REVIEW 4 major objections 6 minor 89 references
AgenticTwin: An Agentic LLM Framework Integrated with Digital Twin for Anomaly Detection
T0 review · 4 major / 6 minor · reviewed 2026-08-16 · deepseek-v4-flash
Pith's one-line read AgenticTwin shows that small grounded LLM agents can outperform a 70B monolithic model on digital-twin anomaly diagnosis, retrieval, and mitigation.
desk verdict Sensible framework and useful benchmark, but the headline agentic-win numbers are internally inconsistent and the reference-based scoring is circular. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The mechanism that carries the argument is the structured anomaly event $E_t = (\tau_t, x_t, \hat{x}_t, r_t, A_t)$, built from the observed sensor vector, the digital twin's predicted state, the residual $r_t = x_t - \hat{x}_t$, and the anomaly classifier's output. Rather than consuming raw time series, each LLM agent operates on this event: the Diagnosis Agent ($F_{diag}(E_t, KB)$) explains root causes from residuals, labels, and the knowledge base; the Retrieval Agent filters and ranks past events from the anomaly repository by shared labels, values, and residual patterns; the Mitigation Agent ($F_{mit}(E_t, DA_t, RA_t, KB)$) turns diagnosis plus retrieved history into corrective actions; and the Supervisor Agent fuses the three outputs into one operator-facing response. The residual is the physics-informed deviation signal, and the curated knowledge base supplies a compact taxonomy of faults (definition, symptoms, causes, mitigation) that keeps generation grounded.
What would settle it
Run the identical agentic and monolithic pipelines on a set of real faults whose root causes and successful mitigation actions are verified by human domain experts, and compare text outputs against those expert judgments; if the agentic advantage shrinks or the knowledge-base gain disappears under expert-grounded scoring, the paper's central claim is refuted. A faster check is to regenerate the benchmark references with a different strong LLM and see whether the relative rankings of models and agents remain stable.
Extended reading notes
Core claim
On its own terms, the paper's central discovery is that task specialization and grounded evidence can substitute for raw model scale in anomaly reasoning. In the central comparison, a single task-specific open-source LLM was assigned to each of three roles — Qwen 2.5 3B for diagnosis, Qwen 2.5 7B for retrieval, DeepSeek-R1-Distill-Llama 8B for mitigation — for a total of 18B parameters, while the strongest monolithic baseline, Llama 3.3 70B, was asked to perform all three tasks in one prompt with the same information. The specialized assembly scored 0.93 diagnosis similarity, 0.95 retrieval MRR, and 0.92 mitigation similarity, against 0.70, 0.74, and 0.78 for the 70B model. The paper also shows that the curated anomaly knowledge base is the main lever for small models: average diagnosis similarity rises from 0.69 to 0.84 and mitigation from 0.66 to 0.84, while relative gains shrink as model size grows. These results are presented as evidence that lightweight open-source LLMs, organized as collaborating agents and grounded in digital-twin residuals, can deliver practical, interpretable anomaly analysis in resource-constrained cyber-physical systems.
Load-bearing premise
The benchmark's reference answers, used as ground truth for diagnosis and mitigation, are themselves generated by a proprietary LLM from the same injected anomaly metadata, digital-twin outputs, and curated knowledge base that the evaluated agents receive; if those references are not themselves correct, the reported similarity scores and the knowledge-base gains built on them do not measure true diagnostic or mitigation quality.
Editorial extensions
If this is right
- Access to the anomaly knowledge base reliably improves diagnosis and mitigation for every evaluated LLM, so adding a structured KB is a low-cost intervention in digital-twin anomaly pipelines.
- Small open-source models gain the most from KB grounding, with relative improvements up to 43.5%, making lightweight deployment practical in resource-constrained cyber-physical environments.
- An 18B-parameter agentic configuration can outperform a 70B monolithic model across all three tasks, so task decomposition is a viable substitute for raw model scale.
- Because the benchmark is constructed from synthetic injections with known labels and structured anomaly events, it enables reproducible comparisons of future LLM-based anomaly reasoning systems.
- Multi-fault classifier results show the evidence layer can support concurrent-fault reasoning, with the selected classifier reaching an F1 of 0.94 under multi-fault conditions.
Reading between the lines
- A human-expert evaluation set with independently verified root causes and mitigation outcomes would test whether the reported similarity gains reflect true diagnostic quality rather than agreement with the LLM-written references.
- The same residual-plus-taxonomy pattern should transfer to other sensor-rich domains, such as manufacturing, energy grids, and medical monitoring, wherever a physics-informed twin and a domain fault taxonomy are available.
- If mitigation recommendations were executed on the twin or the physical asset, the framework could be scored by operational outcomes such as recovery time, giving a closed-loop measure beyond text similarity.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper proposes AgenticTwin, a multi-agent LLM framework integrated with a digital twin (DT) for anomaly diagnosis, historical retrieval, and mitigation recommendation in cyber-physical systems. A regression-based DT estimates expected sensor states; a residual-based classifier flags anomalies; and four agents (DA, RA, MA, SA) generate grounded explanations. The authors construct a benchmark by injecting spike, drift, stuck-at, and replay anomalies into a real-world weather dataset and use Gemini 2.5 Pro to generate 12,000 synthetic operator queries with reference responses. They evaluate 15 LLMs as individual agents, report that adding an anomaly knowledge base improves diagnosis and mitigation similarity, and compare an 18B-parameter AgenticTwin composed of three open-source models against monolithic baselines up to 70B, claiming that task specialization compensates for limited model scale.
