Abstract
The use of AI algorithms in the IoT enhances the ability to analyse big data and various platforms for a number of IoT applications, including industrial applications. AI provides unique solutions in support of managing each of the different types of data for the IoT in terms of identification, classification, and decision making. In industrial IoT (IIoT), sensors, and other intelligence can be added to new or existing plants in order to monitor exterior parameters like energy consumption and other industrial parameters levels. In addition, smart devices designed as factory robots, specialized decision-making systems, and other online auxiliary systems are used in the industries IoT. Industrial IoT systems need smart operations management methods. The use of machine learning achieves methods that analyse big data developed for decision-making purposes. Machine learning drives efficient and effective decision making, particularly in the field of data flow and real-time analytics associated with advanced industrial computing networks.
TopIntroduction
The world has witnessed a change and important contribution in the area of artificial intelligence, industry, robotics and automation technologies. These technologies have a strong impact on knowledge management process which made the foundation of decision making in several fields such as industrial, home, health, financial and many other to create a smarter environment (Ruhul et al, 2020). Intelligent information technology is a new novel feature for industrial Internet of Things (IIoT). Most of the industries are attempt to automate the process of developing and manufacturing products. Using Machine Learning (ML) in Industry 4.0 is a key feature to obtain an IIoT. Due to increasing scale of deployed terminals IoT in industrial, the IIoT becomes heterogeneous, diverse, and dynamically changeable.
Generally, IIoT system consists of an intelligent control technology, network communication technology and information processing technology. ML is used to connect the real world virtually in intelligent ways. ML is a new topic used in IIoT (Ruhul et al, 2020). Many businesses deal with ML models and algorithms to reduce the production and operation costs. ML can be implemented in smart manufacturing, predictive maintenance in industrial, drug discovery, autonomous vehicles and machines for industry, pattern imaging analytics and software testing (Baotong et al, 2019).
In Industry 4.0 machines interact with their environment intelligently and learn to understand adapt their operational behaviour and processes. ML algorithm improves the quality, cost and flexibility of production process (Wen et al, 2019). Nowadays, most of the factories are trying to automate the process of manufacturing products. Due to this kind of automate the system exposed so you need a high-level security. ML is a part of intelligent system in Industry 4.0, is broadly implemented in different scope in manufacturing are to extract knowledge out of existing data (Inés et al, 2018). It provides decision-making process in manufacturing system. But the goal of the ML techniques is to detect the patterns among the data sets or regularities that describe the relationships and structure between those sets (Asharul et al, 2020).
The chapter describes the concepts of ML use in IIoT, the architecture of ML in IIoT, in addition to the optimization methods, algorithms and applications that used in ML for IIoT. The chapter is organised as follows, section 1 provides a brief conception about the Industrial Internet of Things (IIoT) and the Machine Learning contribution to its applications, in addition to recently works related to the machine learning use in IIoT. The background and chapter motivation is presented in section 2. The machine learning architecture layers and their operational functions are illustrated in section 3. In section 4, different machine learning methods are discussed in technical and performance. The implementation of machine learning algorithms in the industrial IoT applications has been discussed in section 5. In section 6, some examples of machine learning applications in IIoT are reviewed. Finally, the industrial IoT future trend and expectations are presented in section 7, and followed by the chapter conclusion.
Key Terms in this Chapter
Support Vector Machine: Is a machine learning approach that uses a linear classifier to classify data into two categories. SVM is the most widely used ML technique-based pattern classification technique available nowadays. The SVM classifies data in feature space based on a hyperplane that separates patients and controls according to class labels. It works well for a high-dimensional dataset by establishing a linear decision boundary.
Blockchain: Is a system for recording information in such a way that it is difficult or impossible to alter, hack or deceive the system. It's a digital record of transactions that are replicated and distributed across the entire network of computer systems on the blockchain. Blockchain, sometimes names as Distributed Ledger Technology (DLT), makes the history of any digital asset unalterable and transparent through the use of decentralization and cryptographic hashing.
Hyperplane: In geometry, a hyperplane is a subspace whose dimension is one less than that of its ambient space. If a space is 3-dimensional then its hyperplanes are the 2-dimensional planes, while if the space is 2-dimensional, its hyperplanes are the 1-dimensional lines. A Support Vector Machine (SVM) performs classification by finding the hyperplane that maximizes the margin between the two classes.
Knowledge Base (KB): Is a technology used to store complex structured and unstructured information used by a computer system. The initial use of the term was in connection with expert systems, which were the first knowledge-based systems. In IoT, to achieve intelligent interactions without human intervention, including devices automatically communicating with each other, making decisions and performing appropriate actions. Knowledge base (KB) provides the abilities of data analysis and logical reasoning, enabling the devices to think autonomously.
Decision Tree: Is a flowchart-like structure in which each internal node represents a “test” on an attribute. i.e., whether a coin flip comes up heads or tails, each branch represents the outcome of the test, and each leaf node represents a class label. Decision taken after computing all attributes.