Industry 4.0 refers to the phase of transition that is taking place, enabling automation and data interchange in industrial technologies and processes. Fog computing architecture can provide real-time processing, nearby storage, extremely low latency, dependability, large data rates, and other requirements for industrial Internet of Things (IIoT) applications. In the context of IoT applications, fog infrastructure and protocols are the main areas of interest. The phrase “fog computing,” sometimes known “edge cloud,” is a new paradigm. Between edge devices and Cloud Core, it adds another layer. Along with providing computing, storage, and networking capabilities, it also fills a need left by the cloud. The main features of fog computing are covered in this chapter, along with current research on the subject and a focus on the difficulties encountered when creating its architectural design.
Top1. Introduction
In reference to the industrial revolution, fog computing, often known as “fogging,” is an expanded type of cloud computing that provides applications and services (low latency and high processing) to autonomous heterogeneous devices inside an industry. The purpose is to place intelligence control, processing, and storage close to data devices. Real-time services with high data processing, maximum capacity, and scalability are crucial for industry 4.0. Due to its many advantages over cloud computing, fog computing offers the finest options for this kind of setting. By introducing the notion of network edge computing, the extension of cloud computing seeks to reduce the load on the cloud.Low latency and improved cache memory are necessary for real-time services and decision-making processesin industrial automation(J. Li, F. R. Yu, G. Deng, 2020)
On the basis of geographic distribution, consider real-time applications, low latency, location awareness, the number of nodes, and edge devices with cache support. In between the internet's cloud of devices and end users' devices are virtualized nodes, also known as cloudlets or fog nodes. With superior QoS parameters performance and coverage of crucial IoT criteria, fog computing offers services and applications similar to those offered by clouds. The following are significant benefits of fog computing that affect IoT adoption .When data is stored on network edge nodes, there is no longer a need to retrieve it from distant clouds, which ends transmission delays.Fog computing helps IoT applications process and analyse data more quickly.The processing and computation time will be reduced by data storage on edge nodes.Cache-enabled nodes will stop transmitting pointless data(A. V. Serbanescu, S. Obermeier, 2020)
Industrial cyber-physical systems make it possible for physical objects and processes to be closely associated with computing, communication, and control systems online. Transmissions are facilitated through cyberphysical interfaces that connect the two worlds. Employingwireless sensors, mobile devices like smartphones and tablets, and others. These cyber-physical interfaces conceptually display “cyber twins,” which are virtual representations of real-world physical things in the cyberspace. To provide operational insights and guide decision-making, these virtual objects may then be individually and/or collectively analysed, questioned, or simulated (A. Banga,et al., 2019)
The ML models and associated IoT data operations include feature extraction, grouping, and classification. Analytics, such as k-means, k-nearest neighbours (k-NN), support vector machines (SVM), liner regression, and DNN, have been thoroughly examined in the design of ML algorithms faces additional challenges because of the resource-constrained context. There are certain universally applicable algorithms that adapt to resource and distribution limits, such as k-NN and some specific neural network techniques. The second, however, necessitates often time-consuming fine-tuning of the granularity of the environment sensor data and the environment model. One possibility is early sensor fusion utilising a 3D voxel environment model (F. Daryabar,et al., 2021).
In addition to using edge intelligence to lessen the quantity of data transmission and storage required on the cloud, the fog-level IoT system produces a faster and more effective knowledge transfer to the cloudAccording to the authors in, in order to allow regional IoT networks to execute edge analytics, future IoT systems need incorporate intelligence at the fog layer. By using local storage, fog-layer analytics, and decision-making, data management may also be enhanced. We need algorithms that are resource-efficient, using little memory and energy, to make this happen. Microsoft recently demonstrated ML models that can be used by small IoT devices.The created algorithm has been disclosed, and it may be used with an Arduino Uno board.(A. Trombetta, 2021).