International Journal of Engineering and Manufacturing @ijem
Journal articles - International Journal of Engineering and Manufacturing
All articles: 649
ACO-QL: Enhancing ACO Algorithm for Routing in MANETs Using Reinforcement Learning
Scientific article
ACO-based routing protocols like AntHocNet have emerged as a solution for adaptive routing in MANETS. Likewise, deep Q-learning based protocols are suitable for complex and dynamic environments like MANETs and utilizing real time data for better decision-making. However, there is lack of studies in enhancing ACO-based protocols using Q-learning in a new hybrid protocol, and comparing it with the established ACO-based protocol AntHocNet. By combining ACO’s strengths (eg. Multi-agent pathfinding and historical data creatd by pheromones) and combine it with key components of Q-learning, then we have a promising protocol ready to be compared with AntHocNet. Previous studies have explored integrating ML with MANET routing, but few of them, if any, have explored enhancing ACO using ML techniques. Therefore, we propose two new protocols: ACO-QL and ACO-DQN. One uses Q-learning and the latter uses deep Q-learning. After conducting many experiments by running implementations of ACO-DQN, ACO-QL, and AntHocNet on a MANET simulation, we found out that AntHocNet is superior to ACO-DQN in terms of execution time, end-to-end delay, and path cost in most cases, but on the other hand ACO-DQN achieved better packet delivery ratio and throughput results. Meanwhile, ACO-QL consistently achieved lower packet delivery ratios than AntHocNet, and mostly matching AntHocNet’s performance in terms and of other metrics, making it a valid lightweight and faster alternative.
Free
AI-Based Smart Prediction of Liquid Flow System Using Machine Learning Approach
Scientific article
Predicting the liquid flow rate in the process industry has proved to be a critical problem to solve. To develop a mathematical, in-depth of physics-based prognostics understanding is often required. However, in a complex process control system, sometimes proper knowledge of system behaviour is unavailable, in such cases, the complement model-based prognostics transform into a smart process control system with the help of Artificial Intelligence. In previous research a number of prognostic methods, based on classical intelligence techniques, such as artificial neural networks (ANNs), Fuzzy logic controller, Adaptive Fuzzy inference system (ANFIS) etc., utilized in a liquid flow process model to predict the effectiveness. Due to system complexity, Computational time &over fitting the performance of the AI has been limited. In this work we proposed three machine learning regression model: Random Forest (RF), decision Tree (DT) & linear Regression (LR) to predict the flow rate of a process control system. The effectiveness of the model is evaluated in terms of training time, RMSE, MAE & accuracy. Overall, this study suggested that the Decision Tree outperformed than other two models RF & LR by achieving the maximum accuracy, least RMSE & Computational time is 98.6%, 0.0859 & 0.115 Seconds respectively.
Free
Absorption, Diffraction and Free Space Path Losses Modeling for the Terahertz Band
Scientific article
With the explosive increase in the data traffic of wireless communication systems and the scarcity of spectrum, terahertz (THz) frequency band is predicted as a hopeful contender to shore up ultra- broadband for the forthcoming beyond fifth generation (5G) communication system. THz frequency band is a bridge between millimeter wave (mmWave) and optical frequency bands. The contribution of this paper is to carry out an in-depth study of the THz channel impairments using mathematical models to evaluate the requirements for designing indoor THz communication systems at 300GHz. Atmospheric absorption loss, diffraction loss and free space path loss were investigated and modeled. Finally, we discuss several potential application scenarios of THz and the essential technical challenges that will be encountered in the future THz communications. Finally, the article finds that propagating in the THz spectrum is strongly dependent on antenna gain.
