Industry 4.0 is a trending revolution in the present society, especially in the circular economy system. The latest advances in the domain of data science, artificial intelligence (AI), robotics, the Internet of Things (IoT), big data, augmented reality, and virtual reality have opened the door for smart waste treatment systems. In India, municipal waste generation has increased dramatically during the past several decades, partly due to the country's rapid population expansion. The amount of waste generated in cities across India has risen from 6 million tons in 1947 to 48 million tons by 1997, and it is anticipated to reach 300 million tons per year by 2047. As a result, several substantial efforts must be taken to resolve the waste management problem. The current system arbitrarily collects waste. As a result, some places are occasionally left unattended, resulting in an unpleasant odor and a threat to public health, as the smell of garbage can potentially be harmful to little children and old-aged people. This chapter focuses on overcoming the hazards by adopting a smart waste management system that includes a cyber—physical system (CPS), which makes use of computers 164to monitor the waste management process, and the Internet of Everything (IoE), which is a type of intelligent computing that connects things, people, services, data, etc., by making use of machine-to-machine communication protocols. These approaches make the waste treatment mechanism smart and could easily enable the waste-the-energy approach, converting energy into electricity, heat, or fuel.

In the digital era, the rise of fraudulent users during exams and online courses presents a significant challenge to the integrity of educational assessments. However, with advancements in technology, there exists a promising solution" Fake User Identification and Duplication Detection using MERN stack ". This innovative system integrates a variety of cutting-edge algorithms, including node.js for backend processing, React.js for frontend interface development, and Convolutional Neural Networks (CNNs) for sophisticated image analysis. By combining these powerful tools, our method aims to establish a robust user-controlling system capable of thwarting fraudulent activity during exams and classroom interactions. By implementing Fake User Identification and Duplication Detection, educational institutions can uphold the integrity of their digital ecosystems while preserving the authenticity of assessments. This technique not only deters fraudulent behavior but also fosters a culture of academic honesty and accountability in online learning environments. As technology continues to advance, our ongoing commitment to innovation ensures that our system remains at the forefront of defending against fraudulent users, safeguarding the future of online education.

One characteristic of type 1 diabetes mellitus (T1DM), a chronic autoimmune disease, is the body's incapacity to produce insulin. Blood glucose levels must be constantly monitored and controlled by people with kind 1 diabetes with the goal to prevent complications. Machine learning (ML) techniques offer a chance to help with personalized diabetes treatment by predicting blood glucose levels. The project's purpose is to build a Web app that employs machine learning methods to forecast blood glucose levels in persons with type 1 diabetics. This research study aims to collect real-time data from T1DM patients on relevant parameters such as exercise, insulin dosage, and consumption of carbohydrates. The information can be utilized for teaching machine learning methods, including cooperative gets closer, neural systems, and model-based regression, to estimate upcoming blood glucose levels. Individuals with type 1 diabetes can use the Flask programs straightforward interface to enter their routines, insulin dosages, and meal information. Subsequently, the software will employ the acquired machine learning methods to generate customized predictions for blood sugar levels. It will additionally involve studies on the factors influencing blood sugar fluctuations and recommendations for bettering diabetes treatment. The project is significant as it offers accurate and personalized blood sugar predictions, which might enhance the standard living for those with type 1 diabetes. By applying machine learning techniques and developing a simple implementation, the project aims to empower T1DM patients to take educated choices regarding their healthcare, thereby improving their general wellness and health.

Nowadays, looking into the world’s environmental crisis in real-time is essential. Forecasting climatic conditions is critical to human survival; it includes collecting data on the temporal dynamics of weather changes. Weather monitoring often includes monitoring emissions of major pollutants such as hydrocarbons, which are organic compounds composed of hydrogen and carbon. Prolonged exposure to hydrocarbons and other gases results in health problems such as cancer and asthma. This project aims to create an effective solution for real- time weather forecasts and air pollution control by leveraging advancements in embedded system design to create a weather monitoring system. With IoT, tracking weather parameters in a location through DHT11, Ultrasonic sensor, MQ6, MQ2, and MQ135 gas sensors is possible. Dirigible satellite monitors weather changes like temperature, humidity, and air pollution, which measures LPG gas, smoke, benzene, and other combustible gases. The data uploaded to the web page is easily accessible from anywhere in the world. This project contributes to developing a comprehensive solution for weather monitoring and air pollution control by integrating advanced embedded systems and IoT technologies.

In medical diagnosis, early and accurate detection of brain tumors is crucial for effective treatment planning and improved patient outcomes. Traditional methods for brain tumor detection in magnetic resonance imaging (MRI) scans and computed tomography (CT) scans can be time-consuming, subjective, and prone to human error. In this paper we overcome the problem in the domain of computer vision with deep learning approach. Deep learning models have emerged as a promising alternative, offering high accuracy and the potential for automation. This work investigates the application of the YOLOv8 deep learning model for real-time brain tumor detection in MRI scans. YOLOv8 offers a compelling balance between speed and accuracy, making it suitable for deployment in clinical settings. We evaluate the performance of YOLOv8 on a benchmark brain tumor dataset and compare it with other commonly used models. The results demonstrate that our YOLOv8 model achieves competitive accuracy (mentioning the specific metrics like accuracy, precision, recall) while maintaining real-time inference speed, paving the way for its potential use in real-time brain tumor detection during clinical examinations.

