Identifying Critical Failure Chains for Resilience in Cyber-Physical Systems using Symbolic Execution

Cyber-physical systems comprise tightly interconnected software and hardware components whose interactions under complex operating conditions can give rise to hazardous emergent behaviors. We present an automated approach to identify potential failure chains across multiple components by systematically exploring feasible inter-component communication paths using symbolic execution directly on component binaries. Building on the angr binary analysis framework, our multi-phase analysis models a message-passing bus, symbolically executes each component to derive input–output constraints, and assembles a directed communication graph whose paths correspond to feasible chains. These chains provide actionable insight for resilience engineering by revealing hidden dependencies and opportunities for containment of error/attack propagation. We report a prototype implementation, MCS Analyzer, capable of extracting paths with minimal manual configuration and without source code, discuss limitations such as path explosion and timing abstraction, and outline how the resulting graph supports safety and security assessments in general Cyber-physical systems settings.