Free
Achieving Performability and Reliability of Data Storage in the Internet of Things
Scientific article
Internet of things (IoT) includes a lot of key technologies; In this emerging field, wireless sensors have a key role to play in sensing and collecting measures on the surrounding environment. In the deployment of large-scale observation systems in remote areas, when there is not a permanent connection with the Internet, the network requires distributed storage techniques for increasing the amount of data storage which decreases the probability of data loss. Unlike conventional networked data storage, distributed storage is constrained by the limited resources of the sensors. In this research, we present a distributed data storage method with the combined K-means and PSO clustering mechanism organized with the binary decision tree C4.5 in the IoT area with considering efficiency and reliability approach. This scheme can provide reliability in responding to inquiries while minimizing the use of energy and computational resources. Simulation results and evaluations show that the proposed approach, due to the distributed data storage with minimal repeat publishing according to the decision tree structure, increases the reliability and availability, reduces the communication costs, and improves the Energy consumption, saving memory consumption without registering the same event and compared to other methods performed in this area have good results.
Free
Scientific article
EMG-based prosthetic control systems are known to suffer from muscle fatigue, electrode displacement, signal drift, and inter-subject variability issues, compromising their robustness and long-term stability. While recently developed deep learning techniques deliver highly competitive gesture recognition accuracy, most existing solutions lack adaptive learning and privacy-preserving model optimization mechanisms required for the real-world implementation. In this paper, we propose a new EMG-based adaptive prosthetic control system combining the Hybrid CNN, ViT, and LSTM architectures and applying federated reinforcement learning (FRL). The CNN component of the proposed architecture extracts local spatial information from multi-channel EMG signals, the Vision Transformer analyzes inter-channel dependencies using a self-attention mechanism, while the LSTM models the temporal dynamics of muscle activation. Additionally, the reinforcement learning algorithm constantly updates control policies according to user feedback, while federated learning allows users to collaborate on model optimization without exposing raw EMG data, thus protecting user privacy. In order to enhance the system robustness in real-world conditions, the Adaptive Muscle Signal Learning (AMSL) technique is used to counteract the adverse effect of signal drift, muscle fatigue, and electrode displacement. An EMG dataset specification, called EMGPro-2026, is introduced to establish the criteria of data acquisition, gesture recognition classes, preprocessing, and evaluation process. Based on computational assessment and simulation analysis, there is proof of concept for the proposed framework that enables high gesture classification accuracy, adaptation for personalization, and efficient learning with privacy preservation for real-time use in prosthetics. The proposed CNN-ViT-LSTM with federated reinforcement learning presents an intelligent solution for the future generation of myoelectric prosthetic control systems. Our future work will entail the validation of our framework using EMG signals recorded ethically. Because the current research provides a conceptual framework, there has been no experimental validation of the performance using human EMG datasets. However, based on the benchmarks developed from the existing state-of-the-art literature, it can be said that the suggested framework for Hybrid CNN–ViT–LSTM with Federated Reinforcement Learning is predicted to deliver approximately 94.5% gesture classification accuracy, which will be better than the conventional techniques such as CNN, CNN–LSTM, and SVM and will also increase adaptability, personalization, and privacy preservation.
Free
Scientific article
The feasibility of utilizing an abundant agricultural waste (desert date seed shell) as an alternative low-cost adsorbent for the removal of hazardous basic dyes [crystal violet (CV) and malachite green (MG)] and hexavalent chromium [Cr(VI)] from synthetic industrial effluent was investigated. Five different adsorbents including the raw, carbonized and chemically activated carbons were prepared and screened with respect to adsorption efficiency of the chosen adsorbates. The prepared adsorbents were characterized using Fourier transform infrared (FTIR) spectroscopy, scanning electron microscopy (SEM) and pH of zero point charge (pHzpc) analyses. The effects of operational variables such as solution pH, contact time and temperature on adsorption have been investigated. The removal of the adsorbates was found to be highly pH-dependent and the optimum pH was determined as 8.0 for the dyes and 2.0 for hexavalent chromium. The screening results revealed that the NaOH activated carbon (NAC) has the best adsorption characteristics with removal efficiencies of 91.10, 99.15 and 91.5 % for CV, MG and Cr(VI), respectively. The process dynamics was evaluated by pseudo-first-order and pseudo-second-order kinetic models. Experimental data have been found to be well in line with the pseudo-second-order model, suggesting therefore, a chemically-based sorption process. Negative Gibbs free energy change (∆G) values obtained from thermodynamic analysis indicated that the adsorption process was spontaneous and had a high feasibility. Positive values for enthalpy change (∆H) showed that the removal process was endothermic, implying that the amount of adsorbate adsorbed increased with increasing reaction temperatures. Additionally, positive values of entropy change (∆S) reflect the high affinity of the adsorbent material to the adsorbates. On the basis of results and their analyses, it has been established that adsorbent derived from desert date seed shell has a promising potential in environmental applications such as removing hazardous substances from industrial effluents. Through this work, it is believed that contributions are provided to the scientific investigations about the decontamination of precious water resources.
