FHA, and FFIP techniques, examining their procedures, along with a documentation of their relative strengths and limitations. Our analysis highlights FFIP's capability in identifying emergent system behaviors, second-order effects, and fault propagation; thus, suggesting it is better suited for the safety needs of modern interconnected systems. Second, we review the existing research on the efforts to integrate each of these methods into MBSE. We find that MBSE integration efforts primarily focus on FMEA, and integration of FHA and FFIP is nascent. Additionally, FMEA-MBSE integration efforts could be organized into four categories: model-to-model transformation, use of external customized algorithms, built-in MBSE packages, and manual use of standard MBSE diagrams. While our findings indicate a variety of MBSE integration approaches, there is no universally established framework or standard. This leaves room for an integration approach that could support the ongoing Digital Engineering transformation efforts by enabling a more synergistic lifecycle safety management methods and tools.