Pith. sign in

REVIEW 3 major objections 6 minor 63 references

Variational Digital Twins

T0 review · 3 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A single Bayesian output layer, added to an ordinary neural surrogate, is enough to make a digital twin that updates in seconds, reports calibrated uncertainty, and learns from far fewer experiments.

desk verdict Useful applied benchmark of Bayesian-last-layer digital twins, but the central calibration claim is asserted and never measured—send it to review and make them quantify coverage. read the letter →

arxiv 2507.01047 v1 pith:DFC6U7TU submitted 2025-06-25 cs.LG cs.SYeess.SY

Logan A. Burnett , Umme Mahbuba Nabila , Majdi I. Radaideh This is my paper · ORCID
classification cs.LGcs.SYeess.SY
keywords digitaltwinsvariationalinferenceBayesianlastlayeruncertaintyquantificationactivelearningenergysystemsrenewablepowerforecastingphysics-informedneuralnetworks
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper sets out to close the gap between digital twins as concepts and digital twins as deployable tools by proposing the variational digital twin (VDT): a single Bayesian output layer attached to an otherwise standard neural network is enough to give the twin calibrated uncertainty, fast Bayesian updating, and data-efficient learning. It demonstrates the idea on four energy problems: critical-heat-flux prediction where uncertainty-guided active learning reaches $R^2 = 0.98$ with 47 percent fewer experiments; three-year solar and wind forecasting maintained by monthly session updates; reactor thermocouple reconstruction that survives 50 percent sensor loss; and Li-ion battery voltage prediction whose rolling updates lower mean-squared error by an order of magnitude versus the best static model. If the claim holds, operators could turn existing surrogate models into self-updating, uncertainty-reporting twins without redesigning architectures or buying specialized hardware.

What carries the argument

The carrying mechanism is the variational Bayesian last layer, a single stochastic linear layer whose weights are drawn from a learned mean-field variational posterior; at inference, multiple stochastic forward passes yield a predictive mean and 95 percent credible intervals. It is paired with a session-based update scheme that re-initializes from previous weights and refits on only the most recent month or block of data, along with an aided active learning protocol that first random-samples a candidate pool and then queries by predictive uncertainty. The paper also contributes a structured sensor-concatenation algorithm that interleaves sensor signals by mean-temperature bins so that recurrent models see quasi-periodic structure. These pieces, rather than architectural novelty, carry the argument.

What would settle it

Compute empirical coverage of the 95 percent credible intervals on held-out test data for one or more of the four case studies (for example, HTTF test temperatures or battery discharge voltages); if far fewer than 95 percent of test points fall inside the intervals, the calibration claim and the active-learning ranking built on predictive variance would fail. A second check would rerun the aided active learning comparison against random sampling under identical compute budgets to see whether the 47 percent sample savings reproduces.

Watch

Extended reading notes

Core claim

The paper's central claim is that augmenting a conventional neural backbone (feedforward, recurrent, or physics-informed) with one variational Bayesian linear output layer, trained by maximizing an evidence lower bound and updated by a session-based algorithm, produces a practical digital twin: one that assimilates new sensor data in seconds on commodity GPUs, reports predictive intervals that track its own confidence, and supports active learning and degraded-instrumentation operation. In the four case studies, the VDT reaches $R^2 = 0.98$ for critical heat flux with 47 percent fewer training samples than random sampling and one-third the training time; keeps solar $R^2$ above 0.95 across three years of monthly updates; reconstructs HTTF solid and fluid temperatures with $R^2$ above 0.99 after half the sensors are withheld; and lowers battery voltage mean-squared error by an order of magnitude relative to static training while widening credible intervals as the cell ages. The intended consequence is a generalizable inverse-loop framework for digital twins, with the forward and generative loops left for future work.

Load-bearing premise

The load-bearing premise is that the 95 percent predictive intervals produced by the Bayesian output layer are genuinely calibrated; the paper uses those intervals to rank experiments, to report model confidence, and to claim reliability, but it never measures whether the stated intervals cover the true values at the claimed rate.

