Modern Trends in Pervasive Intelligence

Data mining and knowledge discovery are crucial in pervasive computing systems, which are distinguished by dynamic user behaviours, contextual variations, and different management requirements. Unlike classical data mining, which uses static databases or data warehouses, ubiquitous data mining works at the communication level, frequently in real time and on flowing data. Streaming data mining is supported by a variety of technologies, each with unique requirements imposed by the nature of pervasive data, such as variability, volume, and velocity. Data transmission time, contextual distinction, and processing accuracy are all important factors to consider. This study investigates the capabilities of current data streaming infrastructures and identifies issues particular to pervasive contexts. We demonstrate the limitations of current systems in handling real-time, adaptive mining activities using hypothetical application scenarios and instructive case studies. These examples demonstrate crucial gaps and growing opportunities, particularly in real-time decision making and adaptive intelligence. To summarise, pervasive data mining is a dynamic and expanding research subject with tremendous promise for transdisciplinary applications in social sciences, healthcare, transportation, and smart environments. 

This study begins with the notion that Human-Computer Interaction (HCI) and affective computing have a common goal: understanding, capturing, and altering human affect and physiological states. However, well-established methods from psychology, communication, and media studies have received limited attention in HCI and affective computing. This presentation provides: (1) a rationale for incorporating insights from these disciplines into effective research; (2) an outline of methodological frameworks used in humanistic studies of affect; and (3) examples of how these methods might enhance and inspire new paths in HCI and affective computing. The paper finishes by discussing the benefits and limitations of interdisciplinary collaboration in this context.

Imagine a world in which computers understand not just our commands but also our emotions. This is the promise of incorporating empathy into human-computer communication as technology progresses; integrating Emotional Intelligence (EI) into robots becomes more possible. However, this trip is not without obstacles. People's belief in the power of emotions is what hinders the advancement of Artificial Intelligence (AI). AI emotions enhance human-machine cooperation, fostering empathy and trust. AI systems can handle emotional input by incorporating EI and empathy, affecting human-computer interaction. This section explains how to connect naturally with robots using their emotional model. The incorporation of EI and empathy in AI systems improves human-computer interaction and collaboration. This chapter explores the design and implementation of empathy in AI, namely the machines' ability to feel, grasp, and react to human emotions. This raises ethical concerns, such as the distinction between simulated emotion and actual human connections. This chapter highlights the importance of emotional intelligence systems in shaping the future of human-computer interaction and presents technologies with human-like thinking and emotions.

Modern IoT sensor networks are complicated, and intelligent decisionmaking is critical for energy efficiency, communication costs, and network flexibility. Traditional rule-based and centralised systems have substantial limits due to the long computing delays needed in such large-scale deployments, as well as the volume of data generated. As a result, Multi-Agent Reinforcement Learning (MARL) appears to be the best solution: the sensor nodes can function as autonomous agents, learning and adapting to changing environmental conditions through decentralised decision making. This chapter explores the use of MARL to improve packet transmission, network longevity, and power consumption in sensor networks. Cite the efficacy of several MARL architectures, including Centralised Training-Decentralized Execution (CTDE), Fully Decentralised Learning, and Hierarchical MARL, in terms of sensor coordination. An experimental framework is built on OpenAI Gym, TensorFlow, and NS3 and used to develop MARL-based sensor nodes that dynamically change their transmission power and task scheduling. The performance indicators (e.g., Packet Delivery Ratio (PDR), energy usage, and network efficiency) are examined, demonstrating that MARL outperforms conventional optimisation methods. It is also discussed in the context of real-world applications such as smart cities, healthcare monitoring, environmental sensing, and industrial IoT (IIoT). Finally, the chapter provides insights into scaling problems, integration with new technologies such as 6G and quantum computing, and security enhancement via blockchain. This study also incorporates recent works such as: [Chebyshev Polynomial Based Emergency Conditions with Authentication Scheme for 5G-Assisted Vehicular Fog Computing], [ANAA-Fog: A Novel Anonymous Authentication Scheme for 5G-Enabled Vehicular Fog Computing], [Password-Guessing Attack-Aware Authentication Scheme Based on Chinese Remainder Theorem for 5G-Enabled Vehicular Networks], [ECA-VFog: An Efficient Certificateless Authentication Scheme for 5G-Assisted Vehicular Fog Computing], [CM-CPPA: Chaotic Map-Based Conditional Privacy-Preserving Authentication Scheme in 5G-Enabled Vehicular Networks], [Provably Secure with Efficient Data Sharing Scheme for Fifth-Generation (5G)-Enabled Vehicular Networks without RSU]. 

