pith. machine review for the scientific record. sign in

arxiv: 2605.09689 · v1 · submitted 2026-05-10 · 📡 eess.SY · cs.SY

Recognition: no theorem link

Nullspace-based Fault Diagnosis for Closed-Loop Mechatronic Systems with Application to Semiconductor Equipment

Jeroen van de Wijdeven, Koen Classens, Maurice Heemels, Tom Oomen

Authors on Pith no claims yet

Pith reviewed 2026-05-12 02:47 UTC · model grok-4.3

classification 📡 eess.SY cs.SY
keywords fault detection and isolationnullspace-based methodsclosed-loop systemsmechatronic systemswafer stagesemiconductor equipmentactuator faultssensor faults
0
0 comments X

The pith

Nullspace-based fault diagnosis conditions can be tailored for closed-loop linear mechatronic systems to detect and isolate actuator and sensor faults.

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

The paper develops methods to apply nullspace-based fault detection and isolation techniques to mechatronic systems operating in closed-loop control. It focuses on linear system models that include actuator and sensor faults. The tailored synthesis conditions are then tested on a prototype wafer stage used in semiconductor manufacturing. Real experiments demonstrate that faults can be identified effectively in this setting. This shows the approach can extend to other production machines and instruments.

Core claim

The authors derive tailored FDI synthesis conditions based on the nullspace method that account for the closed-loop dynamics of linear mechatronic systems subject to actuator and sensor faults, and validate them through application to a large-scale prototype wafer stage with experimental results.

What carries the argument

The nullspace-based FDI synthesis conditions, adapted to handle closed-loop aspects and specific fault types in linear systems.

Load-bearing premise

That the mechatronic systems can be adequately modeled as linear and that the closed-loop dynamics permit direct application of the nullspace-based FDI synthesis conditions without significant unmodeled effects.

What would settle it

If experiments on the wafer stage show that the tailored conditions fail to detect or isolate injected faults accurately, or if unmodeled dynamics lead to frequent false positives in fault detection.

Figures

Figures reproduced from arXiv: 2605.09689 by Jeroen van de Wijdeven, Koen Classens, Maurice Heemels, Tom Oomen.

Figure 1
Figure 1. Figure 1: Closed-loop controlled system equipped with a residual generator. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Open-loop system equipped with a residual generator. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Prototype experimental wafer stage setup. The moving part, the [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 6
Figure 6. Figure 6: Block diagram of the closed-loop controlled prototype wafer stage [PITH_FULL_IMAGE:figures/full_fig_p007_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Bode magnitude plot of the BLA in ( ), measured using multisine excitation and the robust method, and the 20th-order fit in ( ), obtained using the frequency domain simplified refined instrumental variable (SRIVC) method with integrated prediction error minimization (IPEM). where G act f = Gˆ u and G sens f = I4. Since the sensors are of high quality, no disturbance and noise contributions are taken into a… view at source ↗
Figure 8
Figure 8. Figure 8: Absolute and normalized values of the residual signals [PITH_FULL_IMAGE:figures/full_fig_p009_8.png] view at source ↗
read the original abstract

Fault detection and isolation (FDI) systems are critical for modern mechatronic production equipment, as their continuous operation is heavily dependent on the ability to detect and isolate faults in a timely and efficient manner. The aim of this paper is to address closed-loop aspects for linear systems and enable the application of well-known nullspace-based FDI synthesis conditions to mechatronic systems subject to actuator and sensor faults. These tailored FDI synthesis conditions are applied to a large-scale prototype wafer stage, showcasing the proposed approach through real experiments, thereby underlining the usefulness of the derived synthesis conditions for a wide range of production machines and scientific instruments.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 1 minor

Summary. The paper derives tailored nullspace-based FDI synthesis conditions for closed-loop linear mechatronic systems subject to actuator and sensor faults. These conditions are then applied to a large-scale prototype wafer stage, with validation provided through real experiments to demonstrate usefulness for production machines and scientific instruments.

