Olga Fink

Last Name

Fink

First Name

Olga

ORCID

0000-0002-9546-1488

Organisational unit

01359 - Lehre Management, Technologie u. Ök.

Show all metadata

Search Results

Publications 1 - 10 of 53

Fusing physics-based and deep learning models for prognostics
Arias Chao, Manuel; Kulkarni, Chetan; Goebel, Kai; et al. (2022)
Reliability Engineering & System Safety
Physics-based and data-driven models for remaining useful lifetime (RUL) prediction typically suffer from two major challenges that limit their applicability to complex real-world domains: (1) the incompleteness of physics-based models and (2) the limited representativeness of the training dataset for data-driven models. Combining the advantages of these two approaches while overcoming some of their limitations, we propose a novel hybrid framework for fusing the information from physics-based performance models with deep learning algorithms for prognostics of complex safety-critical systems. In the proposed framework, we use physicsbased performance models to infer unobservable model parameters related to a system’s components health by solving a calibration problem. These parameters are subsequently combined with sensor readings and used as input to a deep neural network, thereby generating a data-driven prognostics model with physics-augmented features. The performance of the hybrid framework is evaluated on an extensive case study comprising runto-failure degradation trajectories from a fleet of nine turbofan engines under real flight conditions. The experimental results show that the hybrid framework outperforms purely data-driven approaches by extending the prediction horizon by nearly 127%. Furthermore, it requires less training data and is less sensitive to the limited representativeness of the dataset as compared to purely data-driven approaches. Furthermore, we demonstrated the feasibility of the proposed framework on the original CMAPSS dataset, thereby confirming its superior performance.
Generalization of an Encoder-Decoder LSTM model for flood prediction in ungauged catchments
Zhang, Yikui; Ragettli, Silvan; Molnar, Peter; et al. (2022)
Journal of Hydrology
Flood prediction in ungauged catchments is usually conducted by hydrological models that are parameterized based on nearby and similar gauged catchments. As an alternative to this process-based modelling, deep learning (DL) models have demonstrated their ability for prediction in ungauged catchments (PUB) with high efficiency. Catchment characteristics, the number of gauged catchments, and their level of hydroclimatic heterogeneity in the training dataset used for model regionalization can directly affect the model's performance. Here, we study the generalization ability of a DL model to these factors by applying an Encoder-Decoder Long Short-Term Memory neural network for a 6-hour lead-time runoff prediction in 35 mountainous catchments in China. By varying the available number of catchments and model settings with different training datasets, namely local, regional, and PUB models, we evaluated the generalization ability of our model. We found that both quantity (i.e. number of gauged catchments available) and heterogeneity of the training dataset used for the DL model are important for improving model performance in the PUB context, due to a data synergy effect. The assessment of the sensitivity to catchment characteristics showed that the model performance is mainly correlated to the local hydro-climatic conditions; the more arid the region, the more likely it is to have a poor model performance for prediction in ungauged catchments. The results suggest that the regional ED-LSTM model is a promising method to predict streamflow from rainfall inputs in PUB, and outline the need for preparing a representative training dataset.
Continuous Health State Monitoring of High-Voltage Circuit Breakers
Hsu, Chi-Ching; Frusque, Gaëtan; Fink, Olga; et al. (2025)
IEEE Access
Circuit breakers (CBs) are renowned for their high reliability and long lifespan. As a result, many CBs installed decades ago are now approaching their predefined end of service life. However, this predefined service life does not always reflect the actual condition, as some may still be far from reaching their true end-of-life, depending on their operating conditions and history. To assess their true lifetime, continuous condition monitoring is essential. While previous studies have effectively demonstrated the ability to distinguish between different CB fault types, the evolution of CB degradation remains unclear when faults are artificially introduced. This paper investigates the health condition of two high-voltage CBs continuously through run-to-failure experiments. A comprehensive dataset was collected for all opening and closing operations with various sensors such as vibration, coil current, and travel curve and has been made publicly available for further analysis. Furthermore, features were derived from the sensor data, revealing distinct degradation trajectories over time that can be used to monitor the condition of the CBs. This paper highlights the degradation patterns of these features, some of which are well-suited for continuous condition monitoring due to their gradual changing trend over time that likely correlates with the true degradation condition, while others are less useful as they show abrupt changes only before or at failure. By leveraging these features, we can progress beyond the focus of previous research using only fault diagnosis towards fault prognosis. This shift opens the possibility for accurate prediction of the CB condition over time, enabling more effective maintenance strategies.
DARE-GRAM : Unsupervised Domain Adaptation Regression by Aligning Inverse Gram Matrices
Nejjar, Ismail; Wang, Qin; Fink, Olga (2023)
2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
