Journal of Intelligent Systems and Internet of Things

Smart cities represent a transformative vision of urban living, where technology seamlessly integrates with the urban landscape to enhance sustainability and quality of life. Central to this vision is the effective management of environmental factors, particularly air quality and temperature. This paper presents a comprehensive study on real-time environmental pollution detection within smart cities, utilizing Internet of Things (IoT) sensors. We explore the intricate relationships between air pollutant indicators (o3_AQI, no2_AQI, co_AQI, and pm25_AQI) and temperature, shedding light on the dynamic interactions that underlie urban atmospheric conditions. Our research employs a robust dataset and employs statistical analysis, including Ordinary Least Squares (OLS) regression, to uncover significant correlations between key environmental variables. These insights not only contribute to a deeper understanding of urban pollution dynamics but also enable the development of predictive models for temperature fluctuations based on pollutant levels. Such models hold promise for proactive environmental management and public health interventions. Furthermore, our study highlights the pivotal role of IoT sensors in revolutionizing smart city governance, offering real-time data-driven solutions for sustainable urban living. As cities worldwide strive to enhance their environmental resilience, this research provides valuable insights and tools for harnessing the potential of IoT technologies in the pursuit of cleaner and more livable urban environments.

The Industrial Internet of Things (IoT) has ushered in a new era of predictive maintenance, revolutionizing the way industries manage and maintain their critical equipment. This paper presents a comprehensive exploration of predictive maintenance strategies, with a primary focus on early fault detection and classification in industrial equipment. We introduce the "Triplet Fault Injection Algorithm," capable of injecting three distinct fault types—spike, bias, and stuck—into sensor data for realistic and rigorous testing. Leveraging this algorithm, we employ the powerful Extreme Gradient Boosting (XGBoost) machine learning approach to detect and classify these faults. Our experimental results showcase the superiority of XGBoost over baseline machine learning methods, across various data types commonly found in industrial equipment. The consistent higher accuracy and F1 scores obtained with XGBoost underscore its effectiveness in minimizing false alarms and enhancing the reliability of early fault detection. Moreover, we discuss the transformative role of IoT in predictive maintenance, highlighting its potential to optimize equipment performance and reduce downtime in the industry 4.0 landscape. This paper contributes valuable insights and empirical evidence to the domain of predictive maintenance in IoT-enabled industries, emphasizing the significance of early fault detection for efficient and cost-effective maintenance practices.

Diseases in crops lead to decreased production, which can be addressed through consistent surveillance. Manual surveillance of crop diseases is both arduous and prone to mistakes. The timely identification of crop leaf diseases using Computer Vision and Artificial Intelligence can aid in minimizing the negative impact of diseases and address the limitations of continuous human surveillance.  To classify chili crop diseases, this research paper introduces a new deep feature extraction model based on Transfer Learning using ResNet50, MobileNet, EfficientNetB0, and multiple classifiers. On Plant Village dataset related to the diseases of the chili crop and Private data set, the proposed method was trained and tested. And also analyzed the results by comparing the performance of the pre-trained deep learning models on original data and data filtered through the Image filtering mechanisms and proposed method on the plant village dataset and private dataset, the highest performance accuracy is 99.6% with ResNet50 and the faster CPU time for feature extraction is 29.3 seconds using MobileNet. Comparing the suggested model to the most advanced deep learning models reveals greater accuracy with fewer computational resources.

Wastewater treatment procedures (WWTP) rely heavily on accurate forecasting of treatment results to keep oxygenation levels under control. Conventional biochemical mechanism-driven approaches provide poor results, mainly due to complicated and redundant system factors. As sewage treatment operations expand fast, automated operational solutions are needed to achieve this goal. In the research, data mining was used to model the WWTP to predict the outcomes based on input circumstances and the amount of oxygenation provided to the system. Combined Sustainability Research for Wastewater Treatment procedures (CSR-WWTP) is proposed in this research. Data-driven approaches to modeling WWTP have already been developed but do not consider long-term treatment procedures and structure features. Forecasting and management for the WWTP are described in this article using a combination of convolutional neural networks (CNN) and recurrent neural networks (RNN). The first stage utilizes the CNN structure to dynamically learn and encrypt the local features of each WWTP timestamp in the first phase. The RNN model is applied to the WWTP to express global sequence characteristics using local feature encryption. For this purpose, it conducts a huge number of tests to assess the performance and accuracy of the proposed forecasting framework.