In recent years, the growth of “smart” products, those that integrate several technologies and are equipped with a variety of sensors and cameras, has become more prevalent. The software contained in these products is becoming more complicated, demanding effective support systems to speed up and improve the product development process. As a result, developing next-generation systems and product development technologies that not only improve everyday living but also promote industrial growth has become a major problem. One interesting option is the development of open platforms infused with developing intelligence, which will allow developers to successfully address a wide range of technological difficulties. These platforms should be branded and offered as unique industry solutions. This chapter examines a number of attempts in applied evolutionary computation and optimisation. These innovative approaches play an important part in the development of modern industrial products. In addition, current subjects such as ant colony optimisation and other social intelligence paradigms are investigated. The chapter's title concisely expresses the key goals and significance of the underlying study. 

This study investigates the interaction of pervasive intelligence with evolutionary computation and optimisation techniques. Pervasive intelligence is one aspect of pervasive technology, which includes wireless communication, wearable computing, and context-aware computing. Evolutionary algorithms and optimisation techniques are quickly gaining popularity for handling a wide range of real-world challenges. The purpose of this work is to support the concept of included pervasive intelligence systems with automatic parameter tuning procedures through the use of ubiquitous technology. The existing pervasive intelligence systems' limitations are projected to be greatly decreased by improving their adaptability and plug-and-play capacity. It is assumed that such innovative systems will benefit not only system designers but also end users.

Many machine learning researchers believe that poor prediction results from a model are caused by either noisy data or creating a model with the incorrect method, but the reality is that the ideal parameters necessary are selected based on the dataset. So, rather than standard methodologies, the proposed research focuses on feature selection using genetic algorithms. Optimisation is the concept of determining the best values for the parameters connected with the learning process. During the optimisation process, the algorithm takes some initial values and continuously updates the position, velocity, and other parameters. Firefly is a metaheuristic algorithm that computes the fly's attraction using the Euclidean distance. The attractiveness of the suggested algorithm, the adaptive firefly algorithm, is computed using the KNN algorithm, with K as one of the objective functions and the second objective function being to reduce the number of features. The firefly algorithm's key components are light intensity and brightness, both of which are exponential. 

AI and machine learning are transforming the way we solve problems in our data-driven environment. AI allows machines to perform tasks that need human intelligence, such as comprehending language and making decisions. ML enables systems to learn from data and improve on their own. In recent years, advances in computing, massive data, and deep learning have propelled AI and ML into critical domains such as health, finance, transportation, and entertainment. Computer vision, language processing, and neural networks are all powered by cutting-edge technology that is used to detect fraud, run medical tests, and even drive self-driving cars. While AI and ML hold significant promise, they also raise concerns about justice, privacy, job loss, and environmental harm. Strong standards, clear explanations, and global collaboration to investigate how AI is evolving are effective ways to address these issues. New trends include AI that makes things, individuals working with AI computing on the edge, and AI assisting with climate issues. Combining AI and quantum computing might lead to significant advances in the discovery of new pharmaceuticals and materials, while a new method of learning ensures that data remains private. AI will also help to address major global issues such as poverty and disaster response. To summarise, AI and ML are transforming industries and solving complex problems, but we must be cautious about how they influence people and society. We can use these technologies to build a more intelligent, egalitarian, and sustainable future. Recent works also highlight AI–5G integration and VANET authentication schemes such as CLA-FC5G, FCA-VBN, and D-BlockAuth.

