Pith. sign in

REVIEW 3 major objections 4 minor 35 references

Spiking Neural Networks for Low-Power Vibration-Based Predictive Maintenance

T0 review · 3 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read A recurrent spiking neural network trained on 3-axis vibration data can simultaneously estimate flow, pressure, and speed and flag overpressure and cavitation, with >97% classification accuracy and no missed critical faults.

desk verdict A credible SNN condition-monitoring case study whose headline accuracy may not survive a condition-disjoint test split; still worth peer review. read the letter →

arxiv 2506.13416 v1 pith:R232EA3Y submitted 2025-06-16 cs.LG

Alexandru Vasilache , Sven Nitzsche , Christian Kneidl , Mikael Tekneyan , Moritz Neher , Juergen Becker This is my paper · ORCID
classification cs.LG
keywords spikingneuralnetworkspredictivemaintenancevibrationmonitoringneuromorphiccomputingprogressingcavitypumpedgeAImulti-tasklearningenergyestimation
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 show that a recurrent spiking neural network, fed only with three-axis vibration measurements from an industrial pump, can replace several dedicated sensors by jointly estimating flow, pressure, and pump speed while classifying the pump as normal, overpressured, or cavitating. It reports classification accuracy above 97%, with zero false negatives for the two critical fault classes, and smoothed regression errors below 1% mean relative percentage error for flow and pump speed. The authors also estimate that running the trained network on neuromorphic hardware would consume about 0.0032 J per inference, orders of magnitude less than their estimates for conventional x86 or ARM processors, which would make battery-powered on-sensor predictive maintenance feasible. The energy comparison is explicitly an estimate based on literature constants, and the paper calls for hardware validation.

What carries the argument

The load-bearing mechanism is a recurrent network of leaky integrate-and-fire (LIF) neurons trained with a smooth surrogate for the spike gradient through backpropagation in time. Its input is a hybrid encoding: Step-Forward temporal encoding turns each normalized vibration axis into positive and negative spike trains, and Poisson rate encoding carries the axis's local mean and standard deviation, producing 12 spike channels total. Recurrence lets the network carry state across the 16,384 timesteps of a recording, and outputs are averaged over time before a softmax is applied to the classification heads. The energy comparison rests on a counting rule that charges spiking hardware only for synaptic operations triggered by actual spikes while charging conventional hardware for every connection at every timestep. Magnitude-based pruning and fixed-point quantization reduce the trained model to about 80,000 parameters and 175 KB, making the deployment footprint concrete.

What would settle it

Run the same pruned and quantized network on a physical neuromorphic board and on a representative x86 or ARM edge board, measuring actual joules per 16,384-step inference; if the measured neuromorphic-to-x86 ratio is far below three orders of magnitude, the paper's headline energy advantage fails.

Watch

Extended reading notes

Core claim

The paper reports that, on a custom dataset of roughly 4000 recordings from a progressing cavity pump, a compressed recurrent SNN with two hidden layers of 160 leaky integrate-and-fire neurons performs six simultaneous output tasks from 12 input spike channels derived from 3-axis vibration: it regresses flow, pressure, and pump speed and classifies normal, overpressure, and cavitation states. The model achieves an overall classification accuracy of 97.95%, false negative rates of 0.0% for overpressure and cavitation, and smoothed mean relative percentage errors of 0.93% for flow and 0.93% for pump speed, although pressure prediction remains above typical industrial sensor accuracy. The paper's energy model, based on per-operation constants and an event-driven counting rule, estimates $3.16 imes 10^{-3}$ J per 16,384-step inference on neuromorphic hardware, versus $11.3$ J on x86 and $1.18$ J on ARM.

Load-bearing premise

The three-order-of-magnitude energy advantage depends on literature per-operation energy constants and on a counting rule that charges neuromorphic hardware only for spike-driven operations while charging x86 and ARM for every connection at every timestep, and these values are estimates, not measurements.

Editorial extensions

If this is right

  • A single vibration sensor with on-device SNN inference could replace separate flow, pressure, and speed sensors, reducing the cost and cabling of industrial pump monitoring.
  • Zero false negatives for overpressure and cavitation means the model would not silently miss the two fault modes most likely to damage a progressing cavity pump.
  • Estimated neuromorphic energy near 0.003 J per inference makes continuous local monitoring on battery power plausible instead of streaming high-resolution vibration to the cloud.
  • Flow and speed regression accuracy close to industrial sensor tolerances suggests the SNN can act as a virtual sensor for parameters that are otherwise expensive to instrument.
  • Further pruning or penalizing spike frequency in the training objective could lower the dominant synaptic-operation term and push energy lower still.