A smart city's smart economy thrives in various areas, including political strategy, operational efficiency, and innovation management. Business models in smart urban must be based on a new sustainable development strategy, one that conserves natural resources while safeguarding the environment. Therefore, this paper proposes Statistical Business Models (SBM) to enhance the business strategies for developing the economy in smart cities. Economic status in smart cities and changes in business models are part of SBM, a set of design concepts. Smart Business Models (SBM) are business strategies that take advantage of current economic situations by leveraging the power of influential smart communities. The implementation of data systems and business models is the foundation for a systematic study of managing the economy in a smart city. There are several connections between SDM's critical assessments of business models and the global economy and the business models. The experimental findings suggest that the proposed SBM achieves the highest statistical rate with sales revenue up to 95.23 %, gross margin ratio of 80.5%, consumer satisfaction ratio of 96.34%, efficiency ratio of 93.82%, and maintenance cost ratio of 15.08% compared to another existing method.

Community detection in complex networks has become an important step in understanding the structure and behaviour of networks in many fields. However, both standard node clustering and the relatively new link clustering methods have problems that make it hard to find combined clusters. Networks have been used to depict many types of real-world systems, such as those involving the transmission of information, the movement of funds, and biological processes. Communities are key structures for comprehending the structure of actual networks. The purpose of community detection is to identify meaningful subsets of these networks. Mesoscopically, a community consists of highly interconnected nodes within each subcommunity yet less strong connections across subcommunities. Communities can share a node or numerous nodes with overlapping. Evaluating the performance of a community detection method is crucial. Grouping the network's nodes into a family of subsets called clusters such that each cluster comprises similar nodes concerning the overall network structure is the problem of detecting overlapping communities in a network. Meanwhile, it has been shown that many methods for finding cluster centers have inherent flaws. Most methods are vulnerable to initial seeding and built-in variables, while others fail to highlight substantial overlaps. This article proposes the Structural Centrality Approach for Local Overlapping Community Detection (SCA-LOCD). It provides a novel approach to regional development that emphasizes the role of systems in identifying cluster centers. The fundamental concept behind this strategy is to identify structural centers in societies with coherent structures and then increase these centers using weighted methods and search engine techniques. Experimental results on synthetic and network systems show that the suggested technique is efficient and fascinating for detecting overlapped communities. It shows the success of regional extension strategies in identifying coherent groups and producing reliable classification results.

Credit scoring has grown in importance and has been thoroughly researched by banks and financial institutions. The amount of redundant and irrelevant features present in credit scoring datasets, however, reduces the classification accuracy. As a result, employing effective feature selection methods has become essential. In this study, a hybrid feature selection approach that combines the backpropagation neural network (BPNN) classifier and the pigeon optimization algorithm (POA) is suggested. With hybridization, the POA works to choose characteristic subgroups through the feature selection (FS) process, and the BPNN then assesses the chosen subsets using a fitness function. The experiment findings show that the suggested hybrid technique outperforms other competing approaches in terms of evaluation criteria, according to three benchmark credit scoring datasets.

Athletes health monitoring plays a vital role because the changes in their heart rate reduce their physical activity and contribution. The changes in athlete activities cause developing risk that affects their outcome. Therefore, athletes' heart rates should be monitored frequently to minimize the risk factors and improve their health. This work uses wearable sensor devices to monitor their health condition continuously. The wearable devices on their health record their Electrocardiogram (ECG), which is transferred to the health care centre. With the help of the ECG, this work Sportsperson Heart Rate Monitoring (HRMS-SP) is created. The gathered ECG information is processed using the Fuzzy Clustering (FC) algorithm to predict the Heart Rate Variability (HRV). According to the HRV value, athlete's mental stress level and their sports contribution were also investigated to minimize the computation complexity. In addition, the wearable device-based collected information was investigated using the fuzzy and big data analytics used to monitor people frequently. The predicted information is used to monitor, treat, prevent, and predict the sports person's activities effectively. During the analysis, Hadoop, Visualization, and data mining processes are applied to extract the health information from large datasets that are used to improve the athlete health monitoring systems.

