Industrial Sensor Research Articles

Condition Monitoring of Pneumatic Drive Systems Based on the AI Method Feed-Forward Backpropagation Neural Network.

Machine condition monitoring is used in a variety of industries as a very efficient strategy for equipment maintenance. This paper presents a study on monitoring a pneumatic system using a feed-forward backpropagation neural network as a classifier and compares the results obtained with different sensor signals and associated extracted features as input for classification. The vibrations of the body of a pneumatic cylinder are acquired using both common industrial sensors and low-cost sensors integrated into an Arduino board. Pressure sensors for both chambers and a position sensor are also used. Power spectral density (PSD) is used to extract features from the acceleration signals, as well as statistical indices. Statistical indices are considered for pressure and position sensors. The results, which are based on experimental data obtained on a test bench, show that a feed-forward neural network makes it possible to identify the operating states with a good degree of reliability. Even with low-cost instrumentation, it is possible to realize reliable condition monitoring based on vibrations. This last result is particularly important as it can help to further increase the uptake of this maintenance approach in the industrial environment.

3D-printed sensor electric circuits using atomic layer deposition

3D-printing, also known as additive manufacturing, has enabled the production of dynamically shaped objects often customized for specific applications. Many applications, such as sensors in the aerospace industry, have demanding mass and volume requirements or need to work in challenging environments that necessitate electronics to be protected. The combination of 3D-printing and electronics could open up new applications not feasible previously. We propose a novel manufacturing method capable of integrating a complex electric circuit consisting of several, commonly available electronic components with a 3D-printed object. This is achieved using a commercial printer and atomic layer deposition for coating. Various printable polymers and coatings were tested to identify two polymers that could be printed into one object, allowing selective conductivity when coated with conductive coating Selective conductivity is achieved when one polymer exhibits poorer and more non-continuous coating growth compared to the other. The 3D-printed object’s three-dimensional shape and details were used to create the electrical circuit and aid in achieving selective conductivity. A demonstration consisting of an ultraviolet light (UV) sensor, based on an existing traditional circuit board, was replicated using this method. The 3D-printed circuit was then tested by comparing its output with that of the original when placed under the same UV-light source. The novel circuit output closely followed the original. The presented method can combine an electric circuit with the dynamic capabilities of a 3D-printer, allowing for savings in existing applications as well as new applications.

Multi-mode industrial soft sensor method based on mixture Laplace variational auto-encoder

The industrially collected process data usually exhibit non-Gaussian and multi-mode characteristics. Due to sensor failures, irregular disturbances, and transmission problems, there are unavoidable outliers that make the data exhibit heavy-tailed characteristics. To this end, a variational auto-encoder regression method based on the mixture Laplacian distribution (MLVAER) is proposed, by introducing a type-II multivariate Laplacian distribution in the latent variable space for robust modeling, and further extending it to the mixture form to accommodate multi-mode processes, the corresponding reparameterization trick is finally proposed for the mixture form of this distribution for neural network gradient descent training. The model based on this distribution assumption has higher degrees of freedom than the model based on the traditional multivariate Laplace distribution assumption when the network structure is the same. Numerical simulation and experiments on two industrial examples demonstrate that the proposed algorithm reduces the root mean square error by over 15% compared to other algorithms.

Temperature compensation for piezoresistive pressure sensor based on deep learning on graphs

With the escalating demand for precision in piezoresistive pressure sensors to cater to a wide spectrum of applications, a significant challenge emerges due to the material properties of these sensors, which induce a substantial temperature coefficient. This limitation restricts the operational temperature range and is further exacerbated by ambient temperature variations, resulting in pronounced nonlinearity in sensor responses. To address these challenges, A new method using Graph Neural Networks (GNNs) is introduced to address nonlinearity and temperature drift in piezoresistive pressure sensors. GNNs improve the sensors’ accuracy and reliability across different temperatures, expanding their applicability. Test results show a notable accuracy enhancement, with a maximum full-scale error below 0.05% and even lower in specific ranges (under 0.04%). This precision makes the method ideal for industrial pressure sensor applications and production, offering significant benefits.