Editorial extensions

If this is right

  • Active learning driven by the twin's predictive variance can cut experimental data requirements by roughly half while reaching the same accuracy target.
  • A twin updated on one month of data at a time can track three years of renewable-generation drift and keep solar $R^2$ above 0.95.
  • A recurrent variational twin can reconstruct reactor temperature fields with $R^2$ above 0.99 even when half the sensors are missing, supporting graceful degradation of instrumentation.
  • Rolling variational updates keep a physics-informed battery twin accurate through aging, with voltage mean-squared error about an order of magnitude below static models and intervals that widen near end of life.
  • The approach is architecture-agnostic: the same Bayesian last layer works on feedforward, recurrent, and physics-informed backbones.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A natural extension the paper leaves implicit is applying the same Bayesian last layer to other safety-critical assets, such as structural health monitoring or medical devices, where predictive intervals could gate alarms and retraining decisions.
  • One testable extension is to vary the retraining cadence (monthly versus weekly versus event-triggered) to find the minimum update frequency that keeps error and interval coverage within tolerance.
  • The paper's own finding that static models retain long-memory for wind variability suggests a hybrid twin that retrains only when a drift detector fires, rather than on a fixed schedule.
  • A reader who wants to deploy the framework should first check empirical coverage of the 95 percent intervals, because the active-learning ranking and reliability statements are only as good as the calibration of those intervals.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. This paper proposes a Variational Digital Twin (VDT) framework that attaches a Bayesian output layer to standard neural backbones (FNN, RNN/LSTM/GRU, and PINN) and updates the model incrementally as new data arrive. The method is demonstrated on four energy-sector applications: critical-heat-flux prediction with active learning, three-year renewable forecasting with monthly updates, reactor thermocouple reconstruction under sensor loss, and Li-ion battery voltage prediction with rolling updates. Headline claims include calibrated uncertainty bounds, R2=0.98 with 47% fewer CHF experiments, R2>0.99 after 50% sensor loss, and battery MSE reduced by an order of magnitude relative to static training.

Significance. If the central claims hold, the paper offers a useful practical contribution: a computationally cheap way to add uncertainty quantification and online updating to otherwise standard architectures, validated on four realistic energy benchmarks. The application breadth is a strength, and the authors state that code and data will be made public, which supports reproducibility. However, the strongest claims—calibrated uncertainty and the magnitude of the battery improvement—are not supported by the evidence as currently reported. The core methodology is not entirely new (variational last layers exist), but the digital-twin framing and the session-based updating scheme provide a useful integration study.

major comments (3)
  1. [Abstract; §3.1; §5.2; §6.1; §7] The paper repeatedly claims that the VDT produces 'calibrated' uncertainty bounds, but no calibration metric is reported anywhere. The predictive intervals are defined only operationally in Eqs. (3)–(4) as quantiles of S stochastic forward passes; there is no empirical coverage probability, predicted-interval coverage probability (PICP), expected calibration error (ECE), CRPS, or any other calibration check for any of the four applications. This matters because predictive variance is used both to drive active learning (§3.2, §5.1) and to support reliability statements (§6.1, §7). The paper itself concedes in §5.4 that variational-based models may struggle to calibrate uncertainty when trained on insufficient data—a regime that applies directly to the CHF active-learning start (10 samples) and the battery updates (10 discharges per session). Please add explicit calibration evaluation for each application and report the value of S used to form the intervals.
  2. [Abstract; §6.3; §7; Table 7] The abstract, §6.3, and §7 state that the battery VDT lowers voltage mean-squared error 'by an order of magnitude' relative to the best static model. Table 7 shows the best static MSE is 0.022707 V² (300 runs) and the VDT update MSE is 0.004535 V², a factor of 5.0, not 10. Please correct this claim or provide additional results that genuinely reach an order-of-magnitude improvement.
  3. [Table 4; §5.1; Figure 6] The CHF active-learning result is based on 50 trials, but Table 4 reports only point values for training size and time, and Figure 6 shows mean R2 curves without error bars or variability shading. The text claims AAL exhibits 'reduced variability' relative to random sampling, which is not verifiable without trial-to-trial statistics. Report mean ± standard deviation (or a box plot) for the number of samples and wall-clock time needed to reach the R2=0.98 threshold, and specify how the threshold is applied across the 50 trials.
minor comments (6)
  1. [§3.1, Eq. (4)] 'Confidence intervals' should be 'credible intervals', since these are posterior predictive quantiles, not frequentist confidence intervals.
  2. [§3.1, §5.2, Table 3] The number of stochastic forward passes S used in Eqs. (3)–(4) is never reported for any application; please state the value used for each experiment.
  3. [§6.4] The statement that variational last layers add 'only two trainable parameters per output unit' is not correct for a Bayesian linear layer; the parameter count is 2·(D+1) per output unit for a D-dimensional input layer. Please reword to reflect the actual parameter overhead.
  4. [§3.4, Algorithm 2] Line 12 of Algorithm 2 reads 'Append front element of Lk to and remove it'—the target list O is missing. Also, the choice of K=10 bins and the zig-zag ordering is presented as a key design decision but no ablation or justification is given beyond a visual inspection of Figure 10.
  5. [§5.3, Figure 12] The caption says '60 removed indicating training on only one signal', but the dataset contains 310 sensors (155 solid and 155 fluid). Clarify the x-axis units and how removing 60 sensors corresponds to training on one signal.
  6. [Table 3; §5.4] Minor typos: 'gird search' should be 'grid search'; 'robustivity' should be 'robustness'; 'local minimas' should be 'local minima'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; four case studies benchmark VDT against external baselines, and the calibration concern is an unverified empirical claim rather than a derivation that reduces to its inputs.