Free
Advances in MEMS Technology: An In-Depth Analysis of Evolution, Applications, and Future Directions
Other
Micro-Electro-Mechanical Systems (MEMS) have fundamentally transformed technology by combining microelectronics with mechanical systems to create miniature devices capable of diverse functionalities. This review article provides a thorough exploration of MEMS, tracing its evolution from early developments to the latest advancements. It begins by outlining the fundamental principles behind MEMS design and fabrication, detailing processes such as lithography, deposition, and etching. The paper covers a wide array of MEMS devices, including sensors, actuators, resonators, and microfluidic systems, while focusing on essential design considerations, fabrication techniques, and performance parameters. The versatility of MEMS across sectors like healthcare, automotive, aerospace, consumer electronics, and telecommunications is highlighted, illustrating their role in advancing applications such as medical diagnostics, environmental sensing, and autonomous technologies. Unlike previous reviews, this paper provides a unique synthesis linking fabrication mechanisms with device performance metrics, offering an updated comparative analysis across MEMS subcategories (RF MEMS, microfluidics, and optical MEMS). It also integrates the latest market data (2024–2025) and contextualizes how MEMS devices underpin IoT and Industry 5.0 applications. Furthermore, it emphasizes emerging research directions such as energy harvesting MEMS, bio-inspired microsystems, and security-aware MEMS integration in connected environments. These additions make this review both comprehensive and forward-looking, serving as a reference for researchers and practitioners.
Free
Scientific article
Bangla Sign Language is a unique sign language. Due to a lack of interpreters, the hearing- and speech- impaired community face challenges while communicating with the broader community. Recent studies have been con- ducted to reduce the gap between these two communities. But most of the researchers used a dataset with a controlled environment. We know the performance of a system highly depends on dataset quality. In this paper, we have created a new dataset, “BanglaSignSet” including 46 unique signs with over 10k images. We have carefully annotated and labeled the images using Roboflow. Our proposed dataset, “BanglaSignSet” consists of images with high resolution, good qual- ity, and adequate variation in environment and person. The constructed dataset has been trained using the most recent deep learning model, such as YOLOv8. We have also implemented different versions of the YOLOv8 model, such as YOLOv8n, YOLOv8s, and YOLOv8m. Additionally, we evaluated EfficientNet-B0 as a classification-based baseline to broaden the experimental comparison. The performance of models has been measured using different evaluation metrics such as mAP, precision, recall, and f1 score. A comparative analysis has been conducted based on the performance of the model. By comparative analysis we found a well-suited model, YOLOv8n, to deploy into a web-based application. To find the suitable model to deploy, we have considered factors such as memory requirement and inference speed. We have integrated the YOLOv8n model into a web application using the Python language. We have also tested the web application on Android devices and laptops. The web application detects signs from image input successfully.
Free
Scientific article
A vital component of patient care is the diagnosis of blood cancer, which necessitates prompt and correct classification for efficient treatment planning. The limitations of subjectivity and different levels of skill in manual classification methods highlight the need for automated systems. This study improves blood cancer cell identification and categorization by utilizing deep learning, a subset of artificial intelligence. Our technique uses bespoke U-Net, MobileNet V2, and VGG-16, powerful neural networks to address problems with manual classification. For the purposes biomedical image segmentation U-Net architecture is used, MobileNet V2 is used for lightweight neural network model design and VGG-16 is used for image classification. A hand-picked dataset from Taleqani Hospital in Iran is used for the rigorous training, validation, and testing of the suggested models. The dataset is refined using denoising, augmentation, and linear normalisation, which improves model adaptability. The results show that the MobileNet V2 model outperforms related studies in terms of accuracy (97.42%) when it comes to identifying and categorizing blast cells from acute lymphoblastic leukemia. This work offers a fresh approach that adds to artificial intelligence's potentially revolutionary potential in medical diagnosis.