Reading between the lines

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

  • Because the energy advantage scales with spike count, the same model could consume far less energy on calmer vibration regimes and more during faults; an explicit spike-rate penalty in the loss would make the deployment energy more predictable.
  • The training augmentation used overlapping sliding windows with step size 1, so adjacent training segments share nearly all their data; a validation on non-overlapping windows would give a stricter picture of generalization to unseen recordings.
  • If the approach transfers across pump sizes, media, and operating conditions, the same 12-channel input and recurrent architecture could be retrained for other rotating machinery without changing the sensing hardware.
  • The paper leaves Bluetooth transmission gaps of 13 seconds between recordings; a fully edge-deployed version that processes continuously would need to show the model also tracks transients inside those gaps, not just steady 2.5-second windows.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper investigates a recurrent spiking neural network (SNN) for simultaneous regression of flow, outlet pressure, and pump speed, and multi-label classification of normal, overpressure, and cavitation conditions, using 3-axis vibration data from a progressing cavity pump. The pipeline extracts 12 input channels per time window via Step-Forward spike encoding plus Poisson rate encoding of local mean and standard deviation per axis, feeds them into two recurrent LIF layers of 160 neurons, and trains with surrogate gradients and a composite regression/classification loss. After pruning and quantization, the model is evaluated on a held-out test set of 607 full-length recordings. The authors report smoothed MRPE below 1% for flow and speed, 2.46% for pressure, classification accuracy above 97%, zero false negatives for the two critical fault classes, and estimated energy per inference of 3.16e-3 J on Loihi versus 11.3 J on x86 and 1.18 J on ARM.

Significance. If the experimental evaluation is valid, the paper makes a useful contribution: it demonstrates a multi-task SNN on a realistic industrial pump dataset, with detailed preprocessing, architecture, compression, and energy-estimation methodology, and it explicitly acknowledges several limitations of the energy model. The use of a custom real-world dataset, the clear reporting of hyperparameters and compression details, and the comparison to industrial sensor accuracy specifications are strengths. However, the evaluation protocol is the main risk: the random split over quasi-stationary operating points may overstate generalization, and the fault labels are partly derived from the same quantities that serve as regression targets.

major comments (3)
  1. [Section III-B1 and Section II-C] The data-generation procedure holds speed and pressure constant for 5 minutes per recording, so the 4000 recordings form a set of quasi-stationary operating conditions. A random 70/15/15 split can place recordings from the same (speed, pressure, fault) condition in both training and test sets, allowing the network to memorize discrete setpoints rather than learn condition-invariant vibration signatures. The reported test accuracy in Table IV and the zero FNR therefore do not establish generalization to unseen operating regimes. Please report a condition-disjoint or temporally ordered split, give class counts per split, and provide confidence intervals for the classification and regression metrics.
  2. [Section III-A and Table IV] The input features include Poisson-coded local mean and standard deviation per axis, which are strong correlates of pump speed and pressure, and the SF encoding is applied after local standardization. These features give the model a potential shortcut for both regression and classification, because the fault labels are threshold-derived from pressure. Please provide an ablation that removes or degrades the local-statistic channels, and evaluate whether classification remains accurate under unseen combinations of speed and pressure. Without this, the causal role of vibration dynamics in the high accuracy is unclear.
  3. [Section II-A] The classification labels are defined by thresholds on the same physical quantities that are regression targets: overpressure is discharge pressure above 12 bar, and cavitation is suction pressure below 0.75 bar. Classification success can therefore be partly a restatement of regression success. The paper mentions that an operator also adjusted labels by listening, but it does not specify how often the threshold-based and listening-based labels disagreed. Please clarify the independent label-assignment protocol and quantify the agreement, or otherwise evaluate classification against labels that are not deterministic functions of the regression targets.
minor comments (4)
  1. [Abstract and Conclusion] The abstract states energy savings of 'up to 3 orders of magnitude' while the conclusion states 'up to 4 orders of magnitude'; Table V gives a ratio of about 3576, so the phrasings should be aligned and explicitly labeled as estimates.
  2. [Table III] The MAPE values for pressure (9237% raw, 2447% smoothed) are dominated by target values near zero and can mislead; consider reporting MAPE only for samples with nonzero targets or replacing it with a range-based metric for pressure.
  3. [Figure 6] The x-axis in Figure 6 is labeled up to 500, while the test set has 607 samples; please correct the axis limit or clarify the number of displayed samples.
  4. [Section III-F] There is an inconsistency in the test-set size: Section III-B1 reports 607 test sequences, while Section III-F states 606 samples; please correct this.

