GuardTwin: Causal Root-Cause Diagnosis for IoT and Industrial Control Systems — Research Code

This repository contains the research code implementation for GuardTwin, a framework for causal root-cause diagnosis in Internet of Things (IoT) and Industrial Control Systems (ICS). The implementation provides experimental scripts for evaluating causal diagnosis, intervention-based testing, system drift, and tamper-evident auditing concepts. It includes experiment execution code, figure-generation scripts, configuration files, and example result outputs. The package is intended to support reproducibility and further research into trustworthy diagnosis and security monitoring for connected industrial systems. Research status: The included experimental results are based on synthetic data and proof-of-concept experiments. They should not be interpreted as validation on real-world industrial deployments. Contents: Experiment execution scripts Figure-generation scripts Example experimental results Documentation and reproduction instructions Please refer to the README file included in the archive for installation and execution instructions.