Novel Multiattribute Space-Based LSTM for Industrial Soft Sensor Applications

Novel Multiattribute Space-Based LSTM for Industrial Soft Sensor Applications

A Multihead Attention Self-Supervised Representation Model for Industrial Sensors Anomaly Detection

Industrial sensors capture critical information for intelligent manufacturing maintenance. To promote equipment upgrading and manufacturing processes, intelligent decisions, and information learning play an important role. Although deep learning methods historically obtain excellent results, there is always a trade-off between fine-tuning existing networks or designing models from scratch for sensor data processing. In this paper, we propose the multi-head attention self-supervised (MAS) representation model, which is a self-supervised learning-based sensor feature extraction network. To the best of our knowledge, this is the first time a self-supervised contrastive learning method using positive samples that represent multi-dimensional industry sensor data is being used for anomaly detection. We review alternative data augmentation methods proposed for better-representing sensor sequence data. We use this insight to design a new structure that adapts to the temporal characteristics of the application. We apply our method to a real-world water circulation system that uses a variety of industrial sensors. The effectiveness of the proposed MAS methods is demonstrated.

Adversarial Attacks for Neural Network-Based Industrial Soft Sensors: Mirror Output Attack and Translation Mirror Output Attack

Adversarial Attacks for Neural Network-Based Industrial Soft Sensors: Mirror Output Attack and Translation Mirror Output Attack

Guided Exploration of Industrial Sensor Data

AbstractIn recent years, digitization in the industrial sector has increased steadily. Digital data not only allows us to monitor the underlying production process using machine learning methods (anomaly detection, behaviour analysis) but also to understand the underlying production process. Insights from Exploratory Data Analysis (EDA) play an important role in building data‐driven processes because data scientists learn essential characteristics of the data in the context of the domain. Due to the complexity of production processes, it is usually difficult for data scientists to acquire this knowledge by themselves. Hence, they have to rely on continuous close collaboration with domain experts and their acquired domain expertise. However, direct communication does not promote documentation of the knowledge transfer from domain experts to data scientists. In this respect, changing team constellations, for example due to a change in personnel, result in a renewed high level of effort despite the same knowledge transfer problem. As a result, EDA is a cost‐intensive iterative process. We, therefore, investigate a system to extract information from the interactions that domain experts perform during EDA. Our approach relies on recording interactions and system states of an exploration tool and generating guided exploration sessions for domain novices. We implement our approach in a software tool and demonstrate its capabilities using two real‐world use cases from the manufacturing industry. We evaluate its feasibility in a user study to investigate whether domain novices can reproduce the most important insights from domain experts about the datasets of the use cases based on generated EDA sessions. From the results of this study, we conclude the feasibility of our system as participants are able to reproduce on average 86.5% of insights from domain experts.

Linked data to streaming data sensors sources for direct access

Goal: Industrial operations are complex, and data sensors assure safety and reliable information for production improvements. Multiple stakeholders can take advantage of data acquisition for post-analysis and process control. Providing users and systems with friendly access to operating data is fundamental to the digital transition in the industry 4.0 scenario. Linking data and systems over ontologies and Industry Foundation Classes will boost supply chain performance in many layers. This paper presents the concept of valid data points over Uniform Resource Identifiers for sensor time-series into triples stores via Application Programming Interfaces. Design/Methodology/Approach: A streaming data source approach to integrating industrial sensor data and sharing it via Uniform Resource Identifiers is developed and tested using Node-Red with multiple data connection types, such as the Industry Foundation Classes and open-source time series databases. Results: The detailed proof of concept presented valid the feasibility of sharing sensor data via Uniform Resource Identifiers. The findings provide a backbone of a system able to interop Message Queuing Telemetry Transport data, Resource Description Framework datasets and Industry Foundation Classes schema. Limitations of the investigation: The system envisaged was tested using simulated data. However, it is expected to have similar results from real data use. Nevertheless, more research will be needed to implement more features, such as three-dimensional object integration. Practical implications: The solution designed and tested presented can be used in practice for companies that desire to expand via linked data shareability and interoperability. Also, researchers can advance the solution for specific features, such as creating an open-source data query and manipulation language. Originality/Value: This paper examines future deployments of systems-to-systems interoperability targeting user-friendly data shareability. It is meant to be useful for industrial and academic developments.