Circularity Check

1 steps flagged · score 2.0 of 10

Minor circularity: the overpressure fault label is threshold-derived from the same outlet-pressure quantity that is a regression target, so the zero-FNR classification claim partly restates regression accuracy; the energy and regression claims are otherwise self-contained.

  1. self definitional [Section II-A (Pump System), with regression loss in Section III-B4 (Eq. 1)]
    "Any recordings exceeding this limit, including a small transition zone, were labeled as overpressure. Any conditions with a suction pressure below 0.75 bar were labeled as cavitation... for each 3-axis vibration recording of 16384 points measured at 6664 Hz, a label was assigned based on the measured regression (flow, outlet pressure, pump speed) and classification (normal, overpressure, cavitation) values."

    The overpressure class is defined by thresholding outlet pressure (>12 bar), and outlet pressure is one of the three regression targets supervised in L_reg (Eq. 1: sum over flow, pressure, pump speed of |yhat-y|). Therefore 'zero FNR for overpressure' is not an independent fault-detection result: the information needed for that label is the same physical quantity the network is simultaneously trained to regress. The learned classifier could in principle disagree with the regressor, so the reduction is partial rather than forced, but the classification claim largely restates pressure-regression quality rather than demonstrating a separate capability. Cavitation is less redundant because it uses suction pressure and operator listening, not one of the regression targets.

full rationale

The paper is an empirical training/evaluation study rather than a formal derivation, so most of the chain is self-contained: the SNN is trained on recorded vibration data with external sensor ground truth, and the energy estimates are explicitly labeled literature-based estimates using the Nengo-style counting rule. Self-citations [2], [4], and [20] are used for prior context, a survey, and an encoding tool; none carries a load-bearing uniqueness or theoretical claim. The one partial circularity is the overpressure label construction, which is a threshold on the same outlet-pressure quantity that is a regression target; this makes the zero-FNR classification metric partly a restatement of regression accuracy. The random train/test split over quasi-stationary operating points is a valid generalization concern but is not circularity. Overall circularity is minor because the central energy-comparison and regression claims do not reduce to their inputs.

Assumptions & free parameters 7 free parameters · 4 assumptions · 0 invented entities

The central performance and energy claims rest on tuned hyperparameters (free parameters), literature hardware energy constants, a favorable event-counting model for neuromorphic chips, and a random data split that may allow leakage. No new physical entities are introduced.

free parameters (7)
  • LIF neuron threshold = 0.959
    Tuned with Optuna on validation loss; directly controls spiking activity and thus energy estimates on Loihi.
  • Membrane decay factor beta = 0.900
    Tuned with Optuna; sets the temporal integration of inputs.
  • Learning rate = 1.224e-2
    Tuned with Optuna.
  • Weight initialization parameters = mean -0.048, std 0.238
    Tuned with Optuna; identified as high-impact on performance.
  • Classification loss weight = 10
    Hand-chosen multiplier on weighted CE loss to balance regression and classification.
  • Sliding window segment length = 1024
    Chosen for training; validation/test use full 16384 steps, so there is a train/serve distribution shift.
  • SF encoding thresholds = not reported numerically
    Optimized offline via the Spike Encoding Framework [20] on training data; exact values not given in the paper.
assumptions (4)
  • domain assumption Per-operation energy constants from Table II (x86, ARM, Loihi) are accurate for the comparison.
    Used in Eq. 4; the energy claim depends on these literature values from [27], [28].
  • domain assumption On spiking hardware, synaptic energy is proportional to generated spikes and can exploit connection sparsity; on non-spiking hardware, all connections are computed every timestep.
    This counting rule (Eq. 5) drives the three-order-of-magnitude gap; favorable to Loihi and not empirically validated here.
  • domain assumption Ground-truth fault labels (overpressure, cavitation) are correct as defined by pressure thresholds and operator listening.
    Section II-A; label noise directly affects reported accuracy.
  • domain assumption Random 70/15/15 split of recordings prevents temporal leakage.
    Section III-B1; recordings are taken during 5-minute stationary operating points, so randomly split windows may share the same conditions.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Spiking Neural Networks for Low-Power Vibration-Based Predictive Maintenance." pith.science (2026). https://pith.science/paper/R232EA3Y

