scholarly journals Robust Predictive Maintenance for Robotics via Unsupervised Transfer Learning

2021 ◽  
Vol 34 (1) ◽  
Author(s):  
Arash Golibagh Mahyari ◽  
Thomas Locher
Keyword(s):  
Transfer Learning ◽  
Negative Impact ◽  
Industrial Robots ◽  
Predictive Maintenance ◽  
False Alarms ◽  
Mechanical Condition ◽  
Condition Change ◽  
Maintenance Systems ◽  
Real World Datasets ◽  
Main Shortcoming

Industrial robots play an increasingly important role in a growing number of fields. Since the breakdown of a single robot may have a negative impact on the entire process, predictive maintenance systems have gained importance as an essential component of robotics service offerings. The main shortcoming of such systems is that features extracted from a task typically differ significantly from the learnt model of a different task, incurring false alarms. In this paper, we propose a novel solution based on transfer learning which addresses a well-known challenge in predictive maintenance algorithms by passing the knowledge of the trained model from one task to another in order to prevent the need for retraining and to eliminate such false alarms. The deployment of the proposed algorithm on real-world datasets demonstrates that the algorithm can not only distinguish between tasks and mechanical condition change, it further yields a sharper deviation from the trained model in case of a mechanical condition change and thus detects mechanical issues with higher confidence.

scholarly journals Reliable Learning with PDE-Based CNNs and DenseNets for Detecting COVID-19, Pneumonia, and Tuberculosis from Chest X-Ray Images

Mathematics ◽  
2021 ◽  
Vol 9 (4) ◽  
pp. 434
Author(s):  
Anca Nicoleta Marginean ◽  
Delia Doris Muntean ◽  
George Adrian Muntean ◽  
Adelina Priscu ◽  
Adrian Groza ◽  
...  
Keyword(s):  
Transfer Learning ◽  
Negative Impact ◽  
X Ray ◽  
Domain Specific ◽  
Direct Training ◽  
Public Data ◽  
Normal Chest ◽  
Chest X Ray ◽  
The Stability ◽  
Definition Of

It has recently been shown that the interpretation by partial differential equations (PDEs) of a class of convolutional neural networks (CNNs) supports definition of architectures such as parabolic and hyperbolic networks. These networks have provable properties regarding the stability against the perturbations of the input features. Aiming for robustness, we tackle the problem of detecting changes in chest X-ray images that may be suggestive of COVID-19 with parabolic and hyperbolic CNNs and with domain-specific transfer learning. To this end, we compile public data on patients diagnosed with COVID-19, pneumonia, and tuberculosis, along with normal chest X-ray images. The negative impact of the small number of COVID-19 images is reduced by applying transfer learning in several ways. For the parabolic and hyperbolic networks, we pretrain the networks on normal and pneumonia images and further use the obtained weights as the initializers for the networks to discriminate between COVID-19, pneumonia, tuberculosis, and normal aspects. For DenseNets, we apply transfer learning twice. First, the ImageNet pretrained weights are used to train on the CheXpert dataset, which includes 14 common radiological observations (e.g., lung opacity, cardiomegaly, fracture, support devices). Then, the weights are used to initialize the network which detects COVID-19 and the three other classes. The resulting networks are compared in terms of how well they adapt to the small number of COVID-19 images. According to our quantitative and qualitative analysis, the resulting networks are more reliable compared to those obtained by direct training on the targeted dataset.

scholarly journals Sensor-Based Predictive Maintenance with Reduction of False Alarms—A Case Study in Heavy Industry

Sensors ◽  
2021 ◽  
Vol 22 (1) ◽  
pp. 226
Author(s):  
Marek Hermansa ◽  
Michał Kozielski ◽  
Marcin Michalak ◽  
Krzysztof Szczyrba ◽  
Łukasz Wróbel ◽  
...  
Keyword(s):  
Historical Data ◽  
Cold Start ◽  
Predictive Maintenance ◽  
Use Cases ◽  
False Alarms ◽  
Data Sets ◽  
Outlier Identification ◽  
Machine Operator ◽  
System Operating

In this paper, the problem of the identification of undesirable events is discussed. Such events can be poorly represented in the historical data, and it is predominantly impossible to learn from past examples. The discussed issue is considered in the work in the context of two use cases in which vibration and temperature measurements collected by wireless sensors are analysed. These use cases include crushers at a coal-fired power plant and gantries in a steelworks converter. The awareness, resulting from the cooperation with industry, of the need for a system that works in cold start conditions and does not flood the machine operator with alarms was the motivation for proposing a new predictive maintenance method. The proposed solution is based on the methods of outlier identification. These methods are applied to the collected data that was transformed into a multidimensional feature vector. The novelty of the proposed solution stems from the creation of a methodology for the reduction of false positive alarms, which was applied to a system identifying undesirable events. This methodology is based on the adaptation of the system to the analysed data, the interaction with the dispatcher, and the use of the XAI (eXplainable Artificial Intelligence) method. The experiments performed on several data sets showed that the proposed method reduced false alarms by 90.25% on average in relation to the performance of the stand-alone outlier detection method. The obtained results allowed for the implementation of the developed method to a system operating in a real industrial facility. The conducted research may be valuable for systems with a cold start problem where frequent alarms can lead to discouragement and disregard for the system by the user.

