Anomaly Detection Based on Edge Computing Using Transfer Learning in Industrial IoT

The evolution of the Industrial Internet of Things (IIoTs) greatly increases the volume of data generated by the connected IIoT devices. IIoT data are playing an increasingly important role in various industrial sectors. IIoT data sharing helps enterprises make better production decisions and respond to market changes timely. However, the distrust among IIoT entities and IIoT entities' distrust of data-sharing platforms may hinder the realization of data sharing. In this article, a decentralized IIoT data-sharing scheme based on blockchain and edge computing is proposed. A Proof of Storage and Transmission (PoST) consensus mechanism is proposed to meet data storage and transmission requirements of data owners in IIoT data-sharing networks. Based on the manufacture ties of data owners, shared data request probabilities are derived. The IIoT data sharing interactions between data owners and edge devices are modeled as a multiple-leader and multiple-follower Stackelberg game. The alternating direction method of multipliers (ADMMs) algorithm is used to obtain the optimal IIoT data sharing solutions in a distributed manner. Simulation results show that compared with the cooperative scheme, the total profit of edge devices is maximally increased by 59%, and the total utility of data owners is maximally increased by 52%.

The Industrial Internet of Things (IIoT) is a crucial research field spawned by the Internet of Things (IoT). IIoT links all types of industrial equipment through the network; establishes data acquisition, exchange, and analysis systems; and optimizes processes and services, so as to reduce cost and enhance productivity. The introduction of edge computing in IIoT can significantly reduce the decision-making latency, save bandwidth resources, and to some extent, protect privacy. This paper outlines the research progress concerning edge computing in IIoT. First, the concepts of IIoT and edge computing are discussed, and subsequently, the research progress of edge computing is discussed and summarized in detail. Next, the future architecture from the perspective of edge computing in IIoT is proposed, and its technical progress in routing, task scheduling, data storage and analytics, security, and standardization is analyzed. Furthermore, we discuss the opportunities and challenges of edge computing in IIoT in terms of 5G-based edge communication, load balancing and data offloading, edge intelligence, as well as data sharing security. Finally, we introduce some typical application scenarios of edge computing in IIoT, such as prognostics and health management (PHM), smart grids, manufacturing coordination, intelligent connected vehicles (ICV), and smart logistics.

The Internet of Things (IoT) is an essential component of the digital age, particularly in ensuring reliable and efficient operations as well as the timely and precise recognition of anomalies within IoT systems. However, anomaly detection in time-series data—especially data collected by edge devices—poses several challenges, including concerns over data privacy and communication overhead. To address these issues, this research proposes a novel deep learning (DL)-based anomaly detection model capable of being trained in real-time on edge devices within an IoT environment. The framework ensures user privacy by enabling distributed edge devices to cooperatively train an anomaly recognition system. This study introduces a new, Intelligent Shark Smell Tuned Deep Isolation Forest (ISS-DIF), which effectively detects anomalies and identifies outliers in industrial IoT sensor data with high accuracy. The hybrid model isolates anomalies rather than profiling normal data, making it particularly suitable for identifying rare anomalies within large datasets. Industrial data are collected from real-world manufacturing environments using IoT edge devices. Following data collection, median filtering is applied to reduce noise, and min-max scaling is employed for data normalization. The ISS component allows for fine-tuning the hyperparameters of the Deep Isolation Forest (DIF) to optimize detection performance. The DIF model is tailored to enhance anomaly detection capabilities in sensor data from industrial IoT applications. For validation, 80% of the dataset was used for training, while the remaining 20% served as the test set. The results demonstrate that the proposed ISS-DIF framework achieved superior training accuracy of 99.30%, with precision of 97.89%, recall of 98.76%, and an F1-score of 97.21%, outperforming the testing metrics. This approach integrates real-time anomaly detection with the processing capabilities of edge devices, thereby improving IoT data analytics and providing a scalable, efficient solution that preserves privacy in irregularity recognition within IoT environments.

Efficient heterogeneous signcryption scheme based on Edge Computing for Industrial Internet of Things

The Industrial Internet of Things (IIoT) technology has led to tremendous advancements in the field of real-time data collection, automation and intelligent decision-making. However, with an increased number of IIoT devices, security and privacy concerns, including unauthorized access, data manipulation and limited traceability, have emphasized the importance of addressing the risk. Among all the tools used, blockchain technology has gained prominence as a tool to reduce problems like data manipulation and unauthorized access. The authors of this study have developed a comprehensive review of the existing research, its application and the framework and integration of blockchain technology with IIoT for securing IIoT and related devices. Moreover, the study has also revealed key limitations of applying this technology, including transaction delay, resource consumption and scalability. Finally, this study has included an effective and comparative analysis with other technologies and solutions, providing readers with evidence-based and authentic guidelines on the viability of this technology. Moreover, the study has emphasized certain key factors of this technology and its application with IIoT, including enhanced trust, traceability and security, with consideration of the discussed key limitations. Based on the study, it can be clearly deduced and asserted that blockchain technology can play an effective and critical role in developing a novel and robust security mechanism for IIoT, including other related devices.

