Industrial Monitoring Research Articles

Towards UAVs in Construction: Advancements, Challenges, and Future Directions for Monitoring and Inspection

The use of UAVs for monitoring and inspection in the construction industry has garnered considerable attention in recent years due to their potential to enhance safety, efficiency, and accuracy. The development and application of various types of drones and sensors in the construction industry have opened up new data collection and analysis possibilities. This paper provides a thorough examination of the latest developments in the use of UAVs for monitoring and inspection in the construction industry, including a review of the current state of UAVs and an exploration of the types of drones and sensors applied and their applications. It also highlights the technological advancements in this field. However, as with any new technology, there are challenges and limitations that need to be addressed, such as regulatory and legal concerns, technical limitations, data processing challenges, training and expertise, and safety. Finally, we offer insights into potential solutions to these challenges, such as innovative sensors and imaging technologies, integration with other construction technologies, and the use of machine learning and AI for data analysis, which are some of the potential areas for future investigation, and highlight the prospects for drone-based construction inspection.

Implementing Deep Learning Models in Embedded Systems for Diagnosis Induction Machine

Diagnosis of induction machines based on deep learning models is becoming a trend in maintenance systems of modern industry. The implementation of such systems allows low-cost industrial monitoring due to the characteristics of hardware and software. However, combining deep learning models with current embedded systems is a difficult issue due to the computational cost to run data through neural networks. In particular, balancing the resources of graphics processing unit (GPU) with modern architecture models is a challenge for prototyping induction machine diagnostic systems. In this work, a novel method is proposed to implement trained deep learning models in low-cost embedded systems. We review algorithmic optimizations for running neural network models, balancing computational resources. In addition, we propose the use of a graphical user interface (GUI) as a tool for the end user of the embedded system. The approach proposed in this work can be of great help to future researchers who develop their prototype using deep learning models and low-cost embedded systems.

Multi-platform solution for data acquisition

Analogue data acquisition is a common task which has application in several fields such as scientific research, industry, food production, safety, and environmental monitoring. It can be carried out either using systems designed ad-hoc for a specific application or by using general-purpose Digital Acquisition Boards (DAQ). Several DAQ solutions are nowadays available on the market, however, most of them are extremely expensive and come as commercial closed products, a factor which prevents users to adapt the system to their specific applications and limits the product compatibility to few operating systems or platforms. This paper describes the design and the preliminary metrological characterization of a digital data acquisition solution based on the Teensyduino Development Board. The aim of the project is to create a hardware and software infrastructure suitable to be employed on several operating systems and that can be freely modified by the users when required. Teensyduino board is a well-known development platform which is characterized by high computing performance and USB support. Taking advantage of the Teensyduino features, the proposed system is easy to be calibrated and used, and it provides functions and performance comparable to many commercial DAQs, but at a significantly lower cost.

AI-powered health monitoring of anode baking furnace pits in aluminum production using autonomous drones

Industrial health monitoring in factories is essential for quality assurance, energy and cost reduction, and health and safety. In aluminum factories, anode furnace pits’ flue walls deform over time due to cyclic heating and cooling. They are inspected and classified using manually acquired measurements in a process that takes several hours and is done under high temperatures using specialized equipment. We propose an end-to-end AI-powered system for automated inspection using drones. We fly a drone carrying color and depth cameras to film and navigate a 50-pit furnace floor autonomously in a GPS-denied environment. We then mosaic the recorded videos to produce color and depth mosaics using frame-to-frame motion parameters estimated using the color videos and applied to both. We finish the mosaics using depth-to-color mosaic registration based on maximizing mutual information on gradients. We extract pit images from the mosaics using a YOLOv5 object detection initially trained using a physical floor model with a proposed data augmentation scheme and then fine-tuned for the on-site environment. We achieve a mean average precision of 94.7%. Once pit images are detected and initially classified, we propose a dual-stage segmentation algorithm using the Hough transform and a semantic segmentation network trained using probabilistic feature images with a precision of 96.67%. Segmented pits with depth information allow us to produce 3D models of pits to aid in temporal monitoring and diagnosis confirmation. Our system is cost-effective and reduces inspection downtime by 87%, eliminating the need for human intervention.