Significance. If the reported results were reliable, the paper would make a useful contribution: it offers a modular agentic design for grounding LLM reasoning in DT outputs, targets lightweight open-source models for resource-constrained CPS, and uses objective retrieval metrics (Precision@k, Recall@k, MRR). The breadth of evaluated models and the controlled injection of synthetic anomalies are also strengths. However, the central comparative claims currently rest on internally inconsistent tables and on an evaluation reference that is generated by an LLM from the same knowledge base given to the agents. Until these issues are resolved, the headline conclusions about KB gains and agentic superiority over monolithic models are not supported.
major comments (4)
- [§IV-E, Table VI] The headline agentic-versus-monolithic comparison is internally inconsistent with the component-level results. In AgenticTwin, the DA is Qwen 2.5 3B Instruct operating on (E_t, KB) per Eq. (16), exactly the condition evaluated in Table IV's KB column, yet its diagnosis similarity is 0.93 in Table VI versus 0.84 in Table IV. Likewise, Qwen 2.5 7B as the RA reaches MRR 0.95 in Table VI versus 0.89 in Table V, and DeepSeek-R1-Llama-8B as the MA reaches mitigation similarity 0.92 versus 0.88 in Table IV. No change in query set, prompt template, temperature, or scoring is reported between these evaluations. These gaps are far too large to be sampling noise. Until the authors reconcile these numbers or document the protocol change, the claimed 32.9%, 28.4%, and 17.9% gains over monolithic baselines are uncontrolled, and the central claim of the paper is unsupported.
- [§IV-C, Eq. (19)] The benchmark's diagnosis and mitigation ground truth is circular with respect to the treatment being evaluated. Reference responses are generated by Gemini 2.5 Pro from the same injected anomaly metadata, DT outputs, and curated anomaly KB that the evaluated agents receive (Section IV-C). DA and MA quality is then measured as cosine similarity to these references (Eq. 19). Consequently, the KB gains in Table IV and the agentic gains in Table VI may reflect how closely each model imitates the reference generator rather than the correctness of the diagnosis or mitigation. The statement that references were 'reviewed for consistency and factual alignment' is insufficient: the authors should provide human-expert ratings on a random sample, report agreement between the semantic-similarity metric and expert judgments, and, if possible, use independent ground-truth labels such as known fault taxonomies and verified mitigation actions instead of LLM-generated references.
- [§IV-E] The monolithic baseline condition is not a clean control for task decomposition. Each monolithic model receives the complete event, the KB, and the historical repository in a single prompt and is asked to output all three tasks, whereas AgenticTwin's agents each receive only their subtask inputs (Eqs. 16–18). The reported gap could therefore be due to prompt length, output-format constraints, or the particular choice of monolithic models (selected from Tables IV and V) rather than to task specialization itself. At minimum, the authors should include a monolithic baseline that receives the same decomposed inputs in a pipeline of separate calls, and they should report per-task scores for the monolithic models under the Table VI protocol.
- [§IV-D, Tables IV–VI] All central comparisons are reported as point estimates with no variance or significance testing. For example, the retrieval differences among medium models in Table V (MRR 0.83–0.91) may overlap under sampling variability, and the KB gains in Table IV have no confidence intervals. The paper should report standard deviations across repeated runs or query subsets and apply a significance test to the key claims: the KB effect on diagnosis and mitigation, and the agentic-versus-monolithic difference.
minor comments (6)
- [§III-E, Eq. (19)] Specify the sentence embedding function f(·), including the model name, version, pooling method, and whether the same embedding model is used for all evaluated LLM outputs.
- [§III-A] The paper repeatedly mentions 'physics-guided rules' and 'physics-based constraints,' but no equations or implementation details are provided; the DT is defined only through the regression loss in Eq. (5). Please either specify these constraints or soften the claim.
- [§IV-A] Report the anomaly injection parameters (magnitudes, durations, affected sensors, and rules for train/test contamination) so that the benchmark is reproducible; currently only the four fault formulas in Section III-C are given.
- [Tables II and IV] Some table labels are unclear: Table II's header 'Models LR Precision Recall F1-score' appears to use LR for learning rate, and Table IV's G% is not defined in the caption. Please clarify these captions.
- [§IV-C] Clarify the relationship between the 12,000 synthetic queries and the 6,999 test samples; in particular, state the number of distinct anomaly events used for each agent's evaluation and whether multiple queries are generated per event.
- [§IV-E, Table VI] The notation '18B assigned parameters' conflates total parameter count with inference-time compute; clarify whether the three agents run sequentially, and whether the 70B baseline uses quantization or other memory optimizations.
Circularity Check
Benchmark references for diagnosis and mitigation are generated from the same knowledge base that is the experimental treatment, so the reported KB gain is partly built into the scoring metric.
-
self definitional
[Section IV-C (Benchmark Query Generation), score basis for Tables IV and VI and the Section V conclusion]
"Each event is represented using its observed sensor values, DT predictions, residual features, anomaly labels, and fault metadata. These structured events are provided to an LLM through task-specific prompt templates that instruct it to generate realistic operator questions and corresponding reference responses. ... For each query, Gemini 2.5 Pro also generates task-specific reference responses, using the injected anomaly metadata, DT outputs, and curated anomaly knowledge base."