@misc{pith2026250613416,
  author       = {Pith},
  title        = {Pith review of: Spiking Neural Networks for Low-Power Vibration-Based Predictive Maintenance},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/R232EA3Y}},
  note         = {Machine review of arXiv:2506.13416}
}
read the original abstract

Advancements in Industrial Internet of Things (IIoT) sensors enable sophisticated Predictive Maintenance (PM) with high temporal resolution. For cost-efficient solutions, vibration-based condition monitoring is especially of interest. However, analyzing high-resolution vibration data via traditional cloud approaches incurs significant energy and communication costs, hindering battery-powered edge deployments. This necessitates shifting intelligence to the sensor edge. Due to their event-driven nature, Spiking Neural Networks (SNNs) offer a promising pathway toward energy-efficient on-device processing. This paper investigates a recurrent SNN for simultaneous regression (flow, pressure, pump speed) and multi-label classification (normal, overpressure, cavitation) for an industrial progressing cavity pump (PCP) using 3-axis vibration data. Furthermore, we provide energy consumption estimates comparing the SNN approach on conventional (x86, ARM) and neuromorphic (Loihi) hardware platforms. Results demonstrate high classification accuracy (>97%) with zero False Negative Rates for critical Overpressure and Cavitation faults. Smoothed regression outputs achieve Mean Relative Percentage Errors below 1% for flow and pump speed, approaching industrial sensor standards, although pressure prediction requires further refinement. Energy estimates indicate significant power savings, with the Loihi consumption (0.0032 J/inf) being up to 3 orders of magnitude less compared to the estimated x86 CPU (11.3 J/inf) and ARM CPU (1.18 J/inf) execution. Our findings underscore the potential of SNNs for multi-task PM directly on resource-constrained edge devices, enabling scalable and energy-efficient industrial monitoring solutions.

Figures

Figures reproduced from arXiv: 2506.13416 by the authors.

Figure 1
Figure 1. Progressing cavity pump (PCP) A. Pump System Throughout this study, the NETZSCH NM021BY02S PCP [14] was used, which is a type of rotary positive displacement pump [15]. The pump’s primary components include an elas￾tomer stator, a metal screw rotor, and a joint bar (see [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Pump data recording setup To measure pump speed, the Omron E2B-M12KN05-M1- B1 [19] was used. Its range is between 0 and 2000 min−1 , and offers an accuracy of +/- 0.1% from the measured value. C. Data Generation The dataset was mainly recorded manually, in part due to the operator’s necessary expertise regarding the machine’s limit states. For this purpose, the operator sets the pump’s specified speed on the frequen… view at source ↗
Figure 3
Figure 3. This pipeline converts the raw 3-axis (x, y, z) accelerometer readings, provided as sequences of uint16 values within a defined time window (16384 samples), into a multi-channel spike train format suitable for neuromorphic processing. The pipeline consists of the following steps, exe￾cuted sequentially for each time window [PITH_FULL_IMAGE:figures/full_fig_p002_3.png] view at source ↗
Figures from the paper (3 more)
Figure 3
Figure 3. Figure 3: Preprocessing applied to each axis (x, y, z). Af￾ter global standardization, a numeric time series undergoes local normalization, yielding a normalized time series and the window’s local mean (µx) and standard deviation (σx). The normalized time series is encoded into …
Figure 4
Figure 4. Figure 4: Overview of the recurrent SNN architecture. It pro [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 6
Figure 6. Figure 6: Classification performance on the sorted test set. [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

35 extracted references · 28 canonical work pages

  1. [1]