Climate variability, inefficient resource management, and a lack of real-time data are among the issues facing agriculture today. Many rural farmers deal with smartphone inaccessibility and inadequate digital literacy, making it difficult to gain critical agricultural insights. To address these difficulties, we offer an AI-powered Smart Board, which is a solar-powered, offline-compatible device equipped with IoT sensors (soil moisture, temperature, humidity, and pH sensors), drones, and real-time data processing capabilities. It sends voice-based alerts in local languages with climate change warnings, crop care suggestions, and real-time government updates (minimum support price, subsidies, and market trends). It also collects crop data and recommends pest-prevention and irrigation management strategies. A voice-enabled query system enables farmers to ask questions and receive AI-powered responses, while radio integration provides emergency alerts and expert views. This technology, which features a simple touch interface and an AI-powered hotline, empowers farmers, boosts productivity, and supports sustainable agriculture, ensuring a wiser, data-driven future.

Industry 4.0 is transforming automation by integrating IoT, robots, and artificial intelligence, resulting in intelligent and autonomous systems for industrial applications. With the growing demand for automation in industrial settings, robotic arms are critical for precision-based operations, including manufacturing, assembling, and material handling. The development of robotic arms for industrial automation requires precise modelling, control, and validation. This work advances industrial automation by improving the integration of robotic arms into smart manufacturing and autonomous operations. The study describes a design using Autodesk Fusion 360, simulation with the ROS framework, and validation with hardware integration and an object detection model on a robotic arm. To validate the design, FEA and generative design analysis for mass and volume optimisation are done on the model, revealing that steel and CFRP are the optimal materials for the desired strength. Furthermore, the designs' kinematics and perception are tested on Gazebo and RViz, while the final trained model for object identification using YOLO V8s is tested on the KSR10 robotic arm. The findings highlight the importance of simulation-driven research in improving robotic arm performance and guaranteeing seamless integration into new Industry 4.0 automation environments. 

Pervasive intelligence, powered by data-driven decision-making, automation, and improved user experiences, is transforming the startup ecosystem via Artificial Intelligence (AI), the Internet of Things (IoT), and big data analytics. The chapter discusses the impact of pervasive intelligence on operational efficiency, market adaptation, and innovation, as well as how it might help startups improve their decision-making processes and expand. Startups can acquire a long-term competitive advantage by leveraging intelligent technologies to foresee market trends, manage resource allocation, and personalize consumer interactions. Pervasive intelligence also increases agility, allowing startups to scale efficiently and respond swiftly to changing market conditions. The study also identifies potential challenges, such as the need for qualified workers, complex technology integration, and data breaches. It identifies potential hurdles, such as data privacy problems, technology integration complexities, and a need for specialised labour. Through case studies and empirical analysis, this chapter explains how companies can strategically use pervasive intelligence to accelerate growth, improve consumer engagement, and build a footing in competitive markets. However, it also provides tremendous opportunities for entrepreneurs to innovate, develop quickly, and establish a competitive edge in an increasingly digital economy. This chapter examines real-world applications and case studies to show how companies can strategically harness pervasive intelligence while navigating its hurdles to generate long-term growth. 

Additive manufacturing's original goal of producing prototypes quickly has led to significant advancements, which have caused it to move toward large-scale precision manufacture. One of the primary drivers of this change is the implementation of Artificial Intelligence (AI) mechanisms, which enhance design skills and optimize manufacturing and quality management techniques. The presented review follows the development of AI-based AM since the first introduction of using AI to accelerate the process of creating prototypes, to its recent potential to deploy intelligent autonomous manufacturing systems. This study explains three fundamental technologies of AI used in AM, and how the technologies are applied towards the manufacturing process, with the resulting outcomes of such an application. Upon identifying the existing problems with standardized data systems, the research presents tactical ways of developing adaptive AI-AM systems that are sustainable. The introduction of AM and AI technology has already started transforming the standards of manufacturing in many industries.