The gold-standard responses for the DA and MA are generated by Gemini 2.5 Pro from the same anomaly event metadata, DT outputs, and curated knowledge base that the evaluated agents receive in the KB condition. Performance is then defined as cosine similarity to those references. The KB ablation therefore compares an agent with KB against a reference that itself contains KB-derived content, and the same agent without KB against that same KB-containing reference. The with-KB response shares vocabulary and factual content with the reference by construction, while the without-KB response cannot, so the Table IV gains (average DA 22.3%, MA 26.3%) partly measure imitation of the reference generator rather than independently verified diagnostic or mitigation quality.
full rationale
The framework's DT and anomaly classifier are evaluated against injected ground-truth labels in Tables II-III, and the RA is scored by Precision@k, Recall@k, and MRR against the actual anomaly repository in Table V, so those components retain independent content and are not circular. The one significant circularity is the LLM-generated reference used for DA and MA semantic-similarity scoring: the reference generator is given the same KB and event metadata as the treated agents, making part of the measured KB gain an artifact of shared inputs rather than of verified correctness. This affects the paper's central KB-ablation claim, though not the classification or retrieval results. I did not count the self-citation [6] as load-bearing because it supports only background context about digital twins. I also flag, but do not score as circularity, that Table VI's agentic component scores appear inconsistent with the component tables (e.g., Qwen 2.5 3B Instruct has diagnosis similarity 0.84 with KB in Table IV but 0.93 as the DA in Table VI despite Eq. 16 giving the DA the same E_t and KB inputs); that is a reproducibility and correctness concern rather than a reduction-to-inputs circularity. Overall score 6: one important evaluation claim is partially forced by construction, while other contributions remain independently grounded.
Assumptions & free parameters
free parameters (2)
- Anomaly injection parameters =
a (drift rate), c (stuck-at value), Delta-t (replay lag), n(t) (spike noise)
- DT regression parameters theta =
Learned from the Jena training split
assumptions (4)
- domain assumption The Beutenberg weather sensor stream is a valid proxy for a cyber-physical system.
- ad hoc to paper Gemini 2.5 Pro generated reference responses are correct.
- ad hoc to paper Cosine similarity of sentence embeddings measures diagnosis and mitigation quality.
- domain assumption Physics-guided rules in the digital twin are correct.
Cite this review
Pith. "Pith review of AgenticTwin: An Agentic LLM Framework Integrated with Digital Twin for Anomaly Detection." pith.science (2026). https://pith.science/paper/JJ3I3YTT
@misc{pith2026260811679,
author = {Pith},
title = {Pith review of: AgenticTwin: An Agentic LLM Framework Integrated with Digital Twin for Anomaly Detection},
year = {2026},
howpublished = {\url{https://pith.science/paper/JJ3I3YTT}},
note = {Machine review of arXiv:2608.11679}
}
read the original abstract
Digital twins are increasingly used to monitor and simulate the behavior of cyber-physical systems. Even with skilled operators, interpreting anomalies detected within digital twin pipelines is challenging, as the sheer complexity and volume of raw sensor data make thorough analysis difficult. Recent advances in large language models (LLMs) offer promising capabilities for reasoning and explanation, yet their integration into digital twin-driven anomaly analysis remains underexplored. In this work, we propose AgenticTwin, an agentic framework that integrates LLM-driven reasoning with a digital twin-based anomaly detection pipeline. The framework grounds LLM-generated explanations in outputs from a digital twin-driven anomaly classifier and enables human operators to ask relevant natural-language questions about the system. Beyond the framework itself, we introduce a benchmark-oriented evaluation pipeline constructed over synthetic anomalies injected into a real-world weather sensor dataset, enabling controlled generation of operator queries over anomaly events. We further evaluate the feasibility of deploying lightweight, open-source LLMs for practical cyber-physical environments. Experimental results demonstrate that structured agent collaboration and knowledge-grounded reasoning improve diagnosis quality, contextual retrieval, and mitigation quality across diverse possible anomaly scenarios.
Figures
Reference graph
Works this paper leans on
-
[1]
Digital twin: Mitigating unpredictable, undesirable emergent behavior in complex systems,
M. Grieves and J. Vickers, “Digital twin: Mitigating unpredictable, undesirable emergent behavior in complex systems,” in Transdisciplinary perspectives on complex systems: New findings and approaches. Springer, 2016, pp. 85–113
2016
-
[2]
Dts and cps toward smart manufacturing and industry 4.0: Correlation and comparison,
