Roee Cohen
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Roee Cohen has a total of 14 co-authors. Here are the top co-authors by number of shared publications:
- Jacob Bortman
BGU-PHM Laboratory, Department of Mechanical Engineering, Ben-Gurion University of the Negev, P.O. Box 653, Beer Sheva 8410501, Israel
8 shared publications - Omri Matania
PHM Laboratory, Department of Mechanical Engineering, Ben-Gurion University of the Negev, Beer Sheva, Israel
8 shared publications - Lior Bachar
BGU-PHM Laboratory, Department of Mechanical Engineering, Ben-Gurion University of the Negev, Beer Sheva, Israel
5 shared publications - Renata Klein
R. K. Diagnostics, Gilon, P.O. Box 101, D. N. Misgav 2010300, Israel
2 shared publications Eric Bechhoefer
R&D Department,Green Power Monitoring Systems, Cornwall, VT, USA
1 shared publication- Eric Bechhoefer
GPMS International Inc., 93 Pilgram Place, Waterbury, VT 05676, USA
1 shared publication
Latest Publications
- Open Access
- 24 August 2026
- Published by Elsevier in SSRN Electronic Journal
Transfer across Different Sensors (TDS) using Transfer Function Compensation (TFC) for Gear Wear Classification
The development of machine-learning (ML)-based fault classification models for aerospace applications that can generalize across machines remains a significant and underexplored challenge in the literature. The difficulties stem from the high sensitivity of ML models to different acquisition settings resulting in domain shifts, and the inherent scarcity of real-world faulty-state data. This study formalizes the problem of Transfer across Different Sensors (TDS), a physically driven domain shift induced by sensor relocation within a single mechanical system. Using data from a controlled gear-wear experiment involving two sensors and varying working conditions, we demonstrate that standard ML models fail to generalize across sensor locations. To address this, a novel Transfer Function Compensation (TFC) algorithm, which models the vibration propagation path as a Linear Time-Invariant (LTI) system, is proposed. Unlike most data-driven methods, the TFC algorithm- as a deterministic physical signal conditioning step- requires no target-domain data during training to achieve effective domain adaptation. Validation results confirm that applying TFC substantially improves generalization. Overall, our findings underscore the importance of integrating physical knowledge with ML techniques to achieve zero-shot domain adaptation. Furthermore, they lay the foundation for generalizing across different machines, ultimately enabling robust fault diagnosis in real-world applications.
- 24 August 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessTransfer across Different Sensors (TDS) using Transfer Function Compensation (TFC) for Gear Wear Classification
The development of machine-learning (ML)-based fault classification models for aerospace applications that can generalize across machines remains a significant and underexplored challenge in the literature. The difficulties stem from the high sensitivity of ML models to different acquisition settings resulting in domain shifts, and the inherent scarcity of real-world faulty-state data. This study formalizes the problem of Transfer across Different Sensors (TDS), a physically driven domain shift induced by sensor relocation within a single mechanical system. Using data from a controlled gear-wear experiment involving two sensors and varying working conditions, we demonstrate that standard ML models fail to generalize across sensor locations. To address this, a novel Transfer Function Compensation (TFC) algorithm, which models the vibration propagation path as a Linear Time-Invariant (LTI) system, is proposed. Unlike most data-driven methods, the TFC algorithm- as a deterministic physical signal conditioning step- requires no target-domain data during training to achieve effective domain adaptation. Validation results confirm that applying TFC substantially improves generalization. Overall, our findings underscore the importance of integrating physical knowledge with ML techniques to achieve zero-shot domain adaptation. Furthermore, they lay the foundation for generalizing across different machines, ultimately enabling robust fault diagnosis in real-world applications.
- Open Access
- 19 September 2025
- Published by SAGE Publications in Structural Health Monitoring
Enhanced fault diagnosis of helical gears: advanced dynamic modeling and novel health indicators for early detection of tooth breakage
Helical gears are common in rotating machines, offering smoother operation and better durability compared to spur gears due to their continuous engagement and higher contact ratios. However, these advantages present significant challenges in fault detection, particularly in the early stages. Although advanced methods exist for fault detection, most are designed for spur gears, overlooking the complexities associated with helical gear dynamics. This research sheds new light on the complex physical phenomena associated with localized tooth faults in helical gears while proposing a sensitive fault detection strategy grounded in physical justification and extensive experimentation. The first study presents a new modeling approach for localized tooth breakage faults in helical gears, building on an existing validated dynamic model. This approach, validated experimentally, utilizes the multi-slice method to address the three-dimensional, time-variable contact line. For the first time, we refute the premise that fault severity in helical gears is necessarily correlated with the size of the damaged region. Instead, we provide a physically justified fault characterization through analysis of the gear mesh stiffness (GMS) and vibration signals. The second case study proposes a novel health indicator (HI) designed specifically for detecting localized tooth breakage faults in helical gears, relying on sensitive spectral analysis and energy-based feature extraction. This new HI is tested on vibration data measured through extensive controlled-degradation tests on helical gears, both in healthy conditions and with various severities of tooth breakage faults. The performance of the proposed HI is compared against traditional methods, showing significant superiority both in early and in more advanced stages of the fault. Our findings demonstrate the superiority of the new HI over traditional methods and provide crucial insights into the complex dynamic behavior of helical gears, offering practical value for improving reliability and safety in various engineering applications, such as in automotive, aerospace, and industrial machinery, and also paving the way for advanced fault severity estimation.