CatSight, a direct path to proper multi-variate time series change detection: perceiving a concept drift through common spatial pattern

Detecting changes in data streams, with the data flowing continuously, is an important problem which Industry 4.0 has to deal with. In industrial monitoring, the data distribution may vary after a change in the machine’s operating point; this situation is known as concept drift, and it is key to detecting this change. One drawback of conventional machine learning algorithms is that they are usually static, trained offline, and require monitoring at the input level. A change in the distribution of data, in the relationship between the input and the output data, would result in the deterioration of the predictive performance of the models due to the lack of an ability to generalize the model to new concepts. Drift detecting methods emerge as a solution to identify the concept drift in the data. This paper proposes a new approach for concept drift detection—a novel approach to deal with sudden or abrupt drift, the most common drift found in industrial processes-, called CatSight. Briefly, this method is composed of two steps: (i) Use of Common Spatial Patterns (a statistical approach to deal with data streaming, closely related to Principal Component Analysis) to maximize the difference between two different distributions of a multivariate temporal data, and (ii) Machine Learning conventional algorithms to detect whether a change in the data flow has been occurred or not. The performance of the CatSight method, has been evaluated on a real use case, training six state of the art Machine Learning (ML) classifiers; obtained results indicate how adequate the new approach is.

Determination of Metals in Environmental Waters by Solution Cathode Glow Discharge -Atomic Emission Spectroscopy (SCGD-AES) with a Syringe Pump for Sample Introduction

This study reports a syringe injection pump for sample introduction for solution cathode glow discharge (SCGD) atomic emission spectroscopy (AES). The experimental conditions are optimized at 60 mA discharge current, 2.5 mm gap spacing, 2.0 mm distance from the glass capillary to cathode, 1.60 mL/min flow rate, and pH 1.0 based on the detection limits (DLs) obtained for Li, Cd, Cu, Ag, Mn, Pb, Cr, and Zn. The detection limits for Li, Cd, Cu, Ag, Mn, Pb, Cr, and Zn were 0.24, 12.26, 12.08, 1.09, 46.42, 25.81, 183.05, and 29.01 µg/L, respectively, significantly lower than using the traditional peristaltic pump. The detection limits decreased by 1.4, 1.8, 2.2, 5.0, 1.4, 1.4, 1.4, 1.6, and 2.2-fold for these elements. The flow rate and the solution consumed per unit time were significantly reduced using this design. The spectral detection revealed stable and practical operation of the system with relative standard deviations <0.85%. Water samples from Yangtze River, Baixi River, and Baiyun Reservoir are analyzed using the modified SCGD system and its evaluation is based upon comparison to inductively coupled plasma—optical emission spectroscopy (ICP-OES). The modified device operates at atmospheric pressure without inert gas and integrates a syringe injection pump and a compact spectrometer to optimize solution utility, analytical performance, and operation size for industrial and environmental monitoring applications.

Effect of applied stresses on magnetic properties of Co and Fe-rich glass-coated microwires

In this article we evaluate the possibility of using glass coated magnetic microwires as an alternative to optical fiber devices in structural health monitoring in the railway industry. The effect of applied stress on hysteresis loops of Fe71.80B13.27Si11.02Nb2.99Ni0.92 and Co65.34Si12.00B10.20Cr8.48Fe3.90Mo0.08 microwires has been studied and analyzed. An interpolation function has been obtained relating the coercitivity Hc with the applied stress.

Semi-supervised Discriminative Projective Dictionary Pair Learning and Its Application to Industrial Process

Industrial process data have the characteristics of less label, multimode, high dimension, containing noise, and mixing with outliers, which increase the difficulty of mode identification and anomaly detection in process monitoring using limited labeled data. In this article, to address the effect of these adverse factors, a semi-supervised discriminative projective dictionary pair learning (SSDP-DPL) is proposed for industrial process monitoring. First, a semi-supervised dictionary pair learning framework with a class estimation regularization term is developed to address the problem of lacking labeled data. Based on unlabeled data and their reconstruction errors, the class estimation regularization term is designed to obtain a discriminative extended synthetical dictionary, mining the hidden discriminative information in unlabeled data and reducing the impact of incorrect class estimation. Second, a sparse constraint is added to the analytical dictionary to enhance the robustness of projective dictionary pair. Then, a class discriminative function term is designed to improve the discrimination of dictionary pair, which ensures the intraclass compactness and interclass separation of coding space. Finally, the synthetical dictionary, the analytical dictionary, and the control threshold are obtained by iterating the dictionary pair to a process monitoring model for anomaly detection and mode identification. The effectiveness of the proposed process monitoring method is verified by the continuous stirred tank heater process and Tennessee Eastman benchmark tests. The proposed method is then applied to a real-world aluminum electrolysis industrial process. Experimental results demonstrate the superiority of our SSDP-DPL in contrast to other existing methods.

