Towards Secure Legacy Manufacturing: A Policy-Driven Zero Trust Architecture Aligned with NIST CSF 2.0
Open accessThe DOI matches its Crossref record; 6 of the 6 references it cites were found in a registry or an index; 7 of the 18 sources its claims cite were checked for support.
Abstract
As smart manufacturing environments continue to evolve, operational technology systems are increasingly integrated with external networks and cloud-based platforms. However, many manufacturing facilities still use legacy systems running on end-of-support/life operating systems with discontinued security updates. It is difficult to mitigate the cyber threats and risks for these systems using perimeter-based security models that isolate them from other networks. To address these constraints, a Zero Trust-based security architecture tailored for legacy manufacturing environments with practical field applicability is proposed. Our architecture builds upon the six core functions outlined in National Institute of Standards and Technology Cybersecurity Framework 2.0—identify, protect, detect, respond, recover, and govern—adapting them specifically to manufacturing environment security challenges. To achieve this, the architecture combines asset identification, policy-driven access control, secure SMB gateway transfers, automated anomaly detection and response, clean image recovery, and organizational governance procedures. This study validates the effectiveness and scalability of the proposed architecture through scenario-based simulations. When combining the EoSL defense hardening and gateway-based perimeter control, the architecture achieves approximately 99% overall threat suppression and a 98% reduction in critical-asset infection rates, demonstrating its strong resilience and scalability in large-scale legacy OT environments.
Overview
CoChat Summary
TL;DR
The paper proposes a policy-driven Zero Trust architecture for legacy manufacturing, aligned with NIST CSF 2.0 and centered on an SMB gateway, PBAC, anomaly detection/response, clean-image recovery, and governance. Simulations report that combining EoSL hardening with gateway boundary protection yields approximately 99% overall threat suppression and 98% critical-asset protection.
Key findings
- Plant OS analysis found legacy exposure: Plant A had 10,925 of 20,498 systems (65.7%) running EoSL Windows, and Plant B had 277 of 700 systems (39.6%) on outdated platforms; a cited DCS study found only 22.5% on current Windows and 77.5% on discontinued versions.
- Scenario 1: increasing EoSL defense from 20% to 80% reduced the overall infection rate from 78.51% ± 3.95% to 22.91% ± 13.87% (about 70% reduction) and critical-asset infection from 59.66% ± 3.88% to 7.32% ± 6.31% (about 88% reduction); MTTD stayed near 6 s, while MTTR fell from 682.80 ± 45.69 s to 515.52 ± 89.97 s (about 25%).
- Scenario 2: at EoSL defense 20%, raising gateway TPR from 60% to 98% reduced delivered malicious files from 30.9 to 0.8 and Total Inf from 3.09% to 0.08% (about 97% reduction); internal blocking was mostly 0% in this boundary-only setup.
- Integrated analysis reports approximately 99% overall threat suppression and 98% protection of critical assets when EoSL hardening and gateway-based boundary protection are combined.
Methods
- Design: a ZTA for legacy manufacturing aligned to NIST CSF 2.0’s six functions, integrating asset identification, policy-based access control, SMB gateway file control, anomaly detection and response, clean-image recovery, and governance.
- Data/sample: empirical OS analysis of two manufacturing plants (Plant A: 20,498 systems; Plant B: 700 systems; data collected April 2025) plus Python/SimPy simulations. Scenario 1 used N = 1000 nodes, T = 3600 s, EoSL ratio 0.60, critical ratio 0.20, defense levels 20/40/60/80%, and 20 runs per level. Scenario 2 used gateway TPR values {60, 80, 90, 98}%, fixed FPR = 2.0%, file rate λf = 0.2 s−1, malicious file proportion p_mal = 0.10, inactive internal propagation, and 10 repeats per parameter combination.
- Analysis: event-driven simulation of infection, detection, isolation, and recovery; metrics included blocking rate, total/EoSL/non-EoSL/critical infection rates, MTTD, MTTR, delivered malicious files, and gateway blocking rate, reported as mean ± standard deviation.