F. Tao, Q. Qi, L. Wang, and A. Nee, “Dts and cps toward smart manufacturing and industry 4.0: Correlation and comparison,” Engineering, vol. 5, no. 4, pp. 653–661, 2019
2019
-
[3]
Dt-driven smart manufacturing: Connotation, reference model, applications and research issues,
Y . Lu, C. Liu, I. Kevin, K. Wang, H. Huang, and X. Xu, “Dt-driven smart manufacturing: Connotation, reference model, applications and research issues,”Robotics and computer-integrated manufacturing, vol. 61, p. 101837, 2020
2020
-
[4]
Review of dt applications in manufacturing,
C. Cimino, E. Negri, and L. Fumagalli, “Review of dt applications in manufacturing,”Computers in industry, vol. 113, p. 103130, 2019
2019
-
[5]
Digital twin for healthcare systems,
A. Vall ´ee, “Digital twin for healthcare systems,”Frontiers in Digital Health, vol. 5, p. 1253050, 2023
2023
-
[6]
Advancing Security with Digital Twins: A Comprehensive Survey
B. Airehenbuwa, T. Hasan, S. Sarkar, and U. Guin, “Advancing security with digital twins: A comprehensive survey,”arXiv preprint arXiv:2505.17310, 2025
work page Pith review arXiv 2025
-
[7]
Explainable ai for cybersecurity automation, intelligence and trustworthiness in digital twin: Methods, taxonomy, challenges and prospects,
I. H. Sarker, H. Janicke, A. Mohsin, A. Gill, and L. Maglaras, “Explainable ai for cybersecurity automation, intelligence and trustworthiness in digital twin: Methods, taxonomy, challenges and prospects,”ICT express, vol. 10, no. 4, pp. 935–958, 2024
2024
-
[8]
Digital twin-driven online anomaly detection for an automation system based on edge intel- ligence,
H. Huang, L. Yang, Y . Wang, X. Xu, and Y . Lu, “Digital twin-driven online anomaly detection for an automation system based on edge intel- ligence,”Journal of Manufacturing Systems, vol. 59, pp. 138–150, 2021
2021
Show all 89 references
-
[9]
Explainable, interpretable, and trustworthy ai for an intelligent digital twin: A case study on remaining useful life,
K. Kobayashi and S. B. Alam, “Explainable, interpretable, and trustworthy ai for an intelligent digital twin: A case study on remaining useful life,”Engineering Applications of Artificial Intelligence, vol. 129, p. 107620, 2024
2024
-
[10]
Digital twin-based anomaly detection with curriculum learning in cyber-physical systems,
Q. Xu, S. Ali, and T. Yue, “Digital twin-based anomaly detection with curriculum learning in cyber-physical systems,”ACM Transactions on Software Engineering and Methodology, vol. 32, no. 5, pp. 1–32, 2023
2023
-
[11]
Digital twins for cyber-physical systems security: State of the art and outlook,
M. Eckhart and A. Ekelhart, “Digital twins for cyber-physical systems security: State of the art and outlook,”Security and Quality in Cyber- Physical Systems Engineering: With Forewords by Robert M. Lee and Tom Gilb, pp. 383–412, 2019
2019
-
[12]
Leveraging llms for enhanced dt modeling: Trends, methods, and challenges,
L. Yang, S. Luo, X. Cheng, and L. Yu, “Leveraging llms for enhanced dt modeling: Trends, methods, and challenges,”arXiv preprint arXiv:2503.02167, 2025. 9
2025 arXiv
-
[13]
On integrating resilience and human oversight into llm-assisted modeling workflows for digital twins,
P. Lekshmi and N. Karanjkar, “On integrating resilience and human oversight into llm-assisted modeling workflows for digital twins,” in Proceedings of the ACM SIGSIM Conference on Principles of Advanced Discrete Simulation (PADS), 2026
2026
-
[14]
Semantic-aware llm orchestration for proactive resource management in predictive digital twin vehicular networks,
S. H. Ahmadpanah, “Semantic-aware llm orchestration for proactive resource management in predictive digital twin vehicular networks,” Computing, vol. 108, no. 1, p. 14, 2026
2026
-
[15]
Llm-based task offloading and resource allocation for dt edge computing networks,
Q. Wu, Y . Xie, P. Fan, D. Qin, K. Wang, N. Cheng, and K. B. Letaief, “Llm-based task offloading and resource allocation for dt edge computing networks,”IEEE Transactions on Mobile Computing, 2026
2026
-
[16]
Agentic-ai framework for integrated design, implementation, testing, and operation of dt networks,
R. Khaldi, A. Lehmann, B. Ghita, and U. Trick, “Agentic-ai framework for integrated design, implementation, testing, and operation of dt networks,”IEEE Open Journal of the Communications Society, 2026
2026
-
[17]
Multi-agent systems for manufacturing digital twins: A perspective on agency and large language models,
N. P. Greis, H. P. Cherukuri, and J. C. Outeiro, “Multi-agent systems for manufacturing digital twins: A perspective on agency and large language models,”IFAC-PapersOnLine, vol. 59, no. 10, pp. 1612–1617, 2025
2025
-
[18]
Llm-enabled multi-agent framework for natural language interaction with graph-based digital twins,
Y . Pan, M. Wang, L. Lu, R. Lamsal, E. P ¨arn, S. Zlatanova, and I. Brilakis, “Llm-enabled multi-agent framework for natural language interaction with graph-based digital twins,”Automation in Construction, vol. 183, p. 106791, 2026
2026
-
[19]