Unsupervised Domain Adaptation Regression (DAR) aims to bridge the domain gap between a labeled source dataset and an unlabelled target dataset for regression problems. Recent works mostly focus on learning a deep feature encoder by minimizing the discrepancy between source and target features. In this work, we present a different perspective for the DAR problem by analyzing the closed-form ordinary least square (OLS) solution to the linear regressor in the deep domain adaptation context. Rather than aligning the original feature embedding space, we propose to align the inverse Gram matrix of the features, which is motivated by its presence in the OLS solution and the Gram matrix's ability to capture the feature correlations. Specifically, we propose a simple yet effective DAR method which leverages the pseudo-inverse low-rank property to align the scale and angle in a selected sub-space generated by the pseudo-inverse Gram matrix of the two domains. We evaluate our method on three domain adaptation regression benchmarks. Experimental results demonstrate that our method achieves state-of-the-art performance. Our code is available at https://github.com/ismailnejjar/DARE-GRAM.
Explainable AI guided unsupervised fault diagnostics for high-voltage circuit breakers
Hsu, Chi-Ching; Frusque, Gaëtan; Forest, Florent; et al. (2025)
Reliability Engineering & System Safety
Commercial high-voltage circuit breaker (CB) condition monitoring systems rely on directly observable physical parameters such as gas filling pressure with pre-defined thresholds. While these parameters are crucial, they only cover a small subset of malfunctioning mechanisms and usually can be monitored only if the CB is disconnected from the grid. To facilitate online condition monitoring while CBs remain connected, non-intrusive measurement techniques such as vibration or acoustic signals are necessary. Currently, CB condition monitoring studies using these signals typically utilize supervised methods for fault diagnostics, where ground-truth fault types are known due to artificially introduced faults in laboratory settings. This supervised approach is however not feasible in real-world applications, where fault labels are unavailable. In this work, we propose a novel unsupervised fault detection and segmentation framework for CBs based on vibration and acoustic signals. This framework can detect deviations from the healthy state. The explainable artificial intelligence (XAI) approach is applied to the detected faults for fault diagnostics. The specific contributions are: (1) we propose an integrated unsupervised fault detection and segmentation framework that is capable of detecting faults and clustering different faults with only healthy data required during training (2) we provide an unsupervised explainability-guided fault diagnostics approach using XAI to offer domain experts potential indications of the aged or faulty components, achieving fault diagnostics without the prerequisite of ground-truth fault labels. These contributions are validated using an experimental dataset from a high-voltage CB under healthy and artificially introduced fault conditions, contributing to more reliable CB system operation.
Uncertainty quantification in machine learning for engineering design and health prognostics: A tutorial
Nemani, Venkat; Biggio, Luca; Huan, Xun; et al. (2023)
Mechanical Systems and Signal Processing
On top of machine learning (ML) models, uncertainty quantification (UQ) functions as an essential layer of safety assurance that could lead to more principled decision making by enabling sound risk assessment and management. The safety and reliability improvement of ML models empowered by UQ has the potential to significantly facilitate the broad adoption of ML solutions in high-stakes decision settings, such as healthcare, manufacturing, and aviation, to name a few. In this tutorial, we aim to provide a holistic lens on emerging UQ methods for ML models with a particular focus on neural networks and the applications of these UQ methods in tackling engineering design as well as prognostics and health management problems. Towards this goal, we start with a comprehensive classification of uncertainty types, sources, and causes pertaining to UQ of ML models. Next, we provide a tutorial-style description of several state-of-the-art UQ methods: Gaussian process regression, Bayesian neural network, neural network ensemble, and deterministic UQ methods focusing on spectral-normalized neural Gaussian process. Established upon the mathematical formulations, we subsequently examine the soundness of these UQ methods quantitatively and qualitatively (by a toy regression example) to examine their strengths and shortcomings from different dimensions. Then, we review quantitative metrics commonly used to assess the quality of predictive uncertainty in classification and regression problems. Afterward, we discuss the increasingly important role of UQ of ML models in solving challenging problems in engineering design and health prognostics. Two case studies with source codes available on GitHub are used to demonstrate these UQ methods and compare their performance in the life prediction of lithium-ion batteries at the early stage (case study 1) and the remaining useful life prediction of turbofan engines (case study 2).
Towards Multimodal Open-Set Domain Generalization and Adaptation through Self-supervision
Dong, Hao; Chatzi, Eleni; Fink, Olga (2024)
arXiv