Limitations
- Stated in the paper: network-based/agentless detection may need EDR or HIDS for APTs; non-networked assets are hard to identify; organizational barriers include operator acceptance, PBAC administrative overhead, and security-productivity balance; policy/regulatory gaps include lack of legal mandates and scarce OT-specific ZT guidelines; no full proof-of-concept was performed because reconstructing EoSL operating systems is difficult, with future limited pilots planned at Plant A and Plant B; the architecture is scoped to user-initiated file transfers and does not address strict real-time file-transfer requirements.
- Evident from the text: validation is simulation-based rather than a real deployment; Scenario 2’s boundary-only evaluation sets internal propagation inactive (λ_int = 0), isolating gateway effects; plant-level OS data are not publicly available due to confidentiality agreements.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
6 Valid · 1 Not valid · 7 Not checked · 54 Own results and interpretations
Every source the paper cites for this claim was checked and supports it.
In the paper
“In industrial environments, traditional perimeter-based security approaches like firewalls and IP-based access control are increasingly ineffective against the modern threat landscape due to their inherent limitations [11].”
§ Towards Secure Legacy Manufacturing: A Policy-Driven Zero Trust Architecture Aligned with NIST CSF 2.0 > 2.5. Limitations of Conventional Security Approaches
What the sources say
- Supportedcites [11] Paul, B., and Rao, M. (2023). Zero-Trust Model for Smart Manufacturing Industry. Appl. Sci., 13. source
Section 2.1 says traditional perimeter-based security trusts traffic inside the border, allowing attackers easy lateral access once they cross it. Section 6 describes the resulting access-control weakness in smart manufacturing, including access granted because a source IP is internal. The Abstract says this architecture was designed for static systems and needs to adapt to emerging technologies; together, these passages support the claim's broad characterization. (LLM, conf=0.87)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“The architecture proposed in this study aligns structurally with NIST CSF 2.0, and suggests potential validity and practical applicability as a strategic security model [2].”
- Also stated in: § 5.1. Conformance Evaluation Based on NIST CSF 2.0 · boosted
“As summarized and mapped above, the proposed architecture aligns structurally with the six core functions of NIST CSF 2.0, establishing its validity as a standards-based integrated cybersecurity architecture.”
- Also stated in: § 5.1. Conformance Evaluation Based on NIST CSF 2.0 · boosted
“Table 6 maps the proposed security components to the CSF functions and categories, demonstrating that the architecture fulfills the requirements of a standards-based security framework.”
§ Towards Secure Legacy Manufacturing: A Policy-Driven Zero Trust Architecture Aligned with NIST CSF 2.0 > 5.1. Conformance Evaluation Based on NIST CSF 2.0
What the sources say
- Supportedcites [2] National Institute of Standards and Technology (2024). NIST CSWP 29, National Institute of Standards and Technology (NIST). The NIST Cybersecurity Framework (CSF) 2.0. source
Section 1, “Cybersecurity Framework (CSF) Overview,” identifies this document as NIST CSF 2.0 and describes its structure: a Core organized into Functions, Categories, and Subcategories, alongside Organizational Profiles and Tiers. The citation identifies the framework used for the citing authors’ structural comparison; their assessment of their own architecture and its potential validity is not a finding attributed to NIST. (LLM, conf=0.86)
boostedstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § 5.1. Conformance Evaluation Based on NIST CSF 2.0 · boosted
Every source the paper cites for this claim was checked and supports it.
In the paper
“The perimeter-based models are particularly vulnerable to lateral movements, authentication bypasses, and external intrusions [11].”
§ Towards Secure Legacy Manufacturing: A Policy-Driven Zero Trust Architecture Aligned with NIST CSF 2.0 > 2.5. Limitations of Conventional Security Approaches
What the sources say
- Supportedcites [11] Paul, B., and Rao, M. (2023). Zero-Trust Model for Smart Manufacturing Industry. Appl. Sci., 13. source
Section 6 (page not available) describes a perimeter-based model that implicitly trusts devices inside the network: internal devices can reach manufacturing equipment without authentication, and a remote user who passes a single VPN check can access resources without further validation. This supports the paraphrase about lateral access, authentication bypass, and exposure to external intrusion. (LLM, conf=0.87)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“The application of security patches is frequently delayed or postponed indefinitely to minimize system downtime [11].”
