Open research questions in Fault Detection and Control Systems
97 unresolved questions extracted from the limitations and future-work sections of 619 Fault Detection and Control Systems papers in our library. Each links back to the study that raised it.
What the literature leaves open
The presence of non-stationary features in industrial processes. The lack of effective methods for root cause diagnosis in non-stationary industrial processes. The need for a framework that can accurately identify root causes and infer fault propagation paths in non-stationary industrial processes.
Adaptive Moving-Window Dual-Test Granger Causality for Root Cause Diagnosis of Non-Stationary Industrial Processes · 2026 · DOIThe paper does not provide a real-world application with a large sample size. The simulation studies are limited to specific settings. The method is not compared with other existing methods except Dlasso.
The desparsified LASSO faces an additional challenge from nonstationary regressors modeled as local unit roots. No research has solved the question of how to conduct valid inference for a regressor of primary interest in a high dimensional linear predictive regression model.
The algorithm is tested on a simulated dataset, - The number of iterations is limited to 25000, - The algorithm may not reach a value within 1% of the reference value after 25000 iterations
Robust SGLD Algorithm for Solving Non-convex Distributionally Robust Optimisation Problems · 2026 · DOIThe lack of effective algorithms for solving non-convex distributionally robust stochastic optimisation problems. The need for a non-asymptotic upper bound on the excess risk of the algorithm. The challenge of mitigating the effect of outliers in stochastic optimisation.
Robust SGLD Algorithm for Solving Non-convex Distributionally Robust Optimisation Problems · 2026 · DOIThe paper does not provide a comprehensive comparison with existing model selection methods. The theoretical results are limited to finite-sample robustness guarantees. The implementation of CROMS under a continuous model class is not fully explored. The paper does not discuss the computational complexity of the proposed frameworks.
Exploring CROMS under a continuous model class. Investigating the application of gradient-based methods to solve the ERM problems over the model index λ. Adopting implicit differentiation techniques for convex programs to compute the gradients of ϕ(Y i , z λ (X i )) with respect to λ.
The lack of models for data-driven flaw discovery is a challenge. The complexity of mathematical analysis and the need for external devices are also challenges. The incorporation of non-linear valves or transmitters with temporal delay is another challenge identified in the paper.
One further thing that should be taken into consideration for future research is the incorporation of non-linear valves or transmitters that have a temporal delay and the investigation of their behavior, - The study in question suggests using the approach of resilient transition structure for defect identification and diagnostics
Future research should focus on improving the efficiency of the method for real-time applications. Future research should explore the application of the method to other types of valves and systems. Future research should investigate the use of other machine learning algorithms for fault detection in regulating valves.
A Multistage Physics-Informed Neural Network for Fault Detection in Regulating Valves of Nuclear Power Plants · 2024 · DOIWhen compared with state-of-the-art domain adaptation methods across three datasets, FDCycle-GAN demonstrates its superiority by achieving higher fault diagnosis accuracy in tasks where fault samples are scarce.
AE-based methods face a major limitation: they require retraining whenever a sensor fails or is temporarily unavailable, which limits their practicality in large-scale systems.
Enhanced Adaptive Autoencoder Framework for Continuous Monitoring and Fault Management in Industrial Sensors · 2025 · DOICurrent efforts predominantly focus on classification model calibration, leaving the sequence recognition model calibration analysis underexplored and challenging.
Furthermore, model interpretability, as an essential aspect for practical deployment and decision-making, remains insufficiently explored.
Fine-Grained Time and Hidden Feature Learning for Interpretable Hard Landing Prediction Based on QAR Data · 2025 · DOIThe FD problem has garnered great interest in industrial application, yet methods for integrating process risk into the detection procedure are still scarce.
Risk-Based Fault Detection Using Bayesian Networks Based on Failure Mode and Effect Analysis · 2024 · DOIThis inclusion addresses the diminished predictive performance of soft sensors when labeled samples are insufficient.
RegGAN: A Virtual Sample Generative Network for Developing Soft Sensors with Small Data · 2024 · DOIThe use of data-driven methods for these tasks is limited by the requirement of physically consistent outcomes, particularly in safety-critical systems.
A Multistage Physics-Informed Neural Network for Fault Detection in Regulating Valves of Nuclear Power Plants · 2024 · DOITherefore, it is crucial to effectively extract representative features and identify faults in a scenario with limited data.
Elastic Slow Feature Prototypical Network for Few-Shot Fault Diagnosis of Industrial Processes · 2024 · DOIAlthough deep learning-based detection technology has significantly improved the efficiency and accuracy of fault diagnosis, its development is limited by the differences in sample distribution caused by operating condition changes.
The complexity of nuclear power plants as nonlinear systems with high safety requirements. The lack of effective fault diagnosis methods for cross-reactor-type scenarios. The need to reduce operation and maintenance costs while ensuring safety.