Vehicular ad-hoc network (VANETs) is a promising technology that is used in the maximum of the applications of intelligent transport systems (ITS). VANETs become more attractive due to their communication methods such as vehicle-to-vehicle (V2V) and vehicle-to-roadside unit (RSU) communication. VANETs consist of a few special features such as unpredictable mobility, dynamic inter-vehicle spacing, high speed and so on which make communication ineffective. These features network delay and routing overhead increased which affects the stability and reliability of the network. In this paper Path Scheduling and Bandwidth Utilization for VANETs (PSBU-VANETs) are proposed. Through the path scheduling process, the changing topologies are predicted that the prediction path is scheduled for data transmission which leads to reduce the delay and overhead of the network. Through the effective utilization of bandwidth, the throughput and delivery rate of the network are increased. The simulation is performed in NS2 and SUMO and to measure the outcome the parameters which are considered are packet delivery ratio, end-to-end delay, routing overhead, and throughput. To perform a comparative analysis the results of the proposed PSBU-VANETs are compared with the earlier research works such as TDG-VANETs and ICB-VANETs. The proposed PSBU-VANETs achieve a high packet delivery ratio and throughput as well as lower end-to-end delay and routing overhead when compared with the earlier approaches.

Recognition and modelling of driver behavior (DB) have lately been crucial in intelligence transportation systems (ITS), human-vehicle, and intelligent vehicle systems (IVS). The evidence that drivers are distracted most often causes accidents and incidents involving vehicles is growing. Using camera sensors in the vehicle or sensors worn by the driver can help detect and prevent drivers from engaging in distracting behaviors like talking on the phone, eating, drinking, adjusting the radio, interacting with navigation systems, or even combing their hair while behind the wheel. However, this system requires a lightweight data processing module and a powerful training module for real-time detection. Data must be collected from certain cameras or wearable sensors to detect distracted drivers and ensure a rapid reaction from the administrator on safe driving. Therefore, this paper suggests a Machine Learning Driver Distraction Prediction Model (MLDDPM) with a decision-support system (DSS) that can alert the driver to possible dangers on the road by analyzing both internal (vehicle parameters) and external (road infrastructure messages) data. This research MLDDPM employed semi-supervised algorithms to reduce the expense of labelling training data for driver attention detection in actual driving scenarios. Two attentive and cognitively distracted driving states were used to assess support vector machines: i) as a supplementary parameter for the aggregate risk assessment of driving and ii) as a parameter for providing the driver with the most appropriate message type on possible road dangers. Finding the optimal approach to driver assistance to guarantee secure transportation is the primary goal of this work.

Rapid changes in modern technology and sports have impacted society and lifestyle. Augmentative and Alternative Communication (AAC) technology helps to speak and play videos in various sports applications. In the current sports event, AAC's utilization to validate the players' complex moves exclusively has been considered a significant challenge that includes athlete moves in athletics and penalty shots in Soccer. Deep Learning-based Video Segmentation and Video mining (DL-VSVM) with eyeball tracking assistance are proposed to validate the task modeling of sports event video streaming in AAC. The user could select the specific event in the sport and sub-event using eyeball tracking assistance. The AAC is installed with unique icons to identify circumstances. A deep learning-based Sports Task model is created to recognize the required data to be streamed, and the model will help them view the specific sports event they need to watch. The numerical outcomes demonstrate that the suggested DL-VSVM model enhances the segmentation accuracy ratio of 95.3%, tracking ratio of 97.6%, prediction ratio of 98.7%, and reduces the cost function of 5.6% and the error rate of 20.1% compared to other existing models.