Generative ai-driven digital twin in the manufacturing internet of things: A comprehensive survey,
X. Fu, M. Qin, P. Pace, C. Savaglio, W. Li, and G. Fortino, “Generative ai-driven digital twin in the manufacturing internet of things: A comprehensive survey,”IEEE Internet of Things Journal, 2026
2026
-
[20]
Llm-twin: mini-giant model-driven be- yond 5g dt networking framework with semantic secure communication and computation,
Y . Hong, J. Wu, and R. Morello, “Llm-twin: mini-giant model-driven be- yond 5g dt networking framework with semantic secure communication and computation,”Scientific Reports, vol. 14, no. 1, p. 19065, 2024
2024
-
[21]
Digital twin designs with generative ai: crafting a comprehensive framework for manufacturing systems,
O. Mata, P. Ponce, C. Perez, M. Ramirez, B. Anthony, B. Russel, P. Apte, B. MacCleery, and A. Molina, “Digital twin designs with generative ai: crafting a comprehensive framework for manufacturing systems,”Jour- nal of Intelligent Manufacturing, vol. 37, no. 3, pp. 1049–1072, 2026
2026
-
[22]
Enabling technologies and tools for digital twin,
Q. Qi, F. Tao, T. Hu, N. Anwer, A. Liu, Y . Wei, L. Wang, and A. Y . Nee, “Enabling technologies and tools for digital twin,”Journal of manufacturing systems, vol. 58, pp. 3–21, 2021
2021
-
[23]
Digital twin paradigm: A systematic literature review,
C. Semeraro, M. Lezoche, H. Panetto, and M. Dassisti, “Digital twin paradigm: A systematic literature review,”Computers in industry, vol. 130, p. 103469, 2021
2021
-
[24]
Leveraging llm agents and dts for fault handling in process plants,
M. S. Gill, J. Vyas, A. Markaj, F. Gehlhoff, and M. Mercang ¨oz, “Leveraging llm agents and dts for fault handling in process plants,” in 30th International Conference on Emerging Technologies and Factory Automation (ETFA). IEEE, 2025, pp. 1–8
2025
-
[25]
Predictive dt for offshore wind farms,
A. Haghshenas, A. Hasan, O. Osen, and E. T. Mikalsen, “Predictive dt for offshore wind farms,”Energy Informatics, vol. 6, no. 1, p. 1, 2023
2023
-
[26]
Dt, physics-based model, and ml applied to damage detection in structures,
T. Ritto and F. Rochinha, “Dt, physics-based model, and ml applied to damage detection in structures,”Mechanical Systems and Signal Processing, vol. 155, p. 107614, 2021
2021
-
[27]
The advance of digital twin for predictive maintenance: The role and function of machine learn- ing,
C. Chen, H. Fu, Y . Zheng, F. Tao, and Y . Liu, “The advance of digital twin for predictive maintenance: The role and function of machine learn- ing,”Journal of Manufacturing Systems, vol. 71, pp. 581–594, 2023
2023
-
[28]
Methodology for enabling digital twin using advanced physics-based modelling in predictive maintenance,
P. Aivaliotis, K. Georgoulias, Z. Arkouli, and S. Makris, “Methodology for enabling digital twin using advanced physics-based modelling in predictive maintenance,”Procedia Cirp, vol. 81, pp. 417–422, 2019
2019
-
[29]
Anomaly detection using ml and adopted dt concepts in radio environments,
M. H. Moharam, O. Hany, A. Hany, A. Mahmoud, M. Mohamed, and S. Saeed, “Anomaly detection using ml and adopted dt concepts in radio environments,”Scientific Reports, vol. 15, no. 1, p. 18352, 2025
2025
-
[30]
Sgad-gan: Simultaneous generation and anomaly detection for time- series sensor data with generative adversarial networks,
P. Zhao, Z. Ding, Y . Li, X. Zhang, Y . Zhao, H. Wang, and Y . Yang, “Sgad-gan: Simultaneous generation and anomaly detection for time- series sensor data with generative adversarial networks,”Mechanical Systems and Signal Processing, vol. 210, p. 111141, 2024
2024
-
[31]
Tab: Unified benchmarking of time series anomaly detection methods,
X. Qiu, Z. Li, W. Qiu, S. Hu, L. Zhou, X. Wu, Z. Li, C. Guo, A. Zhou, Z. Shenget al., “Tab: Unified benchmarking of time series anomaly detection methods,”arXiv preprint arXiv:2506.18046, 2025
2025 arXiv
-
[32]
A digital twin-based platform for medical cyber-physical systems,
M. Rahim, W. Lalouani, E. Toubal, and L. Emokpae, “A digital twin-based platform for medical cyber-physical systems,”IEEE Access, vol. 12, pp. 174 591–174 607, 2024
2024
-
[33]
Digital-twin-based cps anomaly diagnosis and security defense countermeasure recommendation,
J. Ma, Y . Guo, C. Fang, and Q. Zhang, “Digital-twin-based cps anomaly diagnosis and security defense countermeasure recommendation,”IEEE Internet of Things Journal, vol. 11, no. 10, pp. 18 726–18 738, 2024
2024
-
[34]
Weather-driven agricultural decision- making under imperfect conditions,
T. Ahmed and M. Hasan, “Weather-driven agricultural decision- making under imperfect conditions,” inProceedings of the 33rd ACM International Conference on Advances in Geographic Information Systems, 2025, pp. 927–930
2025
-
[35]
An architecture for integrating large language models with digital twins and automation systems,
Y . Xia, N. Jazdi, and M. Weyrich, “An architecture for integrating large language models with digital twins and automation systems,” inIEEE 30th International Conference on Emerging Technologies and Factory Automation (ETFA). IEEE, 2025, pp. 1–8
2025
-