Research on a strongly generalizable fault diagnosis method based on adversarial transfer learning · 2026 · DOIThe low efficiency of shallow machine learning algorithms in fault diagnosis under small-sample and unlabeled data conditions. The lack of effective fault diagnosis methods for cross-reactor-type scenarios.
Research on a strongly generalizable fault diagnosis method based on adversarial transfer learning · 2026 · DOIFuture work will focus on extending the proposed approach to online fault diagnosis, investigating lightweight attention mechanisms to further reduce computational overhead, and evaluating generalization performance across different operating modes and industrial processes. First, the evaluation was conducted using an offline benchmark dataset, and the model's behaviour under real-time industrial deployment conditions was not investigated.
Compact CNN-Transformer Model for Robust Fault Diagnosis in Large-Scale Chemical Processes · 2026 · DOITo apply the proposed approach to other systems with linear discrete time-varying dynamics. To investigate the use of other optimization techniques for designing robust residual generators.
There is a need for effective fault diagnosis systems for unmanned aerial vehicles. Existing approaches may not be suitable for systems with linear discrete time-varying dynamics.
Large scale and strong nonlinearity of chemical processes. High complexity of chemical processes. Limited interpretability of data-driven fault diagnosis models.
A nonlinear Takagi-Sugeno-Kang fuzzy system based on information granules for chemical engineering fault diagnosis · 2026 · DOI
Most-cited papers in Fault Detection and Control Systems
- Diagnosing multiple faults · Artificial Intelligence · 1987 · 1,195 citations
- Safe Learning in Robotics: From Learning-Based Control to Safe Reinforcement Learning · Annual Review of Control Robotics and Autonomous Systems · 2022 · 696 citations
- A Review on Soft Sensors for Monitoring, Control, and Optimization of Industrial Processes · IEEE Sensors Journal · 2020 · 595 citations
- Combining discrepant diagnostic information from multiple sources: Are complex algorithms better than simple ones? · Journal of Abnormal Child Psychology · 1992 · 311 citations
- New approach for the diagnosis of extractions with neural network machine learning · American Journal of Orthodontics and Dentofacial Orthopedics · 2015 · 228 citations
- Sensor-Fault Detection, Isolation and Accommodation for Digital Twins via Modular Data-Driven Architecture · IEEE Sensors Journal · 2020 · 223 citations
- Dual processing and diagnostic errors · Advances in Health Sciences Education · 2009 · 196 citations
- Attention-based deep meta-transfer learning for few-shot fine-grained fault diagnosis · Knowledge-Based Systems · 2023 · 196 citations
- The alarm problem and directed attention in dynamic fault management · Ergonomics · 1995 · 194 citations
- Knowledge-Based Fault Diagnosis in Industrial Internet of Things: A Survey · IEEE Internet of Things Journal · 2022 · 192 citations
Most recent work
- SMNet: A Novel Compositional Generalization Model for Industrial Robot Multijoint Fault Diagnosis · IEEE Internet of Things Journal · 2026
- Causal Graph Spatial--Temporal Autoencoder for Reliable and Interpretable Process Monitoring · IEEE Transactions on Neural Networks and Learning Systems · 2026
- A Multi-Source Attention Graph Neural Network for modeling long and short-term dependencies in chemical process forecasting · Advanced Engineering Informatics · 2026
- Optimality of the mean-square value, skewness, kurtosis, and synentropy for detection of weak additive fault signatures in (nearly) Gaussian noise · Mechanical Systems and Signal Processing · 2026
- CausalFPS: A causal discovery-based method for UAV flight parameter selection · Mechanical Systems and Signal Processing · 2026
- Multi-dimensional sequence embedding and improved Informer for prediction of industrial alarm events · Reliability Engineering & System Safety · 2026
- Decision Transformers for RIS-Assisted Systems With Diffusion Model-Based Channel Acquisition · IEEE Transactions on Wireless Communications · 2026
- Optimal Fault Detection for Stochastic Time-Varying Systems via Hotelling’s $T^{2}$ Statistic · IEEE Transactions on Industrial Electronics · 2026
- Dynamic sparse vector autoregressive moving average model for online fault detection in liquid rocket engines · Aerospace Science and Technology · 2026
- Graph autoencoder with causal relationship inference for fault detection and root cause identification in complex industrial process · Advanced Engineering Informatics · 2026
Find a gap in your own Fault Detection and Control Systems sub-topic
This page shows what the Fault Detection and Control Systems literature already flags as unresolved. To narrow it to your specific question, search the Research Gap Finder: the search is free with a free account and lists the papers closest to your topic first. Unlocking that topic (50 credits, charged once) fills the comparison table from our 4.5M-paper local library and writes the gaps from its rows.
Open the Research Gap Finder →