    Establishing the right analytics-based maintenance strategy,

    M. . Company, “Establishing the right analytics-based maintenance strategy,” McKinsey & Company, Tech. Rep., 2021. [Online]. Avail- able: https://www.mckinsey.com/capabilities/operations/our-insights/ establishing-the-right-analytics-based-maintenance-strategy

    work page 2021

  2. [2]

    Ultra-low power machinery fault detection using deep neural networks,

    S. Nitzsche, M. Neher, S. von Dosky, and J. Becker, “Ultra-low power machinery fault detection using deep neural networks,” in Machine Learning and Principles and Practice of Knowledge Discovery in Databases, ser. Communications in Computer and Information Science. Springer International Publishing, 2021, pp. 390–396

    work page 2021

  3. [3]

    Benchmarking keyword spotting efficiency on neuromorphic hardware,

    P. Blouw, X. Choo, E. Hunsberger, and C. Eliasmith, “Benchmarking keyword spotting efficiency on neuromorphic hardware,” 2019. [Online]. Available: https://arxiv.org/abs/1812.01739

    arXiv 2019

  4. [4]

    Low-Power Vibration-Based Predictive Maintenance for Industry 4.0 using Neural Networks: A Survey

    A. Vasilache, S. Nitzsche, D. Floegel, T. Schuermann, S. von Dosky, T. Bierweiler, M. Mußler, F. K¨alber, S. Hohmann, and J. Becker, “Low- power vibration-based predictive maintenance for industry 4.0 using neural networks: A survey,” arXiv preprint arXiv:2408.00516 , 2024

    work page Pith review arXiv 2024

  5. [5]

    Online detection of vibration anomalies using balanced spiking neural networks,

    N. Dennler, G. Haessig, M. Cartiglia, and G. Indiveri, “Online detection of vibration anomalies using balanced spiking neural networks,” in 2021 IEEE 3rd International Conference on Artificial Intelligence Circuits and Systems (AICAS) . IEEE, 2021, pp. 1–4. [Online]. Available: https://ieeexplore.ieee.org/document/9458403/

    work page arXiv 2021

  6. [6]

    A spiking neural network-based approach to bearing fault diagnosis,

    L. Zuo, L. Zhang, Z.-H. Zhang, X.-L. Luo, and Y . Liu, “A spiking neural network-based approach to bearing fault diagnosis,” Journal of Manufacturing Systems, vol. 61, pp. 714–724, 2021. [Online]. Available: https://linkinghub.elsevier.com/retrieve/pii/S0278612520301138

    work page 2021

  7. [7]

    Research on fault diagnosis based on spiking neural networks in deep space environment,

    R. Li and J. Yuan, “Research on fault diagnosis based on spiking neural networks in deep space environment,” in 2022 3rd Asia Service Sciences and Software Engineering Conference . ACM, 2022, pp. 165–170. [Online]. Available: https://dl.acm.org/doi/10.1145/3523181.3523205

    work page arXiv 2022

  8. [8]

    A multi-layer spiking neural network-based approach to bearing fault diagnosis,

    L. Zuo, F. Xu, C. Zhang, T. Xiahou, and Y . Liu, “A multi-layer spiking neural network-based approach to bearing fault diagnosis,” Reliability Engineering & System Safety , vol. 225, p. 108561,

    work page

Show all 35 references
  1. [9]

    Novel spiking neural network model for gear fault diagnosis,

    Y . H. Ali, F. Y . H. Ahmed, A. M. Abdelrhman, S. M. Ali, A. A. Borhana, and R. Ishak Raja Hamzah, “Novel spiking neural network model for gear fault diagnosis,” in 2022 2nd International Conference on Emerging Smart Technologies and Applications (eSmarTA) . IEEE, 2022, pp. 1–...

    2022

  2. [10]

    Damage detection in structural health monitoring with spiking neural networks,

    L. Zanatta, F. Barchi, A. Burrello, A. Bartolini, D. Brunelli, and A. Acquaviva, “Damage detection in structural health monitoring with spiking neural networks,” in 2021 IEEE International Workshop on Metrology for Industry 4.0 & IoT (MetroInd4.0&IoT) . IEEE, 2021, pp. 105–110...