- 19 September 2025
- Published by SAGE Publications in Structural Health Monitoring
Open AccessEnhanced fault diagnosis of helical gears: advanced dynamic modeling and novel health indicators for early detection of tooth breakage
Helical gears are common in rotating machines, offering smoother operation and better durability compared to spur gears due to their continuous engagement and higher contact ratios. However, these advantages present significant challenges in fault detection, particularly in the early stages. Although advanced methods exist for fault detection, most are designed for spur gears, overlooking the complexities associated with helical gear dynamics. This research sheds new light on the complex physical phenomena associated with localized tooth faults in helical gears while proposing a sensitive fault detection strategy grounded in physical justification and extensive experimentation. The first study presents a new modeling approach for localized tooth breakage faults in helical gears, building on an existing validated dynamic model. This approach, validated experimentally, utilizes the multi-slice method to address the three-dimensional, time-variable contact line. For the first time, we refute the premise that fault severity in helical gears is necessarily correlated with the size of the damaged region. Instead, we provide a physically justified fault characterization through analysis of the gear mesh stiffness (GMS) and vibration signals. The second case study proposes a novel health indicator (HI) designed specifically for detecting localized tooth breakage faults in helical gears, relying on sensitive spectral analysis and energy-based feature extraction. This new HI is tested on vibration data measured through extensive controlled-degradation tests on helical gears, both in healthy conditions and with various severities of tooth breakage faults. The performance of the proposed HI is compared against traditional methods, showing significant superiority both in early and in more advanced stages of the fault. Our findings demonstrate the superiority of the new HI over traditional methods and provide crucial insights into the complex dynamic behavior of helical gears, offering practical value for improving reliability and safety in various engineering applications, such as in automotive, aerospace, and industrial machinery, and also paving the way for advanced fault severity estimation.
- Open Access
- 10 August 2025
- Published by SAGE Publications in Structural Health Monitoring
Novel approaches of synchronous averaging of gear and bearing vibrations
Synchronous averaging (SA) is a fundamental technique in synchronous vibration analysis. This article begins by reviewing the conventional SA method and offering new insights into its underlying characteristics. We then frame SA as a special case of identifying the signal that minimizes the mean squared error among a set of segments. Building on this, we present a theoretical analysis of how speed fluctuations impact the accuracy of the computed synchronous average. Finally, we propose a novel algorithm—Angular Synchronization SA—that enables accurate SA for bearing signals, even under fluctuating speed conditions. The effectiveness of the proposed algorithm is demonstrated through three experiments involving bearings with outer race spalls. Given that bearings are critical components in rotating machinery and are prone to failure, our approach has the potential to significantly advance vibration-based diagnostics. By enabling direct analysis of their synchronous average, the proposed method paves the way for improved diagnostic capabilities.
- 10 August 2025
- Published by SAGE Publications in Structural Health Monitoring
Open AccessNovel approaches of synchronous averaging of gear and bearing vibrations
Synchronous averaging (SA) is a fundamental technique in synchronous vibration analysis. This article begins by reviewing the conventional SA method and offering new insights into its underlying characteristics. We then frame SA as a special case of identifying the signal that minimizes the mean squared error among a set of segments. Building on this, we present a theoretical analysis of how speed fluctuations impact the accuracy of the computed synchronous average. Finally, we propose a novel algorithm—Angular Synchronization SA—that enables accurate SA for bearing signals, even under fluctuating speed conditions. The effectiveness of the proposed algorithm is demonstrated through three experiments involving bearings with outer race spalls. Given that bearings are critical components in rotating machinery and are prone to failure, our approach has the potential to significantly advance vibration-based diagnostics. By enabling direct analysis of their synchronous average, the proposed method paves the way for improved diagnostic capabilities.