full rationale

The paper's derivation chain is self-contained in the sense that the central results—active-learning sample savings, session-update accuracy retention, sensor-loss robustness, and rolling-update battery gains—are empirical comparisons against external baselines (random sampling, deterministic GRU/LSTM, static training) and are not entailed by construction. The Bayesian last layer is an existing published method [49], the ELBO update is standard variational inference, and the session-based updating (Algorithm 1) is an explicit retraining scheme that is evaluated rather than assumed. Self-citations appear in the literature review and in references to the authors' prior UQ/control work, but none of these supplies the evidence for the paper's headline claims. The only concern raised by the skeptical review—that 'calibrated uncertainty bounds' are never checked with coverage/CRPS/ECE metrics, and that Section 5.4 admits variational models can struggle to calibrate under data scarcity—is a correctness/validation gap, not a circularity: the quantile intervals in Eq. (4) are defined, and the word 'calibrated' is asserted, but no equation or fitted parameter makes the assertion true by construction. Thus the paper has no significant circularity under the specified tests.

Assumptions & free parameters 8 free parameters · 7 assumptions · 0 invented entities

The central method uses a known variational last layer; the main free choices are the ELBO KL weight, the AAL sampling schedule and stopping threshold, the sensor binning order, and update cadences. The paper builds on standard variational inference and on the BattNN equivalent-circuit model; no new physical entities are postulated. The aided-active-learning design and the zig-zag sensor ordering are ad hoc to this paper. The novel updating algorithm is warm-start fine-tuning, a standard incremental learning scheme.