Significance. If the synthesis conditions correctly incorporate closed-loop dynamics and the experimental results robustly confirm fault detection/isolation performance, the work would provide a practical extension of standard nullspace FDI methods to high-precision mechatronic equipment, with direct relevance to semiconductor manufacturing reliability.

major comments (1)
  1. The central experimental claim (application and validation on the wafer-stage prototype) rests on the assumption that the closed-loop system is adequately linear and that fault signatures are not significantly distorted by unmodeled effects. The manuscript should include explicit model-validation residuals, closed-loop vs. open-loop fault-signature comparisons, and quantitative detection/isolation rates under operating conditions to substantiate that the linear FDI conditions apply directly.
minor comments (1)
  1. Abstract: the phrasing 'tailored FDI synthesis conditions' is repeated without a concise statement of the specific modifications made to the standard nullspace method for closed-loop systems.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive feedback and positive evaluation of the work's relevance to semiconductor manufacturing. We address the major comment on experimental validation below, incorporating revisions where feasible while noting practical constraints of the prototype setup.

read point-by-point responses

  1. Referee: The central experimental claim (application and validation on the wafer-stage prototype) rests on the assumption that the closed-loop system is adequately linear and that fault signatures are not significantly distorted by unmodeled effects. The manuscript should include explicit model-validation residuals, closed-loop vs. open-loop fault-signature comparisons, and quantitative detection/isolation rates under operating conditions to substantiate that the linear FDI conditions apply directly.

    Authors: We agree that stronger substantiation of the linearity assumption would enhance the experimental section. In the revised manuscript we have added explicit model-validation residuals (comparing measured closed-loop responses against the identified linear model) and quantitative detection/isolation rates (true-positive/false-positive percentages) under representative operating conditions. Direct closed-loop versus open-loop fault-signature comparisons cannot be performed experimentally, as the wafer-stage prototype is operated exclusively in closed-loop mode for safety, stability, and precision reasons; open-loop excitation would risk mechanical damage and is outside the intended application domain. To address the spirit of the request we have instead included simulation-based comparisons of fault signatures with and without the closed-loop controller, confirming that the derived nullspace conditions correctly account for the closed-loop dynamics. revision: partial

Circularity Check

0 steps flagged

No circularity in derivation; synthesis conditions derived from standard FDI theory and applied experimentally

full rationale

The paper's core contribution is extending nullspace-based FDI synthesis conditions to handle closed-loop aspects for linear mechatronic systems with actuator/sensor faults. The abstract and description indicate these conditions are derived from established FDI methods and then validated on a wafer-stage prototype via real experiments. No equations or steps reduce by construction to fitted parameters, self-definitions, or load-bearing self-citations. The derivation chain remains self-contained against external FDI benchmarks, with the experimental application serving as independent validation rather than a tautology. Linearity assumptions are noted as a modeling choice but do not create circularity in the claimed results.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The work rests on standard linear systems assumptions common to FDI literature. No free parameters, invented entities, or ad-hoc axioms are indicated in the abstract.

axioms (2)
  • domain assumption The mechatronic systems are linear time-invariant
    Required for nullspace-based FDI synthesis conditions to apply directly.
  • domain assumption Closed-loop dynamics can be incorporated into the FDI synthesis without loss of the nullspace properties
    Central tailoring step stated in the abstract.

pith-pipeline@v0.9.0 · 5413 in / 1167 out tokens · 61561 ms · 2026-05-12T02:47:46.447112+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

39 extracted references · 39 canonical work pages

  1. [1]

    and Heemels, W

    Classens, K. and Heemels, W. P. M. H. and Oomen, T. , booktitle =. doi:10.1109/DTPI52967.2021.9540144 , isbn =

    work page doi:10.1109/dtpi52967.2021.9540144 2021

  2. [2]

    Classens, Koen , year =. Fault

    work page

  3. [3]

    , year =

    Gao, Zhiwei and Cecati, Carlo and Ding, Steven X. , year =. IEEE Trans. Ind. Electron , title =. doi:10.1109/TIE.2015.2417501 , issn =

    work page doi:10.1109/tie.2015.2417501 2015

  4. [4]

    and Kim, S

    Hwang, I. and Kim, S. and Kim, Y. and Seah, C. E. , year =. IEEE Trans. Control Syst. Technol. , title =. doi:10.1109/TCST.2009.2026285 , issn =

    work page doi:10.1109/tcst.2009.2026285 2009

  5. [5]

    Isermann, Rolf , year =. Annu. Rev. Control , title =. doi:10.1016/j.arcontrol.2004.12.002 , issn =

    work page doi:10.1016/j.arcontrol.2004.12.002 2004

  6. [6]