§ Towards Secure Legacy Manufacturing: A Policy-Driven Zero Trust Architecture Aligned with NIST CSF 2.0 > 2.5. Limitations of Conventional Security Approaches
What the sources say
- Could not verifycites [11] Paul, B., and Rao, M. (2023). Zero-Trust Model for Smart Manufacturing Industry. Appl. Sci., 13. sourceExperimental — accuracy not yet measured
Section 7.1.4 (page unavailable) discusses tracking OT patching inventory and identifying noncompliant endpoints, but does not say patches are frequently delayed or postponed to avoid downtime. Searches for patching delays and downtime found no passage addressing that claim. (LLM, conf=0.74)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Could not verify
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“in Plant B, 277 of the 700 systems (approximately 39.6%) were using outdated platforms that were no longer eligible for security patches”
§ Towards Secure Legacy Manufacturing: A Policy-Driven Zero Trust Architecture Aligned with NIST CSF 2.0 > 2.3. OS Status of Manufacturing Process Systems
plainSingle case- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
38 more claims
Create a free account to see all 68 claims and to verify the 2 references still open.
- It may state two separate claims as one. 18
- The quote only partly states this claim. 1
- Probably not a substantial claim. 6
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
“| EoSL | Windows XP, 7, Server 2003, etc. | 10,925 | 65.7% | | Supported | Windows 10, 11, etc. | 7573 | 34.3% | | Total | - | 20,498 | 100% |”
§ Towards Secure Legacy Manufacturing: A Policy-Driven Zero Trust Architecture Aligned with NIST CSF 2.0 > 2.
Reported 20498 · recomputed 18498
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
“The analysis revealed that in Plant A, 10,925 of 20,498 systems (approximately 65.7%) were running EoSL versions of Windows OS”
§ Towards Secure Legacy Manufacturing: A Policy-Driven Zero Trust Architecture Aligned with NIST CSF 2.0 > 2.3. OS Status of Manufacturing Process Systems
Reported 65.7 · recomputed 53.297883
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
“| Supported | Windows 10, 11, etc. | 7573 | 34.3% | | Total | - | 20,498 | 100% |”
§ Towards Secure Legacy Manufacturing: A Policy-Driven Zero Trust Architecture Aligned with NIST CSF 2.0 > 2.
Reported 34.3 · recomputed 36.945068
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Towards Secure Legacy Manufacturing: A Policy-Driven Zero Trust Architecture Aligned with NIST CSF 2.0 > 6.2.5. Summary of Results
“According to the simulation results (T = 3600 s, N = 20)”
§ Towards Secure Legacy Manufacturing: A Policy-Driven Zero Trust Architecture Aligned with NIST CSF 2.0 > 6.2.2. Simulation Environment Configuration
“Scenario 1 (EoSL Defense Impact Simulation) was conducted on an industrial network consisting of 1000 nodes over a period of 60 min (3600 s).”
Citations
Cites
Rose, S., Borchert, O., Mitchell, S., & Connelly, S. (2020) · doi:10.6028/nist.sp.800-207 · Cited by 1,174
The NIST Cybersecurity Framework (CSF) 2.0
National Institute of Standards and Technology (2024) · doi:10.6028/nist.cswp.29 · Cited by 266
Rose, S., Borchert, O., Mitchell, S., & Connelly, S. (2020) · doi:10.6028/nist.sp.800-207-draft2 · Cited by 239
Zero-Trust Model for Smart Manufacturing Industry
Paul, B., & Rao, M. (2022) · Applied Sciences · doi:10.3390/app13010221 · Cited by 53 · cited 7×
“…and equipment, the need for maintaining continuous operations without production downtime, and the inability to upgrade legacy applications because of the absence of vendor support [8,10,11], but they have not provided sufficiently concrete and actionable strategies to overcome these limitations.”