[36]
From simulation to prediction: Enhancing digital twins with advanced generative ai technologies,
Y . Huang, J. Zhang, X. Chen, A. H. Lam, and B. M. Chen, “From simulation to prediction: Enhancing digital twins with advanced generative ai technologies,” inIEEE 18th International Conference on Control & Automation (ICCA). IEEE, 2024, pp. 490–495
2024
-
[37]
Defecttwin: When llm meets dt for railway defect inspection,
R. Ferdousi, M. A. Hossain, C. Yang, and A. El Saddik, “Defecttwin: When llm meets dt for railway defect inspection,”IEEE Access, 2026
2026
-
[38]
Lsdts: Llm-augmented semantic digital twins for adaptive knowledge-intensive infrastructure planning,
N. Li, Z. Ma, R. Yu, and L. Li, “Lsdts: Llm-augmented semantic digital twins for adaptive knowledge-intensive infrastructure planning,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 40, no. 45, 2026, pp. 38 871–38 879
2026
-
[39]
Systematic integration of digital twins and constrained llms for interpretable cyber- physical anomaly detection,
K. E. Kampourakis, V . Gkioulos, and S. Katsikas, “Systematic integration of digital twins and constrained llms for interpretable cyber- physical anomaly detection,” inIFIP International Conference on ICT Systems Security and Privacy Protection. Springer, 2026, pp. 199–212
2026
-
[40]
A survey on applica- tions of large language model-driven digital twins for intelligent network optimization,
Z. Guo, F. Tang, L. Luo, M. Zhao, and N. Kato, “A survey on applica- tions of large language model-driven digital twins for intelligent network optimization,”IEEE Communications Surveys & Tutorials, 2025
2025
-
[41]
Llms for explainable decisions in dynamic dts,
N. Zhang, C. Vergara-Marcillo, G. Diamantopoulos, J. Shen, N. Tziritas, R. Bahsoon, and G. Theodoropoulos, “Llms for explainable decisions in dynamic dts,” inInternational Conference on Dynamic Data Driven Applications Systems. Springer, 2024, pp. 81–89
2024
-
[42]
Fcllm-dt: Enpowering federated continual learning with large language models for digital-twin-based industrial iot,
Y . Xia, Y . Chen, Y . Zhao, L. Kuang, X. Liu, J. Hu, and Z. Liu, “Fcllm-dt: Enpowering federated continual learning with large language models for digital-twin-based industrial iot,”IEEE Internet of Things Journal, vol. 12, no. 6, pp. 6070–6081, 2024
2024
-
[43]
Iiot-enabled digital twin for legacy and smart factory machines with llm integration,
A. Gautam, M. R. Aryal, S. Deshpande, S. Padalkar, M. Nikolaenko, M. Tang, and S. Anand, “Iiot-enabled digital twin for legacy and smart factory machines with llm integration,”Journal of Manufacturing Systems, vol. 80, pp. 511–523, 2025
2025
-
[44]
Large language models forecast patient health trajectories enabling digital twins,
N. Makarov, M. Bordukova, P. Quengdaeng, D. Garger, R. Rodriguez- Esteban, F. Schmich, and M. P. Menden, “Large language models forecast patient health trajectories enabling digital twins,”npj Digital Medicine, vol. 8, no. 1, p. 588, 2025
2025
-
[45]
A multi-agent digital twin framework for ai-driven fitness coaching,
M. Vahdati, K. Gholizadeh HamlAbadi, F. Laamarti, and A. El Saddik, “A multi-agent digital twin framework for ai-driven fitness coaching,” inProceedings of the ACM International Conference on Interactive Media Experiences, 2025, pp. 380–385
2025
-
[46]
Llmasmmkg: Llm assisted synthetic multi-modal knowledge graph creation for smart city cognitive digital twins,
S. Mandal and N. E. O’Connor, “Llmasmmkg: Llm assisted synthetic multi-modal knowledge graph creation for smart city cognitive digital twins,” inProceedings of the AAAI Symposium Series, vol. 4, no. 1, 2024, pp. 210–221
2024
-
[47]
Towards fully automated city operations: integrating agentic ai with urban digital twins,
H. Xu, S. Zlatanova, X. Li, M. Wachowicz, and M. Batty, “Towards fully automated city operations: integrating agentic ai with urban digital twins,”Computers, Environment and Urban Systems, vol. 128, p. 102449, 2026
2026
-
[48]
Virtual cities: from dts to autonomous ai societies,
A. Nechesov, I. Dorokhov, and J. Ruponen, “Virtual cities: from dts to autonomous ai societies,”IEEE access, vol. 13, pp. 13 866–13 903, 2025
2025
-
[49]
Integrating llms and digital twins for adaptive multi-robot task allocation in construction,
M. Deng, B. Fu, L. Li, and X. Wang, “Integrating llms and digital twins for adaptive multi-robot task allocation in construction,”IEEE Transactions on Automation Science and Engineering, 2026
2026
-
[50]
An llm-based digital twin for optimizing human-in-the loop systems,
H. Yang, M. Siew, and C. Joe-Wong, “An llm-based digital twin for optimizing human-in-the loop systems,” inIEEE International Workshop on Foundation Models for Cyber-Physical Systems & Internet of Things (FMSys). IEEE, 2024, pp. 26–31
2024
-
[51]
Simbench: A framework for evaluating and diagnosing llm-based digital-twin generation for multi-physics simulation,
J. Wang, A. Negrut, H. Wang, H. Zhang, and D. Negrut, “Simbench: A framework for evaluating and diagnosing llm-based digital-twin generation for multi-physics simulation,”IEEE Access, 2026