An Optimization Algorithm H-GVSPM for Electrical Impedance Tomography

Electrical impedance tomography (EIT) is a novel industrial monitoring technology with the characteristics of safety, inexpensiveness, and real-time monitoring. Its imaging results are easily to generate artifacts due to the influence of various factors. Therefore, the image quality is low, and the contour of measured objects has a distortion phenomenon. An optimization algorithm H-GVSPM based on the Hadamard product is proposed for EIT to reduce the artifacts, and to improve the image quality. The relative conductivity difference between the measured object and the background becomes more distinguishable with the action of the Hadamard product. Simulation and experimental results show that H-GVSPM effectively reduces the influence of artifacts in the reconstructed images. The maximum increase of the correlation coefficient approximately is 15.56% and 22.80% respectively compared with GVSPM and TV method. The maximum reduction of the voltage relative residual is 22.12% in the simulation, and it is 8.12% in the experiment. The standard deviation of conductivity is about 1/10 of the original. This study provides a reference method for improving image quality, and it contributes to the application of EIT in industry, biomedicine, and other fields in the future.

Battery-Less Industrial Wireless Monitoring and Control System for Improved Operational Efficiency

An industrial wireless monitoring and control system, capable of supporting energy-harvesting devices through smart sensing and network management, designed for improving electro-refinery performance by applying predictive maintenance, is presented. The system is self-powered from bus bars, and features wireless communication and easy-to-access information and alarms. With cell voltage and electrolyte temperature measurements, the system enables real-time cell performance discovery and early reaction to critical production or quality disturbances such as short-circuiting, flow blockages, or electrolyte temperature excursions. Field validation shows an increase in operational performance of 30% (reaching 97%) in the detection of short circuits, which, thanks to a neural network deployed, are detected, on average, 10.5 h earlier compared to the traditional methodology. The developed system is a sustainable IoT solution, being easy to maintain after its deployment, and providing benefits of improved control and operation, increased current efficiency, and decreased maintenance costs.

Portable ZnO-based ammonia detector for continuous monitoring in poultry industry

Portable ZnO-based ammonia detector for continuous monitoring in poultry industry

Detection and quantification of Pb and Cr in oysters using laser-induced breakdown spectroscopy

The quantitative determination of heavy metals in aquatic products is of great importance for food security issues. Laser-induced breakdown spectroscopy (LIBS) has been used in a variety of foodstuff analysis, but is still limited by its low sensitivity when targeting trace heavy metals. In this work, we compare three sample enrichment methods, namely drying, carbonization, and ashing, for increasing detection sensitivity by LIBS analysis for Pb and Cr in oyster samples. The results demonstrate that carbonization can remove a significant amount of the contributions of organic elements C, H, N and O; meanwhile, the signals of the metallic elements such as Cu, Pb, Sr, Ca, Cr and Mg are enhanced by 3–6 times after carbonization, and further enhanced by 5–9 times after ashing. Such enhancement is not only due to the more concentrated metallic elements in the sample compared to the dried ones, but also the unifying of the matter in carbonized and ashed samples from which higher plasma temperature and electron density are observed. This condition favors the detection of trace elements. According to the calibration curves with univariate and multivariate analysis, the ashing method is considered to be the best choice. The limits of detection of the ashing method are 0.52 mg kg−1 for Pb and 0.08 mg kg−1 for Cr, which can detect the presence of heavy metals in the oysters exceeding the maximum limits of Pb and Cr required by the Chinese national standard. This method provides a promising application for the heavy metal contamination monitoring in the aquatic product industry.

Evanescent Wave Lab‐on‐Fiber for High Sensitivity Gas Spectroscopy with Wide Dynamic Range and Long‐Term Stability

AbstractMicro/nano waveguide has been widely exploited for gas sensing due to its strong confinement of evanescent field, compact size, and fast response. However, the state‐or‐the‐art evanescent‐wave gas sensors still struggle to satisfy the requirements of high sensitivity, high selectivity, large dynamic range, and long‐term stability. In this work, microfiber‐based evanescent‐wave mode‐phase‐difference photothermal spectroscopy for trace gas detection is reported. With a microfiber of 5.6 cm in length and 2.36 µm in diameter, noise‐equivalent‐concentration of 160 ppb methane in nitrogen is achieved with a dynamic range over 6 orders of magnitude, response time of less than 6 s, and &lt;3% instability over 7 days. The technique utilizes commercial fiber‐optic components to make cost‐effective lab‐on‐fiber probes with real‐time and remote detection capability for industrial, medical, and environmental monitoring.

