CYBER-PHYSICAL SECURITY FRAMEWORKS FOR INTEGRATED CAD/CAM AND DIGITAL TWIN WORKFLOWS: A STATE-OF-THE-ART SYSTEMATIC REVIEW

Integrated computer-aided design and manufacturing (CAD/CAM), product lifecycle management (PLM) and industrial control and digital-twin platforms provide a continuity from design to production, enhancing traceability, but they also enable an impersonated identity, model, toolpath, software component or sensor stream to travel through the entire workflow. The review brings together 56 peer-reviewed articles and six normative sources, most of which were from the period 2016 to September 2026. OpenAlex, publisher records (as retrieved from Crossref) and backward citation chaining were processed through a PRISMA 2020-driven workflows, and evidence was coded by workflow stage, asset, threat, security objective, control family, and validation strength. Network intrusion, theft or tampering of CAD models, manipulation of CAM parameters and post-processors, falsification of a digital-twin state, insecure cloud collaboration, supply-chain compromise, and unsafe shop-floor commands are the main threats. There are seven families of published countermeasures: standards-based governance, identity-centred access control, segmentation, verifiable provenance, blockchain-backed traceability, watermarking and fingerprinting, and twin-assisted anomaly detection. No family is big enough: source data were not necessarily correct; anomaly detection was based on representative data; and conventional IT controls under-represented operational safety and timing. Due to the influence of process parameters on part quality, such as feed rate, injection pressure or infill density, small, unauthorized changes to the process can result in parts that are defective but appear plausible. As such, a lifecycle defence-in-depth approach is proposed, featuring ownership of artefacts, trust boundaries, a signed manufacturing job package, least privilege, monitored hand-offs, process-aware detection and verified recovery. Future work should validate integrated frameworks on realistic CAD-to-machine testbeds, such as learning factories, and report the outcomes of security, production, safety and recovery in a combined manner.