Process expert knowledge is essential in creating value from data-driven industrial soft sensors

The objective of Industry 5.0 is to (re)centre the human operator amidst digital process automation. This requires new data processing technologies that extract human expertise and integrate it with advanced data modelling techniques to enhance human-computer collaboration. In this work, we present an integrated and systematic approach that combines contemporary data modelling technology with process knowledge gained from engineers. Specifically, we develop, investigate, and compare data- and expert-driven approaches for selecting process variables for a real-time product quality prediction for three parallel-operated dairy processing lines. We show how and why data-driven selections are outperformed by expert-driven selections, but that they can improve upon expert-driven selections when used as a secondary selection step. This leads to explainable soft-sensors that are optimal both in accuracy and in parsimony. This highlights the significance of capturing and preserving human expertise, which will enhance the quality and sustainability of industrial processing.

Instance transfer partial least squares for semi-supervised adaptive soft sensor

In order to deal with the unknown data drift problem caused by the change of working conditions in the process of industrial soft sensor modeling, this paper proposes a semi-supervised soft sensor modeling algorithm based on instance transfer learning, named instance transfer partial least squares. Under the framework of nonlinear partial least squares, the algorithm introduces the instance transfer inner model, and realizes the distribution adaptation of the modeling data and the data to be predicted through sample importance weighting. First, a nonlinear partial least squares model is constructed by using labeled samples in the source domain, which consists of a linear outer model and a nonlinear inner model. The linear outer model maps the original data to the hidden variable space, and the nonlinear inner model realizes the nonlinear modeling of hidden variables through kernel ridge regression. Second, a small number of labeled samples from the target domain are used to update the model weights of the nonlinear inner model, enabling adaptation of the latent variable distribution. The effectiveness of the algorithm is verified on a toy dataset, a simulation dataset and a real pulverized coal power plant dataset.

A Comprehensive Review of the Internet of Medical Things in Healthcare

The continuous and rapid development in Internet of Medical Things (IoMT) technologies has led to the expansion of their applications in the field of healthcare and to increased interest in developing their technologies for the purpose of improving their performance in light of the challenges facing their practical application. This paper aims to provide a comprehensive up-to-date review of the literature on IoMT and telemedicine. The study discusses the growing interest in the Internet of Things (IoT) in healthcare, emphasizing the IoMT. It highlights the significance of the IoMT in intelligent healthcare systems, particularly in pandemics like COVID-19, where remote monitoring becomes crucial. This study examines the utilization of the IoMT systems, emphasizing on the application of sensors in healthcare industry. Furthermore, it highlights the significance of cloud computing, communication technologies, artificial intelligence, and Blockchain in IoMT. The challenges related to accepting IoMT in practice are explored; including data security, integration of protocols, data overload, accuracy and cost. Overall, the analyzed literature confirms that IoMT-based healthcare systems demonstrate significant potential in delivering comprehensive surveillance and immediate monitoring through public catastrophes.

Methods of environmental protection in digital transformation in Russia, Poland and Germany

The article reveals an actual problem of environmental protection in the framework of the digital transformation of society. On the example of Russia, Poland and Germany as countries from different parts of Europe, the principles and measures that are taken to protect the environment in the context of digital transformation are considered. The research methodology is based on a systematic approach and includes a group of general scientific methods (analysis, synthesis, generalization, systematization); as well as a number of special methods: historiographical analysis of scientific literature on the topic of the study; the method of expert assessments, the method of statistical analysis and verification of research results. An empirical study was based on responses from 75 experts on environmental principles and measures across the European continent. The representative sample was formed on a regional basis and consisted of three expert groups: 25 experts from Russia; 25 experts from Poland and 25 experts from Germany. Based on the results of the study, the author comes to the following conclusions: in Russia, digital transformation is reflected in new forms and methods of state control in the field of environmental protection; Poland is developing a system of enforcement legislation in the field of air, water and raw material quality sensors in the food industry; in Germany, the focus is on public-private partnerships in the field of subsidizing the digitalization of private enterprises to equip environmental protection systems.

РОЛЬ СЕНСОРНЫХ ТЕХНОЛОГИЙ В РАЗВИТИИ МЕХАТРОННЫХ СИСТЕМ

The article examines the role of sensor technologies in the development of mechatronic systems. The main types of sensors, their technical characteristics and areas of application are analyzed. The interaction of sensors with other components of mechatronic systems, their contribution to providing feedback and control, as well as improving the performance and reliability of systems are described. Examples of practical application of sensors in various industries, such as industry, medicine and the automotive industry are given. The latest innovations and prospects for the development of sensor technologies in the context of mechatronic systems are discussed.