free parameters (8)
  • ELBO KL weight beta = 1e-4
    Chosen by hand because it generalizes well across applications; controls uncertainty magnitude and active learning behavior (Section 3.1).
  • CHF stopping R2 threshold = 0.98
    Defines the 47% fewer experiments headline; different thresholds change the reported data savings (Section 3.2, Table 4).
  • AAL sample schedule = 10 initial + 20 per iteration, pool 500, 50 trials
    Sampling design choices set how quickly R2 rises; no sensitivity analysis is provided (Section 3.2).
  • Sensor binning count and zig-zag ordering = 10 quantile bins, alternating front/back
    Ad hoc preprocessing that creates artificial periodicity; no comparison to other orderings is given (Algorithm 2).
  • PSML update cadence = 1 month pre-train, monthly test, 33 sessions
    Rolling window size determines how much data the twin sees; no cadence sensitivity study is reported (Section 3.3).
  • Battery update cadence = retrain every 10 discharges
    Update frequency is chosen; no study of other cadences; affects MSE comparison (Section 5.4).
  • Number of stochastic forward passes S
    S used for predictive mean and 95% quantiles in Eqs. (3)-(4) is never stated; uncertainty estimates depend on it (Section 3.1).
  • Physics loss weight lambda
    Defined in Eq. (22) for PINN losses but its value for the battery model is not reported (Sections 3.5, 5.4).
assumptions (7)
  • standard math ELBO and reparameterization trick give unbiased gradient estimates for the variational posterior
    Basis of the Bayesian last layer training (Section 3.1, Eqs. 2-3).
  • standard math Mean-field factorization over network weights is a sufficient posterior approximation
    Used to define variational layers; simplifies ELBO but restricts posterior dependencies (Section 3.1).
  • domain assumption The dynamic system in Eq. (1) can be learned by a neural surrogate with a single variational output layer
    Underpins all four applications; no capacity or identifiability analysis is provided.
  • ad hoc to paper Concatenating sensors by mean-level zig-zag ordering preserves learnable temporal structure
    Algorithm 2 is justified by the observation that arbitrary concatenation may appear chaotic; no formal or comparative evidence is given.
  • domain assumption One month of data is sufficient to initialize and update the renewable twin
    Session-based training uses 1-month windows; no evidence this is optimal or sufficient for wind regime shifts (Sections 3.3, 5.2).
  • domain assumption The BattNN equivalent circuit model with fixed physical parameters can represent degradation through fitted subnetworks
    The battery twin inherits BattNN's model form; only the subnetworks are updated (Sections 3.5, 5.4).
  • ad hoc to paper A random candidate pool followed by uncertainty querying is a valid active learning design
    AAL is motivated by preliminary analysis that it was less greedy; no theoretical characterization is given (Section 3.2).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Variational Digital Twins." pith.science (2026). https://pith.science/paper/DFC6U7TU

@misc{pith2026250701047,
  author       = {Pith},
  title        = {Pith review of: Variational Digital Twins},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DFC6U7TU}},
  note         = {Machine review of arXiv:2507.01047}
}
read the original abstract

While digital twins (DT) hold promise for providing real-time insights into complex energy assets, much of the current literature either does not offer a clear framework for information exchange between the model and the asset, lacks key features needed for real-time implementation, or gives limited attention to model uncertainty. Here, we aim to solve these gaps by proposing a variational digital twin (VDT) framework that augments standard neural architectures with a single Bayesian output layer. This lightweight addition, along with a novel VDT updating algorithm, lets a twin update in seconds on commodity GPUs while producing calibrated uncertainty bounds that can inform experiment design, control algorithms, and model reliability. The VDT is evaluated on four energy-sector problems. For critical-heat-flux prediction, uncertainty-driven active learning reaches R2 = 0.98 using 47 % fewer experiments and one-third the training time of random sampling. A three-year renewable-generation twin maintains R2 > 0.95 for solar output and curbs error growth for volatile wind forecasts via monthly updates that process only one month of data at a time. A nuclear reactor transient cooldown twin reconstructs thermocouple signals with R2 > 0.99 and preserves accuracy after 50 % sensor loss, demonstrating robustness to degraded instrumentation. Finally, a physics-informed Li-ion battery twin, retrained after every ten discharges, lowers voltage mean-squared error by an order of magnitude relative to the best static model while adapting its credible intervals as the cell approaches end-of-life. These results demonstrate that combining modest Bayesian augmentation with efficient update schemes turns conventional surrogates into uncertainty-aware, data-efficient, and computationally tractable DTs, paving the way for dependable models across industrial and scientific energy systems.

Figures

Figures reproduced from arXiv: 2507.01047 by the authors.