    2021

  3. [11]

    Spiking neural network-based near-sensor computing for damage detection in structural health monitoring,

    F. Barchi, L. Zanatta, E. Parisi, A. Burrello, D. Brunelli, A. Bartolini, and A. Acquaviva, “Spiking neural network-based near-sensor computing for damage detection in structural health monitoring,” Future Internet, vol. 13, no. 8, p. 219, 2021. [Online]. Available: https: //w...

    2021

  4. [12]

    Efficient time series classification using spiking reservoir,

    S. Dey, D. Banerjee, A. M. George, A. Mukherjee, and A. Pal, “Efficient time series classification using spiking reservoir,” in 2022 International Joint Conference on Neural Networks (IJCNN) . IEEE, 2022, pp. 1–8. [Online]. Available: https://ieeexplore.ieee.org/document/9892728/

    2022

  5. [13]

    Comparing reservoir artificial and spiking neural networks in machine fault detection tasks,

    V . Kholkin, O. Druzhina, V . Vatnik, M. Kulagin, T. Karimov, and D. Butusov, “Comparing reservoir artificial and spiking neural networks in machine fault detection tasks,” BDCC, vol. 7, no. 2, p. 110, 2023. [Online]. Available: https://www.mdpi.com/2504-2289/7/2/110

    2023

  6. [14]

    NEMO® BY Blockpumpe in Industrieausf ¨uhrung

    “NEMO® BY Blockpumpe in Industrieausf ¨uhrung.” [Online]. Available: https://pumps-systems.netzsch.com/ de/produkte-und-zubehoer/nemo-exzenterschneckenpumpen/ nemo-by-blockpumpe-in-industrieausfuehrung

  7. [15]

    Moineau, “Pompe,” Patent US 1 892 217

    J. Moineau, “Pompe,” Patent US 1 892 217

  8. [16]

    SITRANS SCM IQ, SITRANS CC220, SITRANS MS200

    “SITRANS SCM IQ, SITRANS CC220, SITRANS MS200.” [Online]. Available: https://mall.industry.siemens.com/mall/en/WW/ Catalog/Product/7MP2210-2AB21-2AB1

  9. [17]

    PG2453 - Pressure sensor - ifm

    “PG2453 - Pressure sensor - ifm.” [Online]. Available: https: //www.ifm.com/de/en/product/PG2453

  10. [18]

    Promag W 300, 5W3B1H

    “Promag W 300, 5W3B1H.” [Online]. Available: https://www.de. endress.com/de/messgeraete-fuer-die-prozesstechnik/5W3B1H

  11. [19]

    E2B-M12KN05-M1-B1 OMI | OMRON

    “E2B-M12KN05-M1-B1 OMI | OMRON.” [Online]. Available: https: //industrial.omron.de/de/products/E2B-M12KN05-M1-B1

  12. [20]

    A pytorch-compatible spike encoding framework for energy-efficient neuromorphic applications,

    A. Vasilache, J. Scholz, V . Schilling, S. Nitzsche, F. Kaelber, J. Korsch, and J. Becker, “A pytorch-compatible spike encoding framework for energy-efficient neuromorphic applications,” 2025. [Online]. Available: https://arxiv.org/abs/2504.11026

    2025 arXiv

  13. [21]

    Evolving spatio-temporal data machines based on the neucube neuromorphic framework: Design methodology and selected applications,

    N. Kasabov, N. M. Scott, E. Tu, S. Marks, N. Sengupta, E. Capecci, M. Othman, M. G. Doborjeh, N. Murli, R. Hartono et al. , “Evolving spatio-temporal data machines based on the neucube neuromorphic framework: Design methodology and selected applications,” Neural Networks, vol....

    2016

  14. [22]

    PyTorch 2: Faster Machine Learning Through Dynamic Python Bytecode Transformation and Graph Compilation,

    J. Ansel, E. Yang, H. He, N. Gimelshein, A. Jain, M. V oznesensky, B. Bao, P. Bell, D. Berard, E. Burovski, G. Chauhan, A. Chourdia, W. Constable, A. Desmaison, Z. DeVito, E. Ellison, W. Feng, J. Gong, M. Gschwind, B. Hirsh, S. Huang, K. Kalambarkar, L. Kirsch, M. Lazos, M. Le...