InfluentialImproving Age of Information for Covert Communication With Time-Modulated Arrays
Ma, Y., Ma, R., Lin, Z., et al. (2024) · IEEE Internet of Things Journal · doi:10.1109/jiot.2024.3466855 · Cited by 51 · cited 1×
“Furthermore, recent studies on Age of Information (AoI) have highlighted the importance of maintaining temporal freshness in data-driven decision processes [18,19].”
Zhang, G., Xin, W., Tan, X., et al. (2025) · IEEE Transactions on Communications · doi:10.1109/tcomm.2025.3525566 · Cited by 23 · cited 1×
“Furthermore, recent studies on Age of Information (AoI) have highlighted the importance of maintaining temporal freshness in data-driven decision processes [18,19].”
Research works printed but not linked: 2
- Li, J., Zhou, H., Wu, S., Luo, X., Wang, T., Zhan, X., and Ma, X. (2022, January 10–12). FOAP: Fine-Grained Open-World Android App Fingerprinting. Proceedings of the 31st USENIX Security Symposium (USENIX Security 22), Boston, MA, USA.
- Ni, T., Lan, G., Wang, J., Zhao, Q., and Xu, W. (2023, January 9–11). Eavesdropping Mobile App Activity via Radio-Frequency Energy Harvesting. Proceedings of the 32nd USENIX Security Symposium (USENIX Security 23), Anaheim, CA, USA.
Cited by
Toibin, G. E., Lurie, Y., & Mark, S. (2026) · Telecom · doi:10.3390/telecom7050113 · Cited by 1
Hamza, M., Dharani, B., Sethi, G., et al. (2026) · doi:10.1109/qpain69676.2026.11545723
“Nonetheless, KPI behaviors are dynamic in nature and are affected by internal policies, external market forces, changes in the work force and intervention strategies [2].”
Ma, Z., Shi, C., Xue, G., et al. (2026) · doi:10.1109/iits69561.2026.11648439
“…validated and finalized using the Delegated Proof of Stake (DPoS) consensus mechanism on the consortium chain ’ s main chain, which is consistent with recent secure edge-network and policy-driven trust management studies emphasizing cross-domain verification and controlled resource access [3,4] .”
Akhmetkaliyev, A., Ala’anzy, M. A., Almsafir, M. K., & Musleh, S. (2026) · Computer Networks · doi:10.1016/j.comnet.2026.112695
State-Aware Zero Trust Architecture for Offshore Wind Farms Based on Software-Defined Perimeter
Lim, C.-M., Shin, D.-H., Kim, K.-K., & Euom, I.-C. (2026) · IEEE Access · doi:10.1109/access.2026.3732584
“Min et al. [17] proposed a policy-driven zero trust architecture tailored to legacy manufacturing environments and aligned it with the functional structure of the NIST Cyber-security Framework (CSF) 2.0.”
Influential
Metadata
- Authors
- Min, C.-H., Kim, D.-H., Yang, H., & Kwak, J.
- Affiliations
- Ajou University (KR)
- University of Electronic Science and Technology of China (CN)
- Corresponding author
- Kwak, Jin
- Venue
- Electronics
- Publisher
- MDPI AG
- Published
- 2025-10-20
- Volume
- 14
- Issue
- 20
- Pages
- 4109
- ISSN
- 2079-9292 (ISSN-L)
- OpenAlex
- W4415369977
- Semantic Scholar
- 53c74d172c4f71a9fb7166f91d30ad29e21f7900
- Cited by
- 6
- References
- 4
- Citations by year
- 2026: 6
- FWCI
- 3.61
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Electronics (publishedVersion, cc-by)
- Funders
- ICT R&D program of MSIT/IITP (RS-2024-00438597)
- Korea government (RS-2024-00400302)
- Topics
- Physical Sciences › Computer Science › Information Systems › Information and Cyber Security
- Physical Sciences › Computer Science › Computer Networks and Communications › Network Security and Intrusion Detection
- Physical Sciences › Engineering › Control and Systems Engineering › Smart Grid Security and Resilience
- Keywords
- zero-trust architecture, NIST Cybersecurity Framework 2.0, operational technology security, anomaly detection, image restoration, organizational governance, perimeter security
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-09, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-09 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.