Today, the growth of digitalization has made the ease for livelihood for all the organizations. Cloud computing the storage provider for all the computer resources has made it easy for accessing the data from anywhere anytime. But at the same time the security for cloud data storage is the major drawback which is provided by various cryptographic algorithms. These algorithms convert the data into unreadable format, known as cipher text, Rivest, Shamir and Adleman (RSA) one of the most popularly used asymmetric algorithm. This paper gives detailed review about such different cryptographic algorithms used for the cloud data security. The comparison study is also made for the size of data and to analyze the encryption time and decryption time, which concludes that to enhance the cloud data security some addon techniques are to be used along with these cryptographic algorithms. To increase the security level and to increase the transmission speed of plaintext, integrated method will be proposed by encoding the plaintext to intermediate plaintext and then intermediate plaintext will be compressed using any one of the compression techniques to increase the compression ratio, lastly the compressed file is encrypted to further enhance the security level.

Vehicles are regarded as the primary form of transportation in many nations. People prefer to go by vehicle since it gets them where they're going faster. The majority of driving mistakes on the road result in traffic accidents. Drivers' awareness may be affected by a variety of factors, including visual complexity, the surroundings, and inadequate driving instruction. The traffic signs were kept to alert the driver about the condition of the road, speed limit, speed breakers, restricted zones like schools and hospitals, divisions, turnings and intimate them to control the speed of the vehicle accordingly. For autonomous vehicles, this information is much more essential since there is no manual assistance. The traffic signs should be identified by the autonomous vehicle and necessary control action should be taken in order to avoid accidents and violations of traffic rules. The proposed system suggested a model that can read the traffic sign boards and convert them into instructions for autonomous vehicle by using image processing, object identification and recognition techniques. A deep learning model is used for classification of the captured image.

Accidents at bore wells are now frequent worldwide. Some of the kids are rescued, but most of the time, we lose to save their lives. The main goal of this project is to develop and build an octobot for saving a child from an open borewell. There are several methods available to rescue the child. However, using just one strategy might not be feasible. But, these methods/techniques are time-consuming and may not be successful in most cases. To address the problem, we proposed an octobot capable of taking out the child from the bore well methodically. The proposed system comprises a camera, lighting, oxygen source, and balloon technology. Sensors and the camera monitor the child’s depth, position, and state. The child is kept from slipping back into deep with its assistance. And the child is raised on a balloon-filled pillow. The octobot is controlled manually by a human, who also uses a computer or a mobile application to monitor it. According to ongoing assessments, the child is being saved securely, and the necessary medical arrangements may be made to save the child’s life.

Writing abilities are impacted by dysgraphia, a condition of learning disability. It might be challenging to diagnose dysgraphia at an initial point of a child's upbringing. Problematic abilities linked to Dysgraphia difficulties that is utilized in detecting the learning disorder. The features used in this research to identify dysgraphia include handwriting and geometric features that is reclaimed using kekre-discrete cosine mathematical model. The feature learning step of deep transfer learning makes good use of the obtained features to identify dysgraphia. The results of the data collection indicate that this study can use handwritten images to detect children who have dysgraphia. Compared to past investigations, this experiment has shown a significant improvement in the capacity to identify dysgraphia using handwritten drawings. The proposed approach is compared with the machine learning and deep learning approaches where the Kekre-Discrete Cosine Transform with Deep Transfer Learning (K-DCT-DTL) outperforms the existing approaches. The proposed K-DCT-DTL approach attains 99.75% of highest accuracy that exhibits the efficiency of the proposed method.

In wireless sensor network (WSN) application, one of the most important problems to be considered is power consumption. As a result, a number of methodologies or methods for analyzing the power consumption of the application have been presented. These methodologies may aid in the prediction of the lifespan of WSNs, the providing of suggestions to application developers, and the optimization of the amount of energy utilized by WSN applications. All of the sensors in wireless sensor nodes are rechargeable batteries devices with a limited amount of battery power available. It will be difficult to replace every power supply battery available in the network on an individual basis after deploying of sensor nodes. To sort out these issues, an efficient analyzing technique for analyzing the power consumption of WSN is proposed. In this research work, a prototype model with a simulation model is presented. The proposed research work may lower the amount of power used by each sensor node, while also increasing the life of sensors by shutting down certain of their components in order to achieve higher energy savings and longer life. It is necessary to do an analysis of the power usage of each sensor node before using this power management strategy. As a result of this, the proposed method analyzes the WSN power consumption in real time manner.