Explainable AI for Machine Fault Diagnosis: Understanding Features’ Contribution in Machine Learning Models for Industrial Condition Monitoring

Although the effectiveness of machine learning (ML) for machine diagnosis has been widely established, the interpretation of the diagnosis outcomes is still an open issue. Machine learning models behave as black boxes; therefore, the contribution given by each of the selected features to the diagnosis is not transparent to the user. This work is aimed at investigating the capabilities of the SHapley Additive exPlanation (SHAP) to identify the most important features for fault detection and classification in condition monitoring programs for rotating machinery. The authors analyse the case of medium-sized bearings of industrial interest. Namely, vibration data were collected for different health states from the test rig for industrial bearings available at the Mechanical Engineering Laboratory of Politecnico di Torino. The Support Vector Machine (SVM) and k-Nearest Neighbour (kNN) diagnosis models are explained by means of the SHAP. Accuracies higher than 98.5% are achieved for both the models using the SHAP as a criterion for feature selection. It is found that the skewness and the shape factor of the vibration signal have the greatest impact on the models’ outcomes.

A Fog-Based Architecture for Latency-Sensitive Monitoring Applications in Industrial Internet of Things

In industrial wireless sensors networks (IWSNs) deployments, data management and processing tasks are often delegated to centralized computing facilities. Albeit providing an abundance of resources, these solutions impose several drawbacks, stemming from the higher latency they may introduce, as well as from security/privacy issues. To deal with these shortcomings, in this article, we propose an architecture designed for industrial monitoring duties, based on the Fog computing paradigm. In particular, we devised a Fog layer to host a data aggregation algorithm based on singular value decomposition (SVD). Considering a real-world case study about a distributed monitoring system for industrial motors, the benefits of the proposed solution in terms of network performances, i.e., latency and packet delivery ratio, are analyzed using the OMNET++ simulator. Besides, the system implements a set of security mechanisms at different levels (i.e., IWSN, Fog, and Cloud). In particular, we propose a new lightweight and energy-efficient ciphering algorithm for resource-constrained industrial motes. According to our experiments, the new algorithm reduces energy consumption by a factor of three compared to Advanced Encryption Standard-128.

Biases on visual warnings during multiple target visual searches on mine monitoring interface

With the continuous development of human–computer interaction, some industrial monitoring systems have gradually evolved from a single monitoring device to a comprehensive monitoring system platform, and multi-objective visual search is particularly important in the field of industrial monitoring. This study investigates the effect of conceptual and perceptual biases on similarity bias in multiple target visual searches and proposes relevant design strategies that can be applied to warning interfaces. The results of two experiments show that the effect of the similarity bias is due to the combined effects of conceptual and perceptual similarities, and on this basis, we propose a design strategy for the interface of a gas extraction monitoring system used in mines in a third experiment. An existing interface to monitor gas extraction and a proposed interface with modifications made to the existing interface are evaluated. The results confirm that the modified interface effectively improves visual search performance during multiple target visual searches. The research findings are important for reducing cognitive decision errors during related monitoring work and ensuring the performance of the mine gas safety monitoring system interfaces. The findings can also guide the designing of visual warning elements of different targets for other monitoring interfaces.

Online-Dynamic-Clustering-Based Soft Sensor for Industrial Semi-Supervised Data Streams.

In the era of big data, industrial process data are often generated rapidly in the form of streams. Thus, how to process such sequential and high-speed stream data in real time and provide critical quality variable predictions has become a critical issue for facilitating efficient process control and monitoring in the process industry. Traditionally, soft sensor models are usually built through offline batch learning, which remain unchanged during the online implementation phase. Once the process state changes, soft sensors built from historical data cannot provide accurate predictions. In practice, industrial process data streams often exhibit characteristics such as nonlinearity, time-varying behavior, and label scarcity, which pose great challenges for building high-performance soft sensor models. To address this issue, an online-dynamic-clustering-based soft sensor (ODCSS) is proposed for industrial semi-supervised data streams. The method achieves automatic generation and update of clusters and samples deletion through online dynamic clustering, thus enabling online dynamic identification of process states. Meanwhile, selective ensemble learning and just-in-time learning (JITL) are employed through an adaptive switching prediction strategy, which enables dealing with gradual and abrupt changes in process characteristics and thus alleviates model performance degradation caused by concept drift. In addition, semi-supervised learning is introduced to exploit the information of unlabeled samples and obtain high-confidence pseudo-labeled samples to expand the labeled training set. The proposed method can effectively deal with nonlinearity, time-variability, and label scarcity issues in the process data stream environment and thus enable reliable target variable predictions. The application results from two case studies show that the proposed ODCSS soft sensor approach is superior to conventional soft sensors in a semi-supervised data stream environment.