The task of open-set domain generalization (OSDG) involves recognizing novel classes within unseen domains, which becomes more challenging with multiple modalities as input. Existing works have only addressed unimodal OSDG within the meta-learning framework, without considering multimodal scenarios. In this work, we introduce a novel approach to address Multimodal Open-Set Domain Generalization (MM-OSDG) for the first time, utilizing self-supervision. To this end, we introduce two innovative multimodal self-supervised pretext tasks: Masked Cross-modal Translation and Multimodal Jigsaw Puzzles. These tasks facilitate the learning of multimodal representative features, thereby enhancing generalization and open-class detection capabilities. Additionally, we propose a novel entropy weighting mechanism to balance the loss across different modalities. Furthermore, we extend our approach to tackle also the Multimodal Open-Set Domain Adaptation (MM-OSDA) problem, especially in scenarios where unlabeled data from the target domain is available. Extensive experiments conducted under MM-OSDG, MM-OSDA, and Multimodal Closed-Set DG settings on the EPIC-Kitchens and HAC datasets demonstrate the efficacy and versatility of the proposed approach. Our source code is available at https://github.com/donghao51/MOOSA.
MultiOOD: Scaling Out-of-Distribution Detection for Multiple Modalities
Dong, Hao; Zhao, Yue; Chatzi, Eleni; et al. (2024)
arXiv
Detecting out-of-distribution (OOD) samples is important for deploying machine learning models in safety-critical applications such as autonomous driving and robot-assisted surgery. Existing research has mainly focused on unimodal scenarios on image data. However, real-world applications are inherently multimodal, which makes it essential to leverage information from multiple modalities to enhance the efficacy of OOD detection. To establish a foundation for more realistic Multimodal OOD Detection, we introduce the first-of-its-kind benchmark, MultiOOD, characterized by diverse dataset sizes and varying modality combinations. We first evaluate existing unimodal OOD detection algorithms on MultiOOD, observing that the mere inclusion of additional modalities yields substantial improvements. This underscores the importance of utilizing multiple modalities for OOD detection. Based on the observation of Modality Prediction Discrepancy between in-distribution (ID) and OOD data, and its strong correlation with OOD performance, we propose the Agree-to-Disagree (A2D) algorithm to encourage such discrepancy during training. Moreover, we introduce a novel outlier synthesis method, NP-Mix, which explores broader feature spaces by leveraging the information from nearest neighbor classes and complements A2D to strengthen OOD detection performance. Extensive experiments on MultiOOD demonstrate that training with A2D and NP-Mix improves existing OOD detection algorithms by a large margin. Our source code and MultiOOD benchmark are available at https://github.com/donghao51/MultiOOD.
Contrastive feature learning for railway infrastructure fault diagnostic
Rombach, Katharina; Michau, Gabriel; Ratnasabapathy, Kajan; et al. (2022)
Proceedings of the 32nd European Safety and Reliability Conference (ESREL 2022)
To operate the railway system safely and efficiently, a multitude of assets need to me monitored. Railway sleepers are one of these infrastructure assets, that are safety critical. To automate the monitoring process, data-driven fault diagnostics models have shown great potential. However, in practice, the performance of data-driven models can be compromised if the training dataset is not representative of all possible future conditions. Environmental factors, for example, can continuously change and cause variations in the condition monitoring data. The caused variations can harm the performance of a data-driven diagnostics models if (1) they are not represented in the training datatset and especially, if (2) they are larger compared to variations caused by a change in the health condition. We propose to approach this problem by learning a feature representation that is, on the one hand, invariant to operating or environmental factors but, on the other hand, sensitive to changes in the asset's health condition. We evaluate how contrastive learning can be employed to tackle this challenge on a supervised diagnostics task given a real condition monitoring dataset of railway sleepers.We evaluate the performance of supervised contrastive feature learning given a labeled image dataset that is collected by a diagnostic vehicle. Our results demonstrate that contrastive feature learning significantly improves the performance on the supervised classification task regarding sleepers compared to a state-of-the-art method.
Unseen Visual Anomaly Generation
Sun, Han; Cao, Yunkang; Dong, Hao; et al. (2025)
2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
Visual anomaly detection (AD) presents significant challenges due to the scarcity of anomalous data samples. While numerous works have been proposed to synthesize anomalous samples, these synthetic anomalies often lack authenticity or require extensive training data, limiting their applicability in real-world scenarios. In this work, we propose Anomaly Anything (AnomalyAny), a novel framework that leverages Stable Diffusion (SD)’s image generation capabilities to generate diverse and realistic unseen anomalies. By conditioning on a single normal sample during test time, AnomalyAny is able to generate unseen anomalies for arbitrary object types with text descriptions. Within AnomalyAny, we propose attention-guided anomaly optimization to direct SD’s attention on generating hard anomaly concepts. Additionally, we introduce prompt-guided anomaly refinement, incorporating detailed descriptions to further improve the generation quality. Extensive experiments on MVTec AD and VisA datasets demonstrate AnomalyAny’s ability in generating high-quality unseen anomalies and its effectiveness in enhancing downstream AD performance. Our demo and code are available at https://hansunhayden.github.io/CUT.github.io/.

Publications 1 - 10 of 53

Olga Fink

Last Name

First Name

ORCID

Organisational unit

Filters

Search Results