2026
-
[52]
Generative ai and digital twin integrated intelligent process planning: a conceptual framework,
Q. Xu, G. Zhou, C. Zhang, F. Chang, Y . Cao, and D. Zhao, “Generative ai and digital twin integrated intelligent process planning: a conceptual framework,”preprint https://doi.org/10.21203/rs.3.rs-3652246/v1, 2023
2023 doi
-
[53]
Empowering digital twins with large language models for global temporal feature learning,
Y . Sun, Q. Zhang, J. Bao, Y . Lu, and S. Liu, “Empowering digital twins with large language models for global temporal feature learning,” Journal of Manufacturing Systems, vol. 74, pp. 83–99, 2024
2024
-
[54]
Future factories with 6g: Agentic ai and cyber-physical digital twins,
H. Li, H. Madhukumar, N. Methley, X. Chen, Y . Wu, J. Parra-Ullauri, V . Sharma, J. Lee, A. R. Koblitz, M. Andrewset al., “Future factories with 6g: Agentic ai and cyber-physical digital twins,”IEEE Internet of Things Journal, 2025
2025
-
[55]
Sr-dtma: A digital twin-driven llm multi-agent framework for systemic risk simulation and coordinated decision- making in supply chains,
Y . Ling and W. Liu, “Sr-dtma: A digital twin-driven llm multi-agent framework for systemic risk simulation and coordinated decision- making in supply chains,”Journal of Computer, Signal, and System Research, vol. 3, no. 1, pp. 123–132, 2026
2026
-
[56]
The role of multi-agents in digital twin implementation: Short survey,
Y . Kalyani and R. Collier, “The role of multi-agents in digital twin implementation: Short survey,”ACM Computing Surveys, vol. 57, no. 3, pp. 1–15, 2024
2024
-
[57]
A digital twin framework for enhancing human–agentic ai–machine collaboration,
B. Nicoletti and A. Appolloni, “A digital twin framework for enhancing human–agentic ai–machine collaboration,”Journal of Intelligent Manufacturing, pp. 1–17, 2025
2025
-
[58]
Integrating agentic ai and digital twins for intelligent decision-making systems,
A. Hasan and D. T. Nguyen, “Integrating agentic ai and digital twins for intelligent decision-making systems,”Array, p. 100721, 2026
2026
-
[59]
Llm experiments with simulation: Llm multi-agent system for simulation model parametrization in dts,
Y . Xia, D. Dittler, N. Jazdi, H. Chen, and M. Weyrich, “Llm experiments with simulation: Llm multi-agent system for simulation model parametrization in dts,” in29th International Conference on Emerging Technologies and Factory Automation (ETFA). IEEE, 2024, pp. 1–4
2024
-
[60]
Autonomous industrial control using an agentic framework with large language models,
J. Vyas and M. Mercang ¨oz, “Autonomous industrial control using an agentic framework with large language models,”IFAC-PapersOnLine, vol. 59, no. 6, pp. 349–354, 2025. 10
2025
-
[61]
Agentic ai for digital twin,
A. Timms, A. Langbridge, A. Antonopoulos, A. Mygiakis, E. V oulgari, and F. O’Donncha, “Agentic ai for digital twin,” inProceedings of the AAAI Conference on Artificial Intelligence, vol. 39, no. 28, 2025, pp. 29 703–29 705
2025
-
[62]
State estimation of a physical system with unknown governing equations,
K. Course and P. B. Nair, “State estimation of a physical system with unknown governing equations,”Nature, vol. 622, no. 7982, pp. 261–267, 2023
2023
-
[63]
Beutenberg weather station,
Max Planck Institute for Biogeochemistry, “Beutenberg weather station,” https://weather.bgc-jena.mpg.de/weather data.html, accessed: 10-01-25
-
[64]
Forecasting at scale,
S. J. Taylor and B. Letham, “Forecasting at scale,”The American Statistician, vol. 72, no. 1, pp. 37–45, 2018
2018
-
[65]
Temporal fusion transformers for interpretable multi-horizon time series forecasting,
B. Lim, S. ¨O. Arık, N. Loeff, and T. Pfister, “Temporal fusion transformers for interpretable multi-horizon time series forecasting,” International journal of forecasting, vol. 37, no. 4, pp. 1748–1764, 2021
2021
-
[66]
Physics encoded blocks in residual neural network architectures for digital twin models,
M. S. Zia, C. Houpert, A. Anjum, L. Liu, A. Conway, and A. Pe ˜na- Rios, “Physics encoded blocks in residual neural network architectures for digital twin models,”Machine Learning, vol. 114, no. 8, p. 180, 2025
2025
-
[67]
Physics-informed neural network solution of point kinetics equations for a nuclear reactor digital twin,
K. Prantikos, L. H. Tsoukalas, and A. Heifetz, “Physics-informed neural network solution of point kinetics equations for a nuclear reactor digital twin,”Energies, vol. 15, no. 20, p. 7697, 2022
2022
-
[68]
A comparison of residual-based methods on fault detection,
C.-C. Hsu, G. Frusque, and O. Fink, “A comparison of residual-based methods on fault detection,”arXiv preprint arXiv:2309.02274, 2023
2023 arXiv
-
[69]
A deep neural network for unsupervised anomaly detection and diagnosis in multivariate time series data,
C. Zhang, D. Song, Y . Chen, X. Feng, C. Lumezanu, W. Cheng, J. Ni, B. Zong, H. Chen, and N. V . Chawla, “A deep neural network for unsupervised anomaly detection and diagnosis in multivariate time series data,” inProceedings of the AAAI conference on artificial intelligence, ...