Determinan Fraud Hexagon Theory dan Indikasi Financial Statement Fraud

This study aims to test the influence of fraud hexagon theory on financial statement fraud. Elements of the causative factors of fraud based on fraud hexagon theory cannot be directly investigated but require proxy variables. Stimulus is projected with financial stability. Opportunity is projected with nature of industry and ineffective monitoring. Rationalization is projected by a change of auditor. Capability is projected with a change of director. Arrogance is projected with a frequent number of CEO's picture. Collution is projected by the ownership of the company by the government or a State-Owned Enterprise or is a private company. The test was conducted on healthcare companies listed on the Indonesia Stock Exchange during the period 2016-2020. Sample selection is done using purposive sampling techniques. Data analysis uses regression panel data. Based on the results of the analysis showed that the variables of financial stability, ineffective monitoring, change in auditor, frequent number of CEO's picture, and state-owned enterprises positively affect financial statement fraud. Nature of industry and change of director have no effect on financial statement fraud. This research provides implications for those who need information about the possibility of financial statement fraud in healthcare companies taking into account influencing factors.&#x0D; Keyword : Fraud Hexagon Theory; Financial Statement Fraud; Healthcare Company

Monitoring Chemical Accidents in Industrial Complexes Using Tower-Installed Infrared System for Remote Chemical Detection and Long-Range Video Surveillance System

Chemical industrial complexes are extensive, complex structures with large-scale chemical facilities where large quantities of various chemical substances are handled. Detection equipment must be installed in high locations to monitor these industrial complexes and detect chemical accidents from a distance. In previous studies, individual monitoring equipment was temporarily installed on the ground, on a rooftop, or on a vehicle to detect chemical accidents from a distance. In this study, however, the industrial complex chemical accident monitoring system was developed by combining different technologies and was installed on a tower. For the Yeosu National Industrial Complex (which functioned as a test bed), 70m-high steel towers were built. Additionally, an infrared system for remote chemical detection (SIGIS-2, Bruker) and a long-range video surveillance system (TORUSS-LR2000, Globalsystems) were installed at the top of steel towers to monitor the entire industrial complex. The target substances to be monitored in real time by the infrared system for remote chemical detection were selected, and the monitoring sections were classified to enable each piece of equipment to distinguish the scanned areas. To improve the accuracy of the detection results, the information about the actual handled substances and respective facilities of the sites in the industrial complex was inserted into the database of the system and then connected to the sections. During the three-month test operation, various chemical substances (including 1,3-butadiene, methanol, methylamine, ethyl acetate, ammonia, and vinyl chloride) were detected at each section in 20,034 cases, and the detection results were consistent with the inserted actual information. The accumulated detection data shows that the detection frequency of a specific chemical substance was high in each section. This can be used as a basis for modifying the threshold of the anomaly detection model, thereby improving the accuracy of the system. Therefore, this system can detect and evaluate the leakage of chemical substances and the occurrence of fires or smoke through large-scale scans 24 h per day. Furthermore, it can be used for the early detection of and effective responses to chemical accidents in industrial complexes.

A Review of Real-Time Fault Diagnosis Methods for Industrial Smart Manufacturing

In the era of Industry 4.0, highly complex production equipment is becoming increasingly integrated and intelligent, posing new challenges for data-driven process monitoring and fault diagnosis. Technologies such as IIoT, CPS, and AI are seeing increasing use in modern industrial smart manufacturing. Cloud computing and big data storage greatly facilitate the processing and management of industrial information flow, which helps the development of real-time fault diagnosis (RTFD) technology. This paper provides a comprehensive review of the latest RTFD technologies in the field of industrial process monitoring and machine condition monitoring. The RTFD process is introduced in detail, starting with the data acquisition process. The current RTFD methods are divided into methods based on independent feature extraction, methods based on “end-to-end” neural networks, and methods based on qualitative knowledge reasoning from a new perspective. In addition, this paper discusses the challenges and potential trends of RTFD in future development to provide a reference for researchers focusing on this field.