The development of power industries creates impacts on the intelligent power grids. The power grids are more valuable for transmitting information over the network. Several intermediate activities influence the networks, which are interrupted by traffic, creating network security issues. Therefore, the threats highly influence power grids, and the number of attacks also increased gradually. Several conceptual approaches are introduced to overcome the security issues; however, computation complexity is still a significant problem while detecting network anomalies. This research problem is overcome by applying the Organic Combination of Various Deep Learning (OC-VDL) approach. The introduced method observes the industry standards with the help of the Innovative Blockchain Network (IBN). During this process, IBN observes the infrastructure using the communication protocol and Manufacturing Internet of Things (IoT). The collected information is processed with the help of the Intense Autoencoder Classifier Model (IACM), which manages bilateral traffic control and helps predict abnormal activities. The effective prediction of network traffic minimizes the intermediate activities and improves the overall security up to 98.8% accuracy.

In the Smart city environment, sustainable sewage and wastewater management planning plays a crucial role in industry development. Wastewater management is a serious issue with inadequate treatment, which reduces the smart city efficiency. Therefore, this research work concentrates on creating the Strategic Planning Adaption framework (SP-AF) using the Recurrent Neural Networks (RNN). This framework intends to manage the sewage and wastewater in smart cities. The sewage-related information is continuously collected by a recurrent network that identifies and tracks the wastewater and sewage in the smart city. The SP-AF framework analyses sustainable planning and managing wastewater by understanding the waste origin. In addition, the framework has been generated by understanding the wastewater knowledge, and the required actions are carried out. Then the effectiveness of the wastewater management system efficiency is compared with the existing approaches.

Risk Management is an important task that helps to monitor the business application to eliminate the political, financial, cultural, and social consequences. The organization's risk decision is affected by several characteristics, such as lack of accountability and risk decision-making. The difficulties are resolved by applying the Machine-Learning related Business Decision Making Approach (ML-BDMA). The created framework helps to reduce the difficulties in decision-making while managing the organization's risk. The Business Decision Making process works along with the Optimistic Predictive Techniques (OPT) that are used to identify the risk which leads to attaining the business objective. This process categorizes the risk according to the qualitative characteristics of business data. The system's effectiveness was evaluated using the experimental result in which the system ensures a 98.93% performance rate, 92.25% reliability rate, 93.47% authenticity rate, 91.11% risk management rate, and 97.77% development rate while making a business decision.

The science of achieving a healthy mind, body, and spirit through objectives and activities is known as personal health (PH). We must be aware of our mental, bodily, and social well-being. The term "hygiene" refers to a wide variety of healthy behaviours. Individuals' healthcare costs and quality of life increased by avoiding or reducing the long-term effects of the disease through knowledge and skills. Biomedical apparel includes sutures, vascular grafts, and biodegradable clothes (BC). Biomedical clothing is anything implanted or incorporated into the human body and used near tissue, blood, or cells. Quick, dependable, and energy-efficient connectivity between wireless sensor networks is necessary (WSNs). Physical layers, media access control, networking layers, and control requirements must be co-designed. For those with lesser means, health insurance will increase in cost. There are difficulties with privacy and cyber security, a higher chance of malpractice claims, and increased time and financial expenditures for doctors and patients. In this study, wireless sensor networks-based personal health biomedical clothing (PH-BC-WSN) was utilized to increase access to high-quality healthcare, increase food production through precision agriculture, and raise the standard of human resources. More effective healthcare and medical asset monitoring systems can be developed thanks to the Internet of Things. Eavesdropping on medical data, modification, fabrication of warnings, denial of service, user tracking and location, physical interference with equipment, and electromagnetic threats were extensively discussed. The article gives several instances of current technology, discusses design challenges including energy efficiency, security, and scalability, provides various demonstrations of current technology, and provides a complete analysis of all the advantages and disadvantages. Despite their many benefits, body sensor networks have several significant obstacles and unresolved research problems, which are described along with some potential answers. As a result, the experimental findings demonstrate that PH-BC-WSN enhances accuracy and reduces response time in inpatient health monitoring.