2019
-
[70]
A distributed sensor-fault detection and diagnosis framework using machine learning,
S. U. Jan, Y . D. Lee, and I. S. Koo, “A distributed sensor-fault detection and diagnosis framework using machine learning,”Information Sciences, vol. 547, pp. 777–796, 2021
2021
-
[71]
Cafd: Context-aware fault diagnostic scheme towards sensor faults utilizing machine learning,
U. Saeed, Y .-D. Lee, S. U. Jan, and I. Koo, “Cafd: Context-aware fault diagnostic scheme towards sensor faults utilizing machine learning,” Sensors, vol. 21, no. 2, p. 617, 2021
2021
-
[72]
MITRE,https://attack.mitre.org/techniques/ics/, 2026
2026
-
[73]
Documentation of the weather station on top of the roof of the institute building of the max-planck-institute for biogeochemistry,
O. Kolle, “Documentation of the weather station on top of the roof of the institute building of the max-planck-institute for biogeochemistry,” Recuperado de https://www. bgc-jena. mpg. de/wetter, 2008
2008
-
[74]
Support-vector networks,
C. Cortes and V . Vapnik, “Support-vector networks,”Machine learning, vol. 20, no. 3, pp. 273–297, 1995
1995
-
[75]
Random decision forests,
T. K. Ho, “Random decision forests,” inProceedings of 3rd international conference on document analysis and recognition, vol. 1. IEEE, 1995, pp. 278–282
1995
-
[76]
Xgboost: A scalable tree boosting system,
T. Chen and C. Guestrin, “Xgboost: A scalable tree boosting system,” inProceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining, 2016, pp. 785–794
2016
-
[77]
Lightgbm: A highly efficient gradient boosting decision tree,
G. Ke, Q. Meng, T. Finley, T. Wang, W. Chen, W. Ma, Q. Ye, and T.-Y . Liu, “Lightgbm: A highly efficient gradient boosting decision tree,” Advances in neural information processing systems, vol. 30, 2017
2017
-
[78]
Multilayer perceptrons for classification and regression,
F. Murtagh, “Multilayer perceptrons for classification and regression,” Neurocomputing, vol. 2, no. 5-6, pp. 183–197, 1991
1991
-
[79]
Enhancing the locality and breaking the memory bottleneck of transformer on time series forecasting,
S. Li, X. Jin, Y . Xuan, X. Zhou, W. Chen, Y .-X. Wang, and X. Yan, “Enhancing the locality and breaking the memory bottleneck of transformer on time series forecasting,”Advances in neural information processing systems, vol. 32, 2019
2019
-
[80]
A time series is worth 64 words: Long-term forecasting with transformers,
Y . Nie, N. H. Nguyen, P. Sinthong, and J. Kalagnanam, “A time series is worth 64 words: Long-term forecasting with transformers,”arXiv preprint arXiv:2211.14730, 2022
2022 arXiv
-
[81]
Timer-s1: A billion-scale time series foundation model with serial scaling,
Y . Liu, X. Su, S. Wang, H. Zhang, H. Liu, Y . Wang, Z. Ye, Y . Xiang, J. Wang, and M. Long, “Timer-s1: A billion-scale time series foundation model with serial scaling,”arXiv preprint arXiv:2603.04791, 2026
2026 arXiv
-
[82]
Timesnet: Temporal 2d-variation modeling for general time series analysis,
H. Wu, T. Hu, Y . Liu, H. Zhou, J. Wang, and M. Long, “Timesnet: Temporal 2d-variation modeling for general time series analysis,”arXiv preprint arXiv:2210.02186, 2022
2022 arXiv
-
[83]
Gemini 2.5: Pushing the frontier with advanced reasoning, multimodality, long context, and next generation agentic capabilities,
G. Comanici, E. Bieberet al., “Gemini 2.5: Pushing the frontier with advanced reasoning, multimodality, long context, and next generation agentic capabilities,”arXiv preprint arXiv:2507.06261, 2025
2025 arXiv
-
[84]
Qwen3 technical report,
A. Yang, A. Li, B. Yang, B. Zhang, B. Hui, B. Zheng, B. Yu, C. Gao, C. Huang, C. Lvet al., “Qwen3 technical report,”arXiv preprint arXiv:2505.09388, 2025
2025 arXiv
-
[85]
The llama 3 herd of models,
A. Grattafiori, A. Dubey, A. Jauhri, A. Pandey, A. Kadian, A. Al-Dahle, A. Letman, A. Mathur, A. Schelten, A. Vaughanet al., “The llama 3 herd of models,”arXiv preprint arXiv:2407.21783, 2024
2024 arXiv
-
[86]
Gemma: Open models based on gemini research and technology,
G. Team, T. Mesnard, C. Hardin, R. Dadashi, S. Bhupatiraju, S. Pathak, L. Sifre, M. Rivi `ere, M. S. Kale, J. Loveet al., “Gemma: Open models based on gemini research and technology,”arXiv preprint arXiv:2403.08295, 2024
2024 arXiv
-
[87]
Phi-4 technical report,
M. Abdin, J. Aneja, H. Behl, S. Bubeck, R. Eldan, S. Gunasekar, M. Harrison, R. J. Hewett, M. Javaheripi, P. Kauffmannet al., “Phi-4 technical report,”arXiv preprint arXiv:2412.08905, 2024
2024 arXiv
-
[88]
Deepseek-r1: Incentivizing reasoning capability in llms via rl,
D. Guo, D. Yang, H. Zhang, J. Song, P. Wang, Q. Zhu, R. Xu, R. Zhang, S. Ma, X. Biet al., “Deepseek-r1: Incentivizing reasoning capability in llms via rl,”arXiv preprint arXiv:2501.12948, 2025
2025 arXiv
-
[89]
Openai gpt-5 system card,
A. Singh, A. Fry, A. Perelman, A. Tart, A. Ganesh, A. El-Kishky, A. McLaughlin, A. Low, A. Ostrow, A. Ananthramet al., “Openai gpt-5 system card,”arXiv preprint arXiv:2601.03267, 2025. Touseef Hasan (S’25)is pursuing his Ph.D. from the Department of Electrical and Computer Eng...
2025 arXiv
Reviewed August 16, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.