Figure 10
Figure 10. Full 1D concatenation of fluid temperature sensor signals before and after applying Algorithm 2 for stacking sensor data to enhance sequential VDT performance. Vertical gray lines indicate the boundary where one sensor signal ends and the next begins. Once trained, the VLSTM model demonstrated a strong ability to track the complex thermal evolution of the reactor vessel throughout the depressurized conduction cooldo… view at source ↗
Figure 11
Figure 11. VGRU predictions for 62 solid and fluid temperature thermocouples in the unseen test set. Vertical gray lines indicate the boundary where one sensor signal ends and the next begins. Inset plots show a single sensor prediction and its corresponding location in the reactor on the right. reduced set of observations, but there is also a necessary amount of information required from the physical asset to obtain reliable … view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

63 extracted references · 45 canonical work pages

  1. [1]

    Sharma, E

    A. Sharma, E. Kosasih, J. Zhang, A. Brintrup, A. Calinescu, Digital twins: State of the art theory and practice, challenges, and open research questions, Journal of Industrial Information Integration 30 (2022) 100383

    work page 2022

  2. [2]

    J.-F. Yao, Y. Yang, X. -C. Wang, X. -P. Zhang, Systematic review of digital twin technology and applications, Visual computing for industry, biomedicine, and art 6 (1) (2023) 10

    work page 2023

  3. [3]

    R. He, G. Chen, C. Dong, S. Sun, X. Shen, Data-driven digital twin technology for optimized control in process systems, ISA transactions 95 (2019) 221–234

    work page 2019

  4. [4]

    Alanazi, M

    Y. Alanazi, M. Schram, K. Rajput, S. Goldenberg, L. Vidyaratne, C. Pappas, M. I. Radaideh, D. Lu, P. Ra - muhalli, S. Cousineau, Multi-module-based cvae to predict hvcm faults in the sns accelerator, Machine Learning with Applications 13 (2023) 100484

    work page 2023

  5. [5]

    Segovia, J

    M. Segovia, J. Garcia -Alfaro, Design, modeling and implementation of digital twins, Sensors 22 (14) (2022) 5396

    work page 2022

  6. [6]

    Onaji, D

    I. Onaji, D. Tiwari, P. Soulatiantork, B. Song, A. Tiwari, Digital twin in manufacturing: conceptual framework and case studies, International journal of computer integrated manufacturing 35 (8) (2022) 831–858

    work page 2022

  7. [7]

    M. I. Radaideh, C. Pappas, M. Wezensky, P. Ramuhalli, S. Cousineau, Early fault detection in particle accel - erator power electronics using ensemble learning, International Journal of Prognostics and Health Management 14 (1)

    work page

  8. [8]

    Radaideh, C

    M. Radaideh, C. Pappas, P. Ramuhalli, S. Cousineau, Application of convolutional and feedforward neural networks for fault detection in particle accelerator power systems, in: Annual Conference of the PHM Society, Vol. 14, 2022

    2022

Show all 63 references
  1. [9]

    T. Sun, X. He, Z. Li, Digital twin in healthcare: Recent updates and challenges, Digital Health 9 (2023) 20552076221149651

    2023

  2. [10]

    T. Deng, K. Zhang, Z. -J. M. Shen, A systematic review of a digital twin city: A new pattern of urban governance toward smart cities, Journal of management science and engineering 6 (2) (2021) 125–134

    2021

  3. [11]

    Arafet, R

    K. Arafet, R. Berlanga, Digital twins in solar farms: An approach through time series and deep learning, Algorithms 14 (5) (2021) 156

    2021

  4. [12]

    M. I. Radaideh, C. Pappas, J. Walden, D. Lu, L. Vidyaratne, T. Britton, K. Rajput, M. Schram, S. Cousineau, Time series anomaly detection in power electronics signals with recurrent and convlstm autoencoders, Digital Signal Processing 130 (2022) 103704

    2022

  5. [13]

    Kolahi, S

    M. Kolahi, S. Esmailifar, A. M. Sizkouhi, M. Aghaei, Digital-pv: A digital twin-based platform for autonomous aerial monitoring of large -scale photovoltaic power plants, Energy Conversion and Management 321 (2024) 118963

    2024

  6. [14]

    E. E. Ambarita, A. Karlsen, F. Scibilia, A. Hasan, Industry 4.0 digital twins in offshore wind farms, Wind Energy Science Discussions 2023 (2023) 1–34

    2023

  7. [15]

    Vairagade, S

    H. Vairagade, S. Kim, H. Son, F. Zhang, A nuclear power plant digital twin for developing robot navigation and interaction, Frontiers in Energy Research 12 (2024) 1356624

    2024

  8. [16]

