The growing frequency of fissures in concrete construction presents considerable hazards to the stability and security of the structure. Because traditional inspection techniques are frequently labor-intensive and prone to human error, automated solutions must be investigated. This article proposes a convolutional neural network method for the identification of the presence of cracks on concrete surfaces. The aim is to overcome the shortcomings of manually inspecting the concrete structure, which is time-consuming and prone to human error. The study shall provide a more efficient method with better accuracy for structure monitoring purposes. A CNN model with remarkable accuracy of 93.75% was trained using a comprehensive concrete image dataset in this study. Results show that the proposed automated approach enhances not only structural safety but also cost savings in terms of maintenance and monitoring efficiency. Conclusions from this work point out that integration of machine learning techniques like CNN into crack detection systems could be a huge progress toward development in the field of Structural Health Monitoring.

Safe blood transfusion and emergency medical care depend on a fast and precise grouping of blood. The serological ways are slow though they are the current norm, subjective, and need a skilled worker. Here we suggest a mix model of deep learning, Transformer-CNN, meant for smart and strong blood type sorting via pictures of blood smears. The model mixes the wide attention ability of Transformers with the power of Convolutional Neural Networks for making better classification and handling changes in how the pics are taken. A multi-stage pipeline is developed, including image preprocessing, data augmentation, feature extraction, and hybrid model integration. Blood smear datasets are augmented with rotation, color space conversion, and histogram equalization. The CNN backbone uses ResNet50 and MobileNetV2 while the Transformer layers enable spatial context modeling. It is trained and evaluated on over 70,000 labeled images across eight blood group classes using stratified and cross-domain dataset splitting. Results from the experiments proved that the hybrid model achieved a classification accuracy of 96.57% hence, and performed the standalone architectures ResNet18, InceptionV3, and DenseNet201 in both precision and efficiency of training. Moreover, it also proved to be robust under out-of-distribution testing as well as image degradation scenarios. The evaluation metrics-accuracy, precision, recall, F1-score, and a confusion matrix analysis validate its effectiveness. The Transformer-CNN model suggested provides a scalable reliable interpretable solution for clinical blood group classification with potential application in emergency diagnostics mobile health devices AI-assisted pathology.

Technological developments have come a long way in developing service robots and employing them to serve human needs to the fullest extent. Communication between the robot and its user is the most important aspect, and developing a system to train robots and provide them with the necessary capabilities to accomplish specific tasks is important for creating a suitable service robot. This work provides solutions to most of the challenges in developing such systems using smart systems that can implement multiple and complex commands in a short and appropriate time, such as matching data outputs with what is examined by these systems. For this reason, (RFID) was used to compare data and give commands so that the robot can move in complex environments. Another challenge to consider is keeping the robot on the specified path without deviating from its specified path, so a special sensor was developed to track the robot’s deviation and correct the path. Another challenge to consider is if the robot might collide with an object on the path, this problem has been solved by a wireless sensor, all of which is processed and controlled by the microprocessor and motor control unit. Several practical tasks were implemented for the proposed model, and these tasks were in complex and different environments and under difficult conditions. The proposed service robot model was able to implement them with very high accuracy, so it can be relied upon as an advanced service robot model.

Recent years have witnessed a wide spread of scientific applications related to robot science, especially in the field of interaction between real humans and robots. The service robot application is considered one of the best daily life applications It was extensively spread. in which communication between the robot and its user the biggest issue. In this work, a robot is designed and implemented that operates on the principle of movement by using voice commands to determine its path in movement. Track programming technology was also used to control the movement of this robot. The design used to implement these commands is a robot with four wheels that are moved using DC motors. The mathematical model for assembling the robot was built, as well as the mathematical model for controlling the movement of the robot. The commands were programmed, loaded on the microcontroller, and examined through the Proteus program. The most significant of the many practical and simulated outcomes was the ability to voice-command the movement of the design, program the robot’s path as needed, and avoid any obstructions that might be in the way. The obtained results are characterized by high accuracy, which confirms the possibility of adopting the proposed design as an applied practical model. In the final mode, the Internet of Things (IoT) technology implement to control this design remotely, satisfactory results were obtained as implemented practical mobile service robot with intelligent controller and it has ability to execute many services jobs for human life.

The Internet of Things (IoT) has made a significant impact on improving human life in the modern era, offering convenience and ease. Smart appliances, which are connected to the Internet, have become increasingly popular due to their improved efficiency, robustness, power consumption, and connectivity. The main objective of this paper is to convert a conventional refrigerator into an intelligent appliance by integrating computational intelligence using a microcontroller and a range of sensors. The proposed approach aims to improve food management capabilities by introducing automated techniques and cost-effective intelligence to the refrigerator. The resulting smart refrigerator is designed to be user-friendly and improve the overall quality of life. The system includes a mobile or web application that acts as a bridge between the refrigerator and the user, allowing for remote control and monitoring. The smart refrigerator employs sensors to detect shortages of items and can automatically place orders online or send messages to grocery stores for item delivery. Additionally, the smart module continuously monitors temperature and humidity levels, uploading real-time updates to a cloud platform for user access. The integration of IoT technology into refrigerators offers numerous features and benefits, ultimately enhancing convenience and efficiency in managing food storage and consumption.