- Open Access
Machine Health Indicators and Digital Twins
Health indicators (HIs) are quantitative indices that assess the condition of engineering systems by linking sensor data with monitoring, diagnostic, and prognostic methods to estimate the remaining useful life (RUL). Digital twins (DTs), which serve as digital representations of physical assets, enhance system monitoring, diagnostics, and prognostics by operationalizing analytic capabilities derived from sensor data. This paper explores the integration of HIs and DTs, illustrating their roles in condition-based maintenance and structural health monitoring. The methodologies discussed span data-driven and physics-based approaches, emphasizing their applications in rotary machinery, including bearings and gears. These approaches not only detect anomalies but also predict system failures through advanced modeling and machine learning (ML) techniques. The paper provides examples of HIs derived from vibration analysis and soft sensors and maps future research directions for improving health monitoring systems through hybrid modeling and uncertainty quantification. It concludes by addressing the challenges of data labeling and uncertainties and the role of HIs in advancing performance engineering, making DTs a pivotal tool in predictive maintenance strategies.
Open AccessMachine Health Indicators and Digital Twins
Health indicators (HIs) are quantitative indices that assess the condition of engineering systems by linking sensor data with monitoring, diagnostic, and prognostic methods to estimate the remaining useful life (RUL). Digital twins (DTs), which serve as digital representations of physical assets, enhance system monitoring, diagnostics, and prognostics by operationalizing analytic capabilities derived from sensor data. This paper explores the integration of HIs and DTs, illustrating their roles in condition-based maintenance and structural health monitoring. The methodologies discussed span data-driven and physics-based approaches, emphasizing their applications in rotary machinery, including bearings and gears. These approaches not only detect anomalies but also predict system failures through advanced modeling and machine learning (ML) techniques. The paper provides examples of HIs derived from vibration analysis and soft sensors and maps future research directions for improving health monitoring systems through hybrid modeling and uncertainty quantification. It concludes by addressing the challenges of data labeling and uncertainties and the role of HIs in advancing performance engineering, making DTs a pivotal tool in predictive maintenance strategies.
- Open Access
- 20 January 2025
- Published by Elsevier in Engineering Failure Analysis
- Vol. 170
Few-shot learning for estimating gear wear severity towards digital twinning
- 20 January 2025
- Published by Elsevier in Engineering Failure Analysis
- Vol. 170
Open AccessFew-shot learning for estimating gear wear severity towards digital twinning
- Open Access
- 13 January 2025
- Published by SAGE Publications in Structural Health Monitoring
- Vol. 25 (2) , 1409-1423
Anomaly detection of gear wear: introducing a novel health indicator through experimentation and physical investigation
Monitoring gear wear is important for gear diagnosis, although it remains a challenging endeavor. Unlike localized tooth faults, there are still physical patterns in the vibration signature associated with distributed wear faults that should be unfolded. This study contributes a novel framework for gear wear diagnosis, supported by physical principles and validated against traditional methods through extensive experimentation. First, we characterize gear wear in the vibration signature, utilizing data from dozens of seeded, realistic degrading wear cases across various rotational speeds collected through unique controlled-degradation tests. We examine conventional features that have proven useful for diagnosing localized tooth faults and demonstrate their inadequate performance in diagnosing gear wear. We propose a sensitive spectral analysis of the gear mesh and sideband spectral energy in the spectrum to recognize wear manifestation and monitor its degradation. Drawing on the insights gained in this study, we introduce a novel health indicator for anomaly detection of incipient gear wear, grounded in physical understanding and employing sophisticated yet simple feature engineering techniques. We analyze the performance of the proposed health indicator, demonstrating its superiority over state-of-the-art methods, and underscore the importance of considering assembly and dismantling operations to avoid the issue of test-training leakage. Additionally, we highlight the potential of the proposed physical-based framework combined with a deep-learning approach to detect wear in its early stages. To the best of our knowledge, this is the first study to conduct extensive controlled-degradation tests, rather than endurance tests, to investigate gear wear and offer a hybrid physical data-driven framework that combines advanced signal processing, feature extraction, and feature engineering techniques for early anomaly detection.