    B. Li, W. Tan, A novel framework for integrating solar renewable source into smart cities through digital twin simulations, Solar Energy 262 (2023) 111869

    2023

  9. [17]

    T. I. Zohdi, A digital-twin and machine-learning framework for the design of multiobjective agrophotovoltaic solar farms, Computational Mechanics 68 (2) (2021) 357–370. 31

    2021

  10. [18]

    Y. Guo, Q. Tang, J. Darkwa, H. Wang, W. Su, D. Tang, J. Mu, Multi -objective integrated optimization of geothermal heating system with energy storage using digital twin technology, Applied Thermal Engineering 252 (2024) 123685

    2024

  11. [19]

    L. You, M. Zhu, Digital twin simulation for deep learning framework for predicting solar energy market load in trade-by-trade data, Solar Energy 250 (2023) 388–397

    2023

  12. [20]

    Prantikos, L

    K. Prantikos, L. H. Tsoukalas, A. Heifetz, Physics-informed neural network solution of point kinetics equations for a nuclear reactor digital twin, Energies 15 (20) (2022) 7697

    2022

  13. [21]

    Fahim, V

    M. Fahim, V. Sharma, T.-V. Cao, B. Canberk, T. Q. Duong, Machine learning-based digital twin for predictive modeling in wind turbines, IEEE Access 10 (2022) 14184–14194

    2022

  14. [22]

    Pimenta, J

    F. Pimenta, J. Pacheco, C. Branco, C. Teixeira, F. Magalh ˜aes, Development of a digital twin of an onshore wind turbine using monitoring data, in: Journal of Physics: Conference Series, Vol. 1618, IOP Publishing, 2020, p. 022065

    2020

  15. [23]

    E. Deri, C. Var´e, M. Wintergerst, Development of digital twins of pwr steam generators: Description of two maintenance-oriented use cases, in: International Conference on Nuclear Engineering, Vol. 85246, American Society of Mechanical Engineers, 2021, p. V001T01A002

    2021

  16. [24]

    Khalyutin, I

    S. Khalyutin, I. Starostin, I. Agafonkina, Generalized method of mathematical prototyping of energy processes for digital twins development, Energies 16 (4) (2023) 1933

    2023

  17. [25]

    P. Yu, L. Ma, R. Fu, Y. Liang, D. Qin, J. Yu, S. Liao, Framework design and application perspectives of digital twin microgrid, Energy Reports 9 (2023) 669–678

    2023

  18. [26]

    M. I. Radaideh, H. Tran, L. Lin, H. Jiang, D. Winder, S. Gorti, G. Zhang, J. Mach, S. Cousineau, Model calibration of the liquid mercury spallation target using evolutionary neural networks and sparse polynomial expansions, Nuclear Instruments and Methods in Physics Research S...

    2022

  19. [27]

    Price, M

    D. Price, M. I. Radaideh, B. Kochunas, Multiobjective optimization of nuclear microreactor reactivity control system operation with swarm and evolutionary algorithms, Nuclear Engineering and Design 393 (2022) 111776

    2022

  20. [28]

    M. I. Radaideh, K. Borowiec, T. Kozlowski, Integrated framework for model assessment and advanced un - certainty quantification of nuclear computer codes under bayesian statistics, Reliability Engineering & System Safety 189 (2019) 357–377

    2019

  21. [29]

    X. Wu, T. Kozlowski, H. Meidani, K. Shirvan, Inverse uncertainty quantification using the modular bayesian approach based on gaussian process, part 1: Theory, Nuclear Engineering and Design 335 (2018) 339–355

    2018

  22. [30]

    Zhang, J

    C. Zhang, J. Bu¨tepage, H. Kjellstr ¨om, S. Mandt, Advances in variational inference, IEEE transactions on pattern analysis and machine intelligence 41 (8) (2018) 2008–2026

    2018

  23. [31]

    M. I. Radaideh, L. Tunkle, D. Price, K. Abdulraheem, L. Lin, M. Elias, Multistep criticality search and power shaping in nuclear microreactors with deep reinforcement learning, Nuclear Science and Engineering (2025) 1–13

    2025

  24. [32]

