Cognitive Computing for Smart Environments: Survey, Technologies, and Research Challenges

Introduction

In order to develop new applications and services and accomplish shared objectives, physical, virtual, and digital worlds are coming together to build smart environments that improve the intelligence of cities, transportation, energy, and many other domains. IoT plays a significant role in making equipment accessible that was previously locked away in well-designed silos in production processes. Such development could enable IoT or ICT to penetrate more advanced industrial systems that are digitalized or smarter. The Internet of Things (IoT) will link the industrial unit to a vast array of cutting-edge applications that are used throughout the production. Starting with connecting the workplace to the smart grid, diffusion of facilities in the production service, and/or allowing for more agility and flexibility inside the manufacturing organisations themselves are all possible options. Automation, operations information, and sophisticated analytics are three crucial productivity variables that are integrated in smart Manufacturing, an example of a smart environment. Some of these elements, like tools and machines operating on open platforms and with the ability to “think,” lead to the creation of organisations that may be able to collaborate with one another, analyse data to predict disasters, organise themselves, and adjust to changes in the industrial practise itself. By improving perception and appropriate use of contextual information related to processes and goods, it is hoped to boost throughput in a process and ultimately across the whole pricing chain. This will allow real people to get correct information at the right moment. Most organisations and several industry studies indicate that smart industrial systems are not yet ready to be implemented. Furthermore, according to the studies, technology by itself is insufficient for smart or cognitive manufacturing systems to succeed and must instead be incorporated into more comprehensive strategic design, planning, and management procedures. In addition, additional factors, such as architectural paradigms, frameworks, and models, are necessary for a productive manufacturing system, whether it is intelligent or cognitive. In light of this, the architectural design of CIoT-based systems, including cognitive manufacturing systems, is concerned with architecture styles, networking and communication, cooperative data processing and security, smart things, services of the web with corresponding applications, business models with their corresponding processes, etc.

Technology-wise, the architectural design of the Internet of Things, for instance, should take into account scalability, modularity, extensibility, and interoperability across heterogeneous devices. Alternatively, the installed base for the Internet of Objects will increase to roughly 212 billion items by 2020, including over thirty billion linked things or gadgets, predicts industry research company IDC. IDC acknowledges that this development is primarily driven by intelligent systems that will be installed and collecting data from both consumer and business applications. Prior literature was methodically read while keeping an eye on the manufacturing instances, and numerous alternative methods were looked at. One of the goals of this investigation is to assess what cognitive IoT-based technology, architectural frameworks, and alternative approaches are available and why they are needed. Another goal is to determine how cognitive capability was developed in order to achieve more practical, acceptable, and unstoppable smart or cognitive manufacturing.