- 13 January 2025
- Published by SAGE Publications in Structural Health Monitoring
- Vol. 25 (2) , 1409-1423
Open AccessAnomaly detection of gear wear: introducing a novel health indicator through experimentation and physical investigation
Monitoring gear wear is important for gear diagnosis, although it remains a challenging endeavor. Unlike localized tooth faults, there are still physical patterns in the vibration signature associated with distributed wear faults that should be unfolded. This study contributes a novel framework for gear wear diagnosis, supported by physical principles and validated against traditional methods through extensive experimentation. First, we characterize gear wear in the vibration signature, utilizing data from dozens of seeded, realistic degrading wear cases across various rotational speeds collected through unique controlled-degradation tests. We examine conventional features that have proven useful for diagnosing localized tooth faults and demonstrate their inadequate performance in diagnosing gear wear. We propose a sensitive spectral analysis of the gear mesh and sideband spectral energy in the spectrum to recognize wear manifestation and monitor its degradation. Drawing on the insights gained in this study, we introduce a novel health indicator for anomaly detection of incipient gear wear, grounded in physical understanding and employing sophisticated yet simple feature engineering techniques. We analyze the performance of the proposed health indicator, demonstrating its superiority over state-of-the-art methods, and underscore the importance of considering assembly and dismantling operations to avoid the issue of test-training leakage. Additionally, we highlight the potential of the proposed physical-based framework combined with a deep-learning approach to detect wear in its early stages. To the best of our knowledge, this is the first study to conduct extensive controlled-degradation tests, rather than endurance tests, to investigate gear wear and offer a hybrid physical data-driven framework that combines advanced signal processing, feature extraction, and feature engineering techniques for early anomaly detection.
- 5 November 2024
- Published by Elsevier in Mechanical Systems and Signal Processing
- Vol. 224
Zero-fault-shot learning for bearing spall type classification by hybrid approach
- 5 November 2024
- Published by Elsevier in Mechanical Systems and Signal Processing
- Vol. 224
Zero-fault-shot learning for bearing spall type classification by hybrid approach
- Open Access
- 23 March 2024
- Published by Springer Nature in Nonlinear Dynamics
- Vol. 112 (10) , 8089-8107
A novel method for helical gear modeling with an experimental validation
Dynamic models are important for developing gear diagnostics methods since they allow physical phenomena occurring during operation to be studied in a relatively simple environment. The main challenge in gear modeling is the calculation of the time-variant gear mesh stiffness, and this challenge is even greater in helical gears. The mechanism of helical gears is more complex than in spur gears; the helix angle both adds an axial component to the contact force and also makes the contact line three-dimensional. This study suggests a novel dynamic model for helical gear vibrations that combines an existing validated dynamic model for spur gears with a unique extension for helical gears. The extension is based on a common method called “multi-slice”, according to which the helical tooth width is divided into infinitesimal slices, and each slice is treated as spur tooth. The suggested model introduces a novel implementation of the multi-slice method that overcomes the aforementioned challenges with only few parameters and calculations, depends on the tooth geometry. Furthermore, for the first time in helical gear modeling, the manufacturing profile errors are integrated to the model to generate scatter in the data that can better reflect the reality. The model is validated experimentally and for two different test-rigs by a qualitative comparison of the RMS of the vibration signal. The simulations and the measured data show similar behavior at different ranges of rotational speed and applied load, emphasizing the potential inherent in the model for future work on gear fault diagnosis.
- 23 March 2024
- Published by Springer Nature in Nonlinear Dynamics
- Vol. 112 (10) , 8089-8107
Open AccessA novel method for helical gear modeling with an experimental validation
Dynamic models are important for developing gear diagnostics methods since they allow physical phenomena occurring during operation to be studied in a relatively simple environment. The main challenge in gear modeling is the calculation of the time-variant gear mesh stiffness, and this challenge is even greater in helical gears. The mechanism of helical gears is more complex than in spur gears; the helix angle both adds an axial component to the contact force and also makes the contact line three-dimensional. This study suggests a novel dynamic model for helical gear vibrations that combines an existing validated dynamic model for spur gears with a unique extension for helical gears. The extension is based on a common method called “multi-slice”, according to which the helical tooth width is divided into infinitesimal slices, and each slice is treated as spur tooth. The suggested model introduces a novel implementation of the multi-slice method that overcomes the aforementioned challenges with only few parameters and calculations, depends on the tooth geometry. Furthermore, for the first time in helical gear modeling, the manufacturing profile errors are integrated to the model to generate scatter in the data that can better reflect the reality. The model is validated experimentally and for two different test-rigs by a qualitative comparison of the RMS of the vibration signal. The simulations and the measured data show similar behavior at different ranges of rotational speed and applied load, emphasizing the potential inherent in the model for future work on gear fault diagnosis.
- 12 September 2023
- Published by Elsevier in Mechanical Systems and Signal Processing
- Vol. 204
A novel hybrid physical AI-based strategy for fault severity estimation in spur gears with zero-shot learning
- 12 September 2023
- Published by Elsevier in Mechanical Systems and Signal Processing
- Vol. 204
A novel hybrid physical AI-based strategy for fault severity estimation in spur gears with zero-shot learning