    K. K. Abdulraheem, S. Choi, Q. Shen, B. Kochunas, M. I. Radaideh, A load following reactivity control system for nuclear microreactors, Progress in Nuclear Energy 184 (10567) (2025) 6

    2025

  25. [33]

    Price, N

    D. Price, N. Roskoff, M. I. Radaideh, B. Kochunas, Thermal modeling of an evinci™-like heat pipe microreactor using openfoam, Nuclear Engineering and Design 415 (2023) 112709

    2023

  26. [34]

    Yadav, H

    V. Yadav, H. Zhang, C. Chwasz, A. Gribok, C. Ritter, N. Lybeck, R. Hays, T. Trask, P. Jain, V. Badalassi, 32 et al., The state of technology of application of digital twins, US Nuclear Regulatory Commission, Tech. Rep. TLR/RES-DE-REB-2021-01

    2021

  27. [35]

    F. Tao, H. Zhang, C. Zhang, Advancements and challenges of digital twins in industry, Nature Computational Science 4 (3) (2024) 169–177

    2024

  28. [36]

    Iranshahi, J

    K. Iranshahi, J. Brun, T. Arnold, T. Sergi, U. C. Mu¨ller, Digital twins: Recent advances and future directions in engineering fields, Intelligent Systems with Applications (2025) 200516

    2025

  29. [37]

    M. I. Radaideh, T. Kozlowski, Surrogate modeling of advanced computer simulations using deep gaussian processes, Reliability Engineering & System Safety 195 (2020) 106731

    2020

  30. [38]

    M. I. Radaideh, C. Pigg, T. Kozlowski, Y. Deng, A. Qu, Neural-based time series forecasting of loss of coolant accidents in nuclear power plants, Expert Systems with Applications 160 (2020) 113699

    2020

  31. [39]

    Y.-P. Chen, V. Karkaria, Y. -K. Tsai, F. Rolark, D. Quispe, R. X. Gao, J. Cao, W. Chen, Real-time decision - making for digital twin in additive manufacturing with model predictive control using time -series deep neu - ral networks, Journal of Manufacturing Systems 80 (2025) 4...

    2025 doi

  32. [40]

    Schena, P

    L. Schena, P. A. Marques, R. Poletti, S. Ahizi, J. Van den Berghe, M. A. Mendez, Reinforcement twinning: From digital twins to model-based reinforcement learning, Journal of Computational Science 82 (2024) 102421

    2024

  33. [41]

    M. I. Radaideh, K. Du, P. Seurin, D. Seyler, X. Gu, H. Wang, K. Shirvan, Neorl: Neuroevolution optimization with reinforcement learning—applications to carbon-free energy systems, Nuclear Engineering and Design 412 (2023) 112423

    2023

  34. [42]

    H. Song, M. Song, X. Liu, Online autonomous calibration of digital twins using machine learning with appli- cation to nuclear power plants, Applied Energy 326 (2022) 119995

    2022

  35. [43]

    Kakhaia, P

    S. Kakhaia, P. Zun, D. Ye, V. Krzhizhanovskaya, Inverse uncertainty quantification of a mechanical model of arterial tissue with surrogate modelling, Reliability Engineering & System Safety 238 (2023) 109393

    2023

  36. [44]

    M. I. Radaideh, L. Lin, H. Jiang, S. Cousineau, Bayesian inverse uncertainty quantification of the physical model parameters for the spallation neutron source first target station, Results in Physics 36 (2022) 105414

    2022

  37. [45]

    Huang, G

    J. Huang, G. Yang, Z. Wang, J. J. Park, Diffusionpde: Generative pde-solving under partial observation, arXiv preprint arXiv:2406.17763

    arXiv

  38. [46]

    Price, M

    D. Price, M. I. Radaideh, B. Kochunas, Simplified matching pursuits applied to 3d nuclear reactor temperature distribution construction, Applied Mathematical Modelling 131 (2024) 134 –158

    2024

  39. [47]

    D. M. Blei, A. Kucukelbir, J. D. McAuliffe, Variational inference: A review for statisticians , Journal of the American Statistical Association 112 (518) (2017) 859–877. doi:10.1080/01621459.2017.1285773. URL http://dx.doi.org/10.1080/01621459.2017.1285773