Research on Modeling of Industrial Soft Sensor Based on Ensemble Learning

Research on Modeling of Industrial Soft Sensor Based on Ensemble Learning

Siamese Neural Network and Multimodal Data Fusion Approach for Small-Sample Learning in Industrial Soft Sensor Modeling

Siamese Neural Network and Multimodal Data Fusion Approach for Small-Sample Learning in Industrial Soft Sensor Modeling

Prox-STA-LSTM: A Sparse Representation for the Attention-Based LSTM Networks for Industrial Soft Sensor Development

Prox-STA-LSTM: A Sparse Representation for the Attention-Based LSTM Networks for Industrial Soft Sensor Development

ВЕЛИКІ ДАНІ ТА ОЗЕРА ДАНИХ ЯК СКЛАДОВІ ІНФОРМАЦІЙНОЇ ЕКОНОМІКИ

The article is devoted to consideration of the importance of such components of the information economy as big data and data lakes. It is shown that in the process of unfolding the modern information technology revolution, information and information technologies are gaining increasing importance in the economy and social life. This is especially important in the context of the formation and development of the information economy. It is noted that the characteristic features of the information economy are its virtualization, the processing of large volumes of information based on the digitization of data, the use of quantum and neurotechnologies, artificial intelligence, wireless communication, the industrial Internet, sensors and robotics, digital platforms, etc., and the key technologies of the digital transformation of the economy and society are the Internet of Things, processing of structured and unstructured data of huge volumes, artificial intelligence and blockchain. Key commonalities and differences in big data and data lakes are shown, as well as their role in storing, processing and using large amounts of information. In particular, it states that "Big Data" has become shorthand for the technologies and analytical tools that help manage and query "big data" and represent three steps to a new way of analyzing information that will transform how we think about society and its organization: the first step - rejection of the sampling method; the second is the rejection of precision, and, finally, the third step is the rejection of centuries-old traditions of the search for causality. It is shown that along with big data, which stores processable and structured information, there are also data lakes, which are a system or centralized data repository where all structured, semi-structured, unstructured and binary data are stored in their natural, i.e., their own raw format. The further development of the information economy and the growth of the role and importance of information will lead to new requirements for the technologies of processing, storage and use of information resources.

Quality-Driven Regularization for Deep Learning Networks and Its Application to Industrial Soft Sensors.

The growth of data collection in industrial processes has led to a renewed emphasis on the development of data-driven soft sensors. A key step in building an accurate, reliable soft sensor is feature representation. Deep networks have shown great ability to learn hierarchical data features using unsupervised pretraining and supervised fine-tuning. For typical deep networks like stacked auto-encoder (SAE), the pretraining stage is unsupervised, in which some important information related to quality variables may be discarded. In this article, a new quality-driven regularization (QR) is proposed for deep networks to learn quality-related features from industrial process data. Specifically, a QR-based SAE (QR-SAE) is developed, which changes the loss function to control the weights of the different input variables. By choosing an appropriate inductive bias for the weight matrix, the model provides quality-relevant information for predictive modeling. Finally, the proposed QR-SAE is used to predict the quality of a real industrial hydrocracking process. Comparative experiments show that QR-SAE can extract quality-related features and achieve accurate prediction performance.

Recent advances in the synthesis of ZnO-based electrochemical sensors

Until now, various composites based on zinc oxide (ZnO) have been investigated in electrochemical sensors. The physical and electrochemical properties of ZnO and its structure can improve the selectivity, sensitivity, and adaptability of nanocomposites. Therefore, the focus on the fabrication of cheap ZnO-based electrodes with affordable and easy transportability has increased. In addition, the electrochemical behavior is affected by the structure and morphology of the ZnO-based composite in detecting pollutants such as volatile organic compounds, heavy metals, and toxins. Furthermore, ZnO-based nanostructures are efficient in the fabrication of electrochemical sensors in the food industry, pharmaceutical analysis, and medical diagnostics. In this review, various techniques in the synthesis of ZnO-based electrodes and their effect on the particle size, shape, and morphology of compounds have been collected. Since the performance of chemical sensors has a direct relationship with the structure of the composite used in its electrode, it is necessary to discuss the new production methods, new concepts, strategies, and challenges. Additionally, new gains highlight recent developments and sensing of various analytes in the monitoring systems. These sensors have demonstrated a strong growth acceleration which could lead to the development of recent technologies. At last, an optimistic outlook is provided on the future of ZnO-based sensors and their challenges.