Towards Generalization of Intelligent Fault Detection for Roller Element Bearings via Distinct Dataset Transfer Learning

2021 ◽  
Author(s):  
Justin Larocque-Villiers ◽  
Patrick Dumond
Keyword(s):  
Fault Detection ◽  
Open Source ◽  
Transfer Learning ◽  
Data Augmentation ◽  
Operating Conditions ◽  
Predictive Maintenance ◽  
Training Data ◽  
Fault Classification ◽  
Target Domain ◽  
Bearing Faults

Abstract Through the intelligent classification of bearing faults, predictive maintenance provides for the possibility of service schedule, inventory, maintenance, and safety optimization. However, real-world rotating machinery undergo a variety of operating conditions, fault conditions, and noise. Due to these factors, it is often required that a fault detection algorithm perform accurately even on data outside its trained domain. Although open-source datasets offer an incredible opportunity to advance the performance of predictive maintenance technology and methods, more research is required to develop algorithms capable of generalized intelligent fault detection across domains and discrepancies. In this study, current benchmarks on source–target domain discrepancy challenges are reviewed using the Case Western Reserve University (CWRU) and the Paderborn University (PbU) datasets. A convolutional neural network (CNN) architecture and data augmentation technique more suitable for generalization tasks is proposed and tested against existing benchmarks on the Pb U dataset by training on artificial faults and testing on real faults. The proposed method improves fault classification by 13.35%, with less than half the standard deviation of the compared benchmark. Transfer learning is then used to leverage the larger PbU dataset in order to make predictions on the CWRU dataset under a challenging source-target domain discrepancy in which there is minimal training data to adequately represent unseen bearing faults. The transfer learning-based CNN is found to be capable of generalizing across two open-source datasets, resulting in an improvement in accuracy from 53.1% to 68.3%.

Using Rule Quality Measures for Rule Base Refinement in Knowledge-Based Predictive Maintenance Systems

2020 ◽  
Vol 51 (2) ◽  
pp. 161-176
Author(s):  
Qiushi Cao ◽  
Cecilia Zanni-Merk ◽  
Ahmed Samet ◽  
François de Bertrand de Beuvron ◽  
Christoph Reich
Keyword(s):  
Quality Measures ◽  
Predictive Maintenance ◽  
Rule Base ◽  
Knowledge Based ◽  
Maintenance Systems

A Novel Tagging Augmented LDA Model for Clustering

2019 ◽  
Vol 16 (3) ◽  
pp. 59-77
Author(s):  
Yi Zhao ◽  
Yu Qiao ◽  
Keqing He
Keyword(s):  
Transfer Learning ◽  
Real World ◽  
Latent Dirichlet Allocation ◽  
State Of The Art ◽  
Positive Influence ◽  
The State ◽  
Clustering Methods ◽  
Automatic Clustering ◽  
Real World Datasets ◽  
High Representation

Clustering has become an increasingly important task in the analysis of large documents. Clustering aims to organize these documents, and facilitate better search and knowledge extraction. Most existing clustering methods that use user-generated tags only consider their positive influence for improving automatic clustering performance. The authors argue that not all user-generated tags can provide useful information for clustering. In this article, the authors propose a new solution for clustering, named HRT-LDA (High Representation Tags Latent Dirichlet Allocation), which considers the effects of different tags on clustering performance. For this, the authors perform a tag filtering strategy and a tag appending strategy based on transfer learning, Word2vec, TF-IDF and semantic computing. Extensive experiments on real-world datasets demonstrate that HRT-LDA outperforms the state-of-the-art tagging augmented LDA methods for clustering.

Predictive Maintenance of Cogeneration Engines Through the Analysis of the Circulating Fluids

2002 ◽  
Author(s):  
J. R. Gonza´lez ◽  
J. Velayos ◽  
M. Comamala
Keyword(s):  
Fuzzy Logic ◽  
Expert System ◽  
Diesel Engines ◽  
Predictive Maintenance ◽  
Maintenance System ◽  
Maintenance Systems

In this article we present a fluid-based predictive maintenance system based on an expert system which uses fuzzy logic. The programme uses information from the circulating fluids of the machine to provide an evaluation of the maintenance status of the engine. Specifically, the programme is aimed at diesel engines in a half rate cogeneration, and so we will compare our results with other commercial maintenance systems, such as FAMM (Texaco) and ADOC (Repsol), which provide corresponding responses.

scholarly journals Disambiguating Energy Disaggregation: A Collective Probabilistic Approach

2017 ◽  
Author(s):  
Sabina Tomkins ◽  
Jay Pujara ◽  
Lise Getoor
Keyword(s):  
Energy Consumption ◽  
Energy Demand ◽  
Negative Impact ◽  
Probabilistic Approach ◽  
Ground Truth ◽  
Solution Quality ◽  
Ground Truth Data ◽  
Hinge Loss ◽  
Energy Disaggregation ◽  
Real World Datasets

Reducing household energy usage is a priority for improving the resiliency and stability of the power grid and decreasing the negative impact of energy consumption on the environment and public health.Relevant and timely feedback about the power consumption of specific appliances can help household residents to reduce their energy demand. Given only a total energy reading, such as that collected from a residential meter, energy disaggregation strives to discover the consumption of individual appliances. Existing disaggregation algorithms are computationally inefficient and rely heavily on high-resolution ground truth data. We introduce a probabilistic framework which infers the energy consumption of individual appliances using a hinge-loss Markov random field (HL-MRF), which admits highly scalable inference. To further enhance efficiency, we introduce a temporal representation which leverages state duration. We also explore how contextual information impacts solution quality with low-resolution data. Our framework is flexible in its ability to incorporate additional constraints; by constraining appliance usage with context and duration we can better disambiguate appliances with similar energy consumption profiles. We demonstrate the effectiveness of our framework on two public real-world datasets, reducing the error relative to a previous state-of-the-art method by as much as 50%.

Sign in / Sign up

Export Citation Format

Share Document