    2017

  40. [48]

    Zhang, J

    C. Zhang, J. Bu¨tepage, H. Kjellstr ¨om, S. Mandt, Advances in variational inference, IEEE Transactions on Pattern Analysis and Machine Intelligence 41 (8) (2019) 2008–2026. doi:10.1109/TPAMI.2018.2889774

    2019

  41. [49]

    Harrison, J

    J. Harrison, J. Willes, J. Snoek, Variational bayesian last layers, in: The Twelfth International Conference on Learning Representations, 2024. URL https://openreview.net/forum?id=Sx7BIiPzys

    2024

  42. [50]

    S. Hochreiter, The vanishing gradient problem during learning recurrent neural nets and problem solutions, International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems 6 (02) (1998) 107–116. 33

    1998

  43. [51]

    Hochreiter, J

    S. Hochreiter, J. Schmidhuber, Long short-term memory, Neural computation 9 (8) (1997) 1735–1780

    1997

  44. [52]

    K. Cho, B. Van Merr i¨enboer, C. Gulcehre, D. Bahdanau, F. Bougares, H. Schwenk, Y. Bengio, Learning phrase representations using rnn encoder-decoder for statistical machine translation, arXiv preprint arXiv:1406.1078

    arXiv

  45. [53]

    Bahdanau, K

    D. Bahdanau, K. Cho, Y. Bengio, Neural machine translation by jointly learning to align and translate, arXiv preprint arXiv:1409.0473

    arXiv

  46. [54]

    Raissi, P

    M. Raissi, P. Perdikaris, G. E. Karniadakis, Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations, Journal of Computational physics 378 (2019) 686–707

    2019

  47. [55]

    G. E. Karniadakis, I. G. Kevrekidis, L. Lu, P. Perdikaris, S. Wang, L. Yang, Physics-informed machine learning, Nature Reviews Physics 3 (6) (2021) 422–440

    2021

  48. [56]

    P. Ren, Y. Xiao, X. Chang, P. -Y. Huang, Z. Li, B. B. Gupta, X. Chen, X. Wang, A survey of deep active learning, ACM computing surveys (CSUR) 54 (9) (2021) 1–40

    2021

  49. [57]

    F. Wang, Q. Zhi, Z. Zhao, Z. Zhai, Y. Liu, H. Xi, S. Wang, X. Chen, Inherently interpretable physics-informed neural network for battery modeling and prognosis, IEEE Transactions on Neural Networks and Learning Systems 36 (1) (2025) 1145–1159. doi:10.1109/TNNLS.2023.3329368

    2025

  50. [58]

    J.-M. L. Corre, G. Delipei, X. Wu, X. Zhao, Benchmark on artificial intelligence and machine learning for scientific computing in nuclear engineering. phase 1: Critical heat flux exercise specifications, NEA Working Papers

  51. [59]

    Zheng, N

    X. Zheng, N. Xu, L. Trinh, D. Wu, T. Huang, S. Sivaranjani, Y. Liu, L. Xie, A multi -scale time-series dataset with benchmark for machine learning in decarbonized energy grids, Scientific Data 9 (1) (2022) 359

    2022

  52. [60]

    Woods, Osu high temperature test facility design technical report, revision 2

    B. Woods, Osu high temperature test facility design technical report, revision 2. URL https://www.osti.gov/biblio/1599410

  53. [61]

    Woods, Instrumentation plan for the osu high temperature test facility, revision 4

    B. Woods, Instrumentation plan for the osu high temperature test facility, revision 4. URL https://www.osti.gov/biblio/1599628

  54. [62]

    Gutowska, B

    I. Gutowska, B. G. Woods, J. Halsted, Developing pcc and dcc integral effects test experiments at the high temperature test facility, Frontiers in Energy Research 11. doi:10.3389/fenrg.2023.1088070. URL https://www.frontiersin.org/journals/energy-research/articles/10.3389/fenr...

    2023

  55. [63]

    NASA Open Data Portal, Randomized battery usage 2: Room temperature random walkDataset published 20 Oct 2022

    2022

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.