    2024

  15. [23]

    Scikit-learn: Machine learning in Python,

    F. Pedregosa, G. Varoquaux, A. Gramfort, V . Michel, B. Thirion, O. Grisel, M. Blondel, P. Prettenhofer, R. Weiss, V . Dubourg, J. Vander- plas, A. Passos, D. Cournapeau, M. Brucher, M. Perrot, and E. Duch- esnay, “Scikit-learn: Machine learning in Python,” Journal of Machine ...

    2011

  16. [24]

    Adam: A method for stochastic optimization,

    D. P. Kingma and J. Ba, “Adam: A method for stochastic optimization,” arXiv preprint arXiv:1412.6980 , 2014

    2014 arXiv

  17. [25]

    Training spiking neural networks using lessons from deep learning,

    J. K. Eshraghian, M. Ward, E. Neftci, X. Wang, G. Lenz, G. Dwivedi, M. Bennamoun, D. S. Jeong, and W. D. Lu, “Training spiking neural networks using lessons from deep learning,” Proceedings of the IEEE , vol. 111, no. 9, pp. 1016–1054, 2023

    2023

  18. [26]

    Optuna: A next- generation hyperparameter optimization framework,

    T. Akiba, S. Sano, T. Yanase, T. Ohta, and M. Koyama, “Optuna: A next- generation hyperparameter optimization framework,” in Proceedings of the 25th ACM SIGKDD international conference on knowledge discovery & data mining , 2019, pp. 2623–2631

    2019

  19. [27]

    Assessing trends in performance per watt for signal processing applications,

    B. Degnan, B. Marr, and J. Hasler, “Assessing trends in performance per watt for signal processing applications,” IEEE Transactions on Very Large Scale Integration (VLSI) Systems, vol. 24, no. 1, pp. 58–66, 2015

    2015

  20. [28]

    Loihi: A neuromorphic manycore processor with on-chip learning,

    M. Davies, N. Srinivasa, T.-H. Lin, G. Chinya, Y . Cao, S. H. Choday, G. Dimou, P. Joshi, N. Imam, S. Jain et al. , “Loihi: A neuromorphic manycore processor with on-chip learning,” Ieee Micro, vol. 38, no. 1, pp. 82–99, 2018

    2018

  21. [29]

    Nengo: a Python tool for building large-scale functional brain models,

    T. Bekolay, J. Bergstra, E. Hunsberger, T. DeWolf, T. Stewart, D. Ras- mussen, X. Choo, A. V oelker, and C. Eliasmith, “Nengo: a Python tool for building large-scale functional brain models,” Frontiers in Neuroinformatics, vol. 7, no. 48, pp. 1–13, 2014

    2014

  22. [30]

    What is Flow Meter Accuracy?

    “What is Flow Meter Accuracy?” [Online]. Available: https://koboldusa. com/articles/common-questions/what-is-flow-meter-accuracy/

  23. [31]

    Triboelectric rotational speed sensor integrated into a bearing: A solid step to industrial application,

    Z. Xie, J. Dong, Y . Li, L. Gu, B. Song, T. Cheng, and Z. L. Wang, “Triboelectric rotational speed sensor integrated into a bearing: A solid step to industrial application,” Extreme Mechanics Letters , vol. 34, p. 100595, 2020

    2020

  24. [32]

    Comparison of single and double electrostatic sensors for rotational speed measurement,

    L. Wang, Y . Yan, and K. Reda, “Comparison of single and double electrostatic sensors for rotational speed measurement,” Sensors and Actuators A: Physical , vol. 266, pp. 46–55, 2017

    2017

  25. [33]

    Rotational speed measurement through electrostatic sensing and correlation signal processing,

    L. Wang, Y . Yan, Y . Hu, and X. Qian, “Rotational speed measurement through electrostatic sensing and correlation signal processing,” IEEE Transactions on Instrumentation and Measurement , vol. 63, no. 5, pp. 1190–1199, 2013

    2013

  26. [34]

    MIP Series Heavy Duty Pressure Transducers | Honeywell

    “MIP Series Heavy Duty Pressure Transducers | Honeywell.” [Online]. Available: https://automation.honeywell.com/de/de/products/ sensing-solutions/sensors/pressure-sensors/mip-series

  27. [2022]

    Available: https://linkinghub.elsevier.com/retrieve/pii/ S0951832022002095

    [Online]. Available: https://linkinghub.elsevier.com/retrieve/pii/ S0951832022002095

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.