A credit card is used to permit the customer to make settlements of huge amount of money without carrying a lot of money which is need by the customers. It has changed the way reforming of making credits only in installments and made making such an installment helpful for the purchaser. One of the biggest problems faced in the banking sector is predicting a credit card client as a solution this paper uses Supervised Learning algorithm that must be applied for credit card approval for solving the problem. Additionally, the accuracy of the outcome is compared to several known standards using Supervised Learning techniques. The result shows that the ideal accuracy for KNN, Decision Tree, Naive Bayes, Logistic Regression are 98.50%, 98.13%, 99.62% and 84.32% respectively. The relative result exhibits that Decision Tree performs superior to other algorithms as in case of credit card assent. A website is created using FLASK python libraries, to connect the bank with the new entry customer. The proposed idea will help to avoid the risk in identifying and approving credit cards for the eligible customers.

Healthcare is one of the largest sector in India, telemedicine services is a component of healthcare that is evolving in recent days. This has drawn considerable interest to develop a smart and credible health care monitoring system that can be utilized by the front end health care professionals to monitor a patient’s health. This lead to the employment of Internet of Things (IOT) in booming health care sector wherein remote monitoring of a patient’s health can be facilitated. Hence we develop an IOT dependent health observing system that uses ATmega 328 microcontroller, heartbeat, temperature and accelerometer sensor along with GSM module. The sensor’s data is gathered and is then transferred to a could based server (thingspeak), Healthcare professionals can retrieve the patient’s information from the server via an application. Further more our proposed model also uses Global System for Mobile (GSM) module to convey patient’s information to parent’s and doctors in case of any emergency. Hence this system improves the operating efficiency, lessons the time and cost of support.

On streets and highways, traffic signs have a specific set of characteristics that may be used to distinguish one from another. A real-time traffic sign recognition and identification process based on neural networks. A quicker R-CNN technique is utilised to recognise traffic signals, and a convolutional neural network is employed to categorise them using the Inception V2 architecture. Some elements, such as light, occlusion, blurring, and others, might make it difficult. This proposed work could be used in a variety of fields, including the Advanced Driving Assistant System and self-driving automobiles. The classification network is used to execute particular classification tasks before computing the bounding box regression. The tests are carried out using a standard benchmark dataset, and the results indicate that the approach is both successful and resilient.

Recent decade has proved an increase in learning disabilities and directly impacts the education of children worldwide. This problem has to be addressed with a solution since they integrate into society for a bright future. Abnormalities in brain has an impact on memory, intellectual thinking and its balance is getting affected. Because of which achievements in academics and learning ability among children is observed to be lagging. Conventional methodologies involved in detection and interventions in Specific Learning Disability affects the controlling functions creating a barrier in education. The views provided enhances the development of cognitive skills and progress in academics. The tools provided in this work has observed academic progress of an individual. This survey focuses on various learning disabilities and their impact on education. The review findings focus on the tools and various solutions provided in the recent past, supporting the children suffering from learning disabilities. The studies have provided effective guidance for both affected students and teachers who contribute a major role in their progress. This survey provides an exploratory approach to discuss detection and diagnosis tools for three disabilities dyslexia, dysgraphia, and dyscalculia with its various impacts and possible solutions. It provides a broad view for future researchers to provide a better solution compared to existing ones.

For the purpose of generation of pure sinewave AC output from the battery source of DC, inverter circuits like class D, class E, etc., were used. This class type inverters generates the PWM signals in sinusoidal forms from the sinewave input frequency, which is fed to the BJT driver of H bridge that generates pure sinewave AC output as discussed. This paper deals with the modelling and the simulation of class D and class DE for induction heating application. To generate the heat in the induction cooker, class DE circuit is connected parallel to the parallel inverter amplifier. This class DE amplifier system consists of reactive elements and four switches in it, so that the class DE amplifier system generates currents at higher frequencies to meet the load demand. In this paper, to bring out the efficient system for the purpose of induction heating, two types of inverter circuits, class D and class DE amplifiers were taken, modelled, simulated and compared. The open loop system of the class D and class DE amplifier system were simulated and the comparative study has been made to show out the efficient system for the induction heating application.

Encryption is essential for protecting sensitive data, especially images, against unauthorized access and exploitation. The goal of this work is to develop a more secure image encryption technique for image-based communication. The approach uses particle swarm optimization, chaotic map and magic square to offer an ideal encryption effect. This work introduces a novel encryption algorithm based on magic square. The image is first broken down into single-byte blocks, which are then replaced with the value of the magic square. The encrypted images are then utilized as particles and a starting assembly for the PSO optimization process. The correlation coefficient applied to neighboring pixels is used to define the ideal encrypted image as a fitness function. The results of the experiments reveal that the proposed approach can effectively encrypt images with various secret keys and has a decent encryption effect. As a result of the proposed work improves the public key method's security while simultaneously increasing memory economy.