The industrial Internet of things (IIoT) is a framework that enables companies to improve their processes and increase their efficiency by using smart sensors and actuators. It is also used to collect data from various industrial devices and make use of this data to improve their operations. With the help of connected sensors and actuators, companies can save time. Due to the increasing number of industrial Internet of things (IIoT) devices, the data generated by these networks is constantly being analyzed. However, due to the high dimensionality of the data, it can cause fragmentation. Also, datasets collected by IIoT nodes are prone to displaying anomalous events. Unlike traditional networks, which are usually composed of several applications and protocols, IIoT networks are different. They require more stringent security measures to protect the integrity and confidentiality of their data. Due to the unique features of IIoT networks, they are prone to attack. This paper briefly describes the various requirements and challenges faced when it comes to protecting these networks. It also presents evaluation of various machine learning classifiers to detect anomalous behaviors in these networks. Due to the nature of the data that these networks contain anomalies, their analysis is critical to prevent attacks. In this paper, we analyze the usage of various machine learning algorithms to check the efficiency and accuracy of detecting anomalies in data generated by crucial IIoT devices.

The Industrial Internet of Things (IIoT) and Artificial Intelligence (AI) intersected in the realm of industry, transforming the industrial system to allow making real-time decisions at the edge. Latency issues, bandwidth consumption, and data security also pose a challenge to the traditional cloud-based system, and are of utmost priority in the industry where time is limited. Edge AI combines AI features with edge computing, moving the intelligence to the devices in the IIoT, allowing decisions to be made faster and more responsive to the context without depending on centralized infrastructures that might overwhelm the cloud. The paper summarizes the present-day developments in edge AI as a decision-making tool in IIoT and identifies the future directions. It discusses architectures, algorithms and applications that enable intelligent decision making at the network edge with particular focus on manufacturing, predictive maintenance, supply chain optimization and energy management. It is reviewed that facilitating technologies, such as lightweight deep learning models, federated learning, and hardware accelerators, and problems, such as scalability, interoperability, and cybersecurity are discussed. As depicted in the literature section, edge AI has been found to enhance efficiency of distributed industrial systems by reducing latencies, reliability, and independent decision-making. However, barriers such as the shortage of resources, failure to interface with current systems and standardized structures still persist. It is anticipated that future studies will be based on adaptive AI models and edge-cloud collaboration with application of 6G-enabled IIoT ecosystems. This paper summarizes the synthesis of state-of-the-art methods to inform the next generation of industrial automation and digital transformation powered by edge AI. It indicates the need of powerful, hardy, and scalable systems to open up the complete scope of opportunities that edge AI can introduce to decisions made by IIoT.

The rapid expansion of the Industrial Internet of Things (IIoT) in Industry 4.0 has increased the need for reliable anomaly detection to protect industrial services and critical operations. Deep learning–based IDS solutions can achieve strong performance, yet practical deployment may be affected by tuning effort and suboptimal search behavior when optimization is trapped in local optima, particularly when feature selection and model optimization are handled as separate stages. This paper develops an enhanced Improved Honey Badger Optimization (IHBO) algorithm and integrates it with a Bidirectional Gated Recurrent Unit (BiGRU) model for IIoT anomaly detection. To strengthen exploration and maintain population diversity during optimization, the proposed IHBO incorporates a logistic chaotic map as a controlled diversification mechanism. In addition, chaotic binary hashing is used to map continuous candidate representations into discrete binary decisions, enabling effective wrapper-based feature selection within the same optimization framework. The main novelty lies in performing integrated feature selection and BiGRU optimization within a single framework rather than treating these steps independently. Experiments on two benchmark datasets—an industrial IIoT dataset and UNSW-NB15—show that the proposed BiGRU–IHBO approach achieves 94.70% accuracy on the IIoT dataset and 99.29% accuracy on UNSW-NB15, with consistent improvements in precision, recall, and F1-score compared with baseline models reported in the same experimental setting.

Enhanced Fault Diagnosis in IoT: Uniting Data Fusion with Deep Multi-Scale Fusion Neural Network

Industrial Internet of Things (IIoT) is ushering in huge development opportunities in the era of Industry 4.0. However, there are significant data security and privacy challenges during automatic and real-time data collection, monitoring for industrial applications in IIoT. Data security and privacy in IIoT applications are closely related to the reliability of users, which is determined by user authentication that have been widely used as an effective approach. However, the existing user authentication mechanisms in IIoT suffer from single factor authentication and poor adaptability with the rapid growth of the number of users and the diversity of user categories. To solve the aforementioned issues, this article proposes a novel Authentication mechanism based on Transfer Learning empowered Blockchain, coined ATLB. In ATLB, blockchains are applied to achieve the privacy preservation for industrial applications. In addition, by introducing the transfer learning based authentication mechanism, trustworthy blockchains are built such that the privacy preservation for industrial applications is further enhanced. Specifically, ATLB first employs a guiding deep deterministic policy gradient algorithm to train the user authentication model of a specific region, which is then transferred locally for foreign user authentication or cross-regionally for another region's user authentication such that the model training time is significantly reduced. Experimental results show that the proposed ATLB not only provides accurate authentications for IIoT applications but also achieves high throughput and low latency.