    Frank, P. M. and Ding, X. , title =. J. Process Control , year =. doi:10.1016/S0301-4770(08)60756-3 , issn =

    work page doi:10.1016/s0301-4770(08)60756-3

  7. [7]

    , journal =

    Zhang, Ping and Ding, Steven X. , journal =. An Integrated Trade-off Design of Observer Based Fault Detection Systems , doi =

    work page

  8. [8]

    Wang, J. L. and Yang, G. and Liu, J. , journal =. An. doi:10.1016/j.automatica.2007.02.019 , issn =

    work page doi:10.1016/j.automatica.2007.02.019 2007

  9. [9]

    , booktitle =

    Gertler, J. , booktitle =. Diagnosing. doi:10.1109/ACC.1995.529780 , pages =

    work page doi:10.1109/acc.1995.529780 1995

  10. [10]

    and Mostard, M

    Classens, K. and Mostard, M. and van de Wijdeven, J. and Heemels, W. P. M. H. and Oomen, T. , booktitle =. doi:10.1016/j.ifacol.2022.11.271 , title =

    work page doi:10.1016/j.ifacol.2022.11.271 2022

  11. [11]

    and Gonz

    van der Hulst, M. and Gonz. Frequency domain identification for multivariable motion control systems: Applied to a prototype wafer stage , location =. Joint 10th IFAC Symposium on Mechatronic Systems and 14th Symposium on Robotics , doi =

    work page

  12. [12]

    and Gonz

    van der Hulst, M. and Gonz. Mech. Syst. Signal Process. , year =. doi:10.1016/j.ymssp.2026.113948 , pages =

    work page doi:10.1016/j.ymssp.2026.113948 2026

  13. [13]

    Ding, S. X. and Jeinsch, T. and Frank, P. M. and Ding, E. L. , title =. Int. J. Adapt. Control Signal Process. , keywords =. doi:10.1002/1099-1115(200011)14:7<725::AID-ACS618>3.0.CO;2-Q , pages =

    work page doi:10.1002/1099-1115(200011)14:7

  14. [14]

    , journal =

    Henry, D. , journal =. Theories for Design and Analysis of Robust. doi:10.1016/j.jfranklin.2020.11.006 , issn =

    work page doi:10.1016/j.jfranklin.2020.11.006 2020

  15. [15]

    and Xie, Xiaochen and Luo, Hao , journal =

    Yin, Shen and Ding, Steven X. and Xie, Xiaochen and Luo, Hao , journal =. A. doi:10.1109/TIE.2014.2301773 , issn =

    work page doi:10.1109/tie.2014.2301773 2014

  16. [16]

    , journal =

    Lei, Yaguo and Yang, Bin and Jiang, Xinwei and Jia, Feng and Li, Naipeng and Nandi, Asoke K. , journal =. Applications of Machine Learning to Machine Fault Diagnosis:. doi:10.1016/j.ymssp.2019.106587 , issn =

    work page doi:10.1016/j.ymssp.2019.106587 2019

  17. [17]

    Springer , file =

    Varga, Andreas , year =. Springer , file =. doi:10.1007/978-3-031-35767-1 , publisher =

    work page doi:10.1007/978-3-031-35767-1

  18. [18]

    Varga, Andreas , year =. Annu. Rev. Control , title =. doi:10.1016/j.arcontrol.2013.03.001 , number =

    work page doi:10.1016/j.arcontrol.2013.03.001 2013

  19. [19]

    , journal =

    Travé-Massuyès, L. , journal =. Bridging control and artificial intelligence theories for diagnosis:. doi:10.1016/j.engappai.2013.09.018 , issn =

    work page doi:10.1016/j.engappai.2013.09.018 2013

  20. [20]

    , year =

    Ding, Steven X. , year =. Model-Based Fault Diagnosis Techniques: Design Schemes, Algorithms, and Tools , doi =. Model-based Fault Diagnosis Techniques: Design Schemes, Algorithms, and Tools , file =

    work page

  21. [21]

    In: ACC (2021).https://doi.org/10

    Classens, K. and Heemels, W. P. M. H. and Oomen, T. , booktitle =. doi:10.23919/ACC50511.2021.9482785 , isbn =

    work page doi:10.23919/acc50511.2021.9482785 2021

  22. [22]

    , title =

    Gertler, J. , title =

    work page

  23. [23]

    On Computing Achievable Fault Signatures , doi =

    Varga, Andras , journal =. On Computing Achievable Fault Signatures , doi =

    work page

  24. [24]

    On Computing Nullspace Bases — a Fault Detection Perspective , doi =

    Varga, Andras , journal =. On Computing Nullspace Bases — a Fault Detection Perspective , doi =

    work page

  25. [25]

    , journal =

    Kimura, H. , journal =. Geometric Structure of Observers for Linear Feedback Control Laws , doi =

    work page

  26. [26]

    , booktitle =

    Varga, A. , booktitle =. Reliable Algorithms for Computing Minimal Dynamic Covers , doi =

    work page

  27. [27]

    and Doyle, J

    Zhou, K. and Doyle, J. and Glover, K. , title =. doi:10.1016/s0005-1098(97)00132-5 , publisher =

    work page doi:10.1016/s0005-1098(97)00132-5

  28. [28]

    Proceedings of the IEEE Conference on Decision and Control , keywords =

    Liu, Nike and Zhou, Kemin , booktitle =. Proceedings of the IEEE Conference on Decision and Control , keywords =. 2007 , title =. doi:10.1109/CDC.2007.4434123 , eventtitle =

    work page doi:10.1109/cdc.2007.4434123 2007

  29. [29]

    doi:10.1109/CDC.2011.6160723 , booktitle =

    Glover, Keith and Varga, Andras , title =. doi:10.1109/CDC.2011.6160723 , booktitle =

    work page doi:10.1109/cdc.2011.6160723 2011

  30. [30]

    Pintelon, Rik and Schoukens, Johan , year =. System

    work page

  31. [31]

    Identification of

    Gonz. Identification of. doi:10.1016/j.automatica.2024.112013 , journal =

    work page doi:10.1016/j.automatica.2024.112013 2024

  32. [32]

    Closed-Loop Issues in System Identification , doi =. Annu. Rev. Control , keywords =

    work page

  33. [33]

    and Kordestani, R

    Rezamand, M. and Kordestani, R. and Carriveau, R. and Ting, D.S.K. and Saif, M , title =. IEEE Sens. J. , year =. doi:10.1109/JSEN.2019.2948997 , pages =

    work page doi:10.1109/jsen.2019.2948997 2019

  34. [34]

    and Bouadjenek, M.R

    Neupane, D. and Bouadjenek, M.R. and Dazeley, R. and Aryal, S. , title =. Neurocomputing , year =. doi:10.1016/j.neucom.2025.129588 , pages =

    work page doi:10.1016/j.neucom.2025.129588 2025

  35. [35]

    and Li, N

    Lei, Y. and Li, N. and Guo, L. and Ningbo, L. and Tan, T. and Lin, J. , title =. Mech. Syst. Signal Process. , year =. doi:10.1016/j.ymssp.2017.11.016 , pages =

    work page doi:10.1016/j.ymssp.2017.11.016 2017

  36. [36]

    and Deng, X

    Mian, Z. and Deng, X. and Dongg, X. and Tian, Y. and Cao, T. and Chen, K. and Al Jaber, T. , title =. Eng. Appl. Artif. Intell. , year =. doi:10.1016/j.engappai.2023.107357 , pages =

    work page doi:10.1016/j.engappai.2023.107357 2023

  37. [37]

    and Andrade, E

    Leite, D. and Andrade, E. and Rativa, D. and Maciel, A.M.A. , title =. Sensors , year =. doi:10.3390/s25010060 , pages =

    work page doi:10.3390/s25010060

  38. [38]

    and Li, H

    Li, W. and Li, H. and Gu, S. and Chen, T. , title =. Control Eng. Pract. , year =. doi:10.1016/j.conengprac.2020.104637 , pages =

    work page doi:10.1016/j.conengprac.2020.104637 2020

  39. [39]

    and Reimann, P

    Wilhelm, Y. and Reimann, P. and Gauchel, W. and Mitschang, B. , title =. Procedia CIRP , year =. doi:10.1016/j.procir.2021.03.041 , pages =

    work page doi:10.1016/j.procir.2021.03.041 2021