Intrusion detection plays a critical role in safeguarding computer networks from malicious activities and unauthorized access. Traditional machine learning (ML) models have been widely employed for intrusion detection, but their effectiveness can be limited due to the complex and evolving nature of network attacks. An efficient Convolutional Neural Network (CNN) that utilizes one-dimensional convolutions, batch normalization, and max pooling layers is designed to capture the attack patterns in network traffic data. By leveraging the inherent feature extraction capabilities of CNNs, the proposed architecture demonstrates improved performance in identifying network intrusions compared to traditional ML models. The effectiveness of the novel CNN architecture is evaluated by conducted comprehensive experiments on two benchmark intrusion detection datasets. The designed CNN performance was compared with four other ML models such as logistic regression, k-nearest neighbors, decision trees, and random forest. The experimental results revealed that the proposed CNN architecture outperformed the traditional ML models with highest accuracy of 97.68% and 97.39% and highest F1-score of 97.68% and 97.40% for. UNSW-NB15 and CICIDS2017 datasets, respectively. The proposed CNN architecture serves as a promising alternative to traditional ML models, offering improved performance in identifying and mitigating network attacks.

Accurate rainfall forecasting is crucial in activities such as agricultural planning, water resource management, and flood preparedness, so failure to accurately forecast rainfall might undermine these activities. This study aimed to improve the accuracy of daily rainfall prediction in Hai district, Kilimanjaro region, Tanzania, a region characterised by a bimodal rainfall pattern, by comparing rainfall prediction accuracies of Traditional Machine Learning (TML) and Deep Learning (DL) models. The 25-year daily rainfall dataset for Hai district was first pre-processed by cleaning, normalisation, and splitting into training (first 60%), validation (next 20%), and test (last 20%) sets. TML models (K-Nearest Neighbours (KNN) through its variant KNN Regressor and Support Vector Machine (SVM) through its variant Support Vector Regressor (SVR)) were developed, trained, and their performances compared with developed and trained DL models (Long Short-Term Memory (LSTM) and Gated Recurrent Units (GRU)). Each model’s test performance was evaluated using the Root Mean Squared Error (RMSE) metric. The findings revealed that the DL models, LSTM and GRU, achieved test RMSE scores of 6.35 mm and 6.49 mm, respectively, while the TML models, KNN and SVM, achieved test RMSE scores of 6.61 mm and 10.42 mm, respectively. The study demonstrated that DL models are more effective for accurate rainfall forecasting than TML models in regions with complex rainfall patterns. The insights gained will guide the development of more reliable, location-specific weather prediction systems, ultimately supporting informed decision-making in agriculture, environmental management, and flood risk reduction in Tanzania and other regions facing similar climatic challenges

The Internet of Things (IoT) must seamlessly integrate different communication systems for effective real-time data analysis and decision making. It is envisaged that IoT devices will make use of several means including embedded intelligence, wired and wireless communication to achieve effective connectivity between all IoT devices. For reliable industrial IoT and the smart grid, communicating devices must co-exist with other machines and must communicate robustly without human intervention. This paper present result of channel capacity of a short-length (6 m) densely populated wired IoT channel. Signal to noise ratio (SNR) of the wired channel decreases as the number of electrical loads on the channel increases, with overall range between no-load and full-load conditions being 40.84 dB and −26.97 dB. In addition, in-depth coverage is given to how time-series edge computing may be done in real time for large scale networks such as the smart grid and manufacturing industries. The result of this work will be useful in guiding standard bodies and manufacturers of industrial IoT and smart grid communication devices on channel capacity and effective machine to machine communication rates when all edge devices are communicating simultaneously over a densely populated, short length wired IoT channel.

Efficient railway kilometer marker recognition via spatio-temporal slimming and multi-view fusion

Enabled by the increasing availability of sensor data monitored from production machinery, condition monitoring and predictive maintenance methods are key pillars for an efficient and robust manufacturing production cycle in the Industrial Internet of Things. The employment of machine learning models to detect and predict deteriorating behavior by analyzing a variety of data collected across several industrial environments shows promising results in recent works, yet also often requires transferring the sensor data to centralized servers located in the cloud. Moreover, although collaborating and sharing knowledge between industry sites yields large benefits, especially in the area of condition monitoring, it is often prohibited due to data privacy issues. To tackle this situation, we propose an Autoencoder-based Federated Learning method utilizing vibration sensor data from rotating machines, that allows for a distributed training on edge devices, located on-premise and close to the monitored machines. Preserving data privacy and at the same time exonerating possibly unreliable network connections of remote sites, our approach enables knowledge transfer across organizational boundaries, without sharing the monitored data. We conducted an evaluation utilizing two real-world datasets as well as multiple testbeds and the results indicate that our method enables a competitive performance compared to previous results, while significantly reducing the resource and network utilization.