Free
Advancing Road Scene Semantic Segmentation with UNet-EfficientNetb7
Scientific article
Semantic segmentation is an essential tool for autonomous vehicles to comprehend their surroundings. Due to the need for both effectiveness and efficiency, semantic segmentation for autonomous driving is a difficult task. Present-day models’ appealing performances typically come at the cost of extensive computations, which are unacceptable for self-driving vehicles. Deep learning has recently demonstrated significant performance improvements in terms of accuracy. Hence, this work compares U-Net architectures such as UNet-VGG19, UNet-ResNet101, and UNet-EfficientNetb7, combining the effectiveness of compound-scaled VGG19, ResNet101, and EfficientNetb7 as the encoders for feature extraction. And, U-Net decoder is used for regenerating the fine-grained segmentation map. Combining both low-level spatial information and high-level feature information allows for precise segmentation. Our research involves extensive experimentation on diverse datasets, including the CamVid (Cambridge-driving Labeled Video Database) and Cityscapes (a comprehensive road scene understanding dataset). By implementing the UNet-EfficientNetb7 architecture, we achieved notable mean Intersection over Union (mIoU) values of 0.8128 and 0.8659 for the CamVid and Cityscapes datasets, respectively. These results outshine alternative contemporary techniques, underscoring the superior precision and effectiveness of the UNet-EfficientNetb7 model. This study contributes to the field by addressing the crucial challenge of efficient yet accurate semantic segmentation for autonomous driving, offering insights into a model that effectively balances performance and computational demands.
Free
Scientific article
The increasing urbanization of many third-world cities has led to increased generation of solid wastes which are often ill-managed and indiscriminately dumped, posing grave challenges to local environmental engineers and designers. This has consequently reduced the sustainability of many built and natural African environments. Therefore, this study was aimed at practically applying aesthetics in solid waste management as a means of optimizing sustainability in urbanizing West African environments. Adopting a descriptive approach supported with direct observation, with a total sample size of 610, respondents were purposively sampled in selected research sites in Nigeria. Following one hypothesis testing, the study showed a significant association between improved environmental affordance (derived from aesthetics) and the alleviation of negative user responses to the environment (such as indiscriminate dumping of solid wastes). The study also showed that more aesthetically negative environments offers more negative environmental affordance than positive environmental affordance. The results confirm that the majority of users of the environments (humans) exhibit more positive environmental behaviours when positive affordance is perceived from the environment. The study therefore established the significance of the practical application of aesthetics in the management of solid wastes in urbanizing third world environments.
Free
Alleviating Unwanted Recommendations Issues in Collaborative Filtering Based Recommender Systems
Scientific article
The overabundance of information on the internet and ecommerce has resulted to the development of recommender system to discover interesting items or contents that are recommendable to the user. The recommended items might be of no interest or unwanted to the users and can make users to lose interest in the recommendations. In this work, a Collaborative Filtering (CF) based method which exploits the initial top-N recommendation lists of an item-based CF algorithm based on unwanted recommendations penalisation is presented. The method utilises a relevance feedback mechanism to solicit for users preferences on the recommendations while popularise similarity function minimises the chances of recommending unwanted items. The work explains the proposed algorithm in detail and demonstrates the improvements required on existing CF to provide some adjustments required to improve subsequent recommendations to users.
Free
Alternative device for non-ionizing radiation detection
Scientific article
Detection of non-ionizing electromagnetic radiation is central to managing health and environmental hazards resulting from its exposure. This research focused on the design and development of a non-ionizing electromagnetic radiation detector that is sensitive to the medium frequency of 50Hz to 30MHz and their corresponding power density. The device consists of the sensing, amplifying, filtering and microcontroller sections. The sensing section is made up of a coil wound on a ferrite rod, it detects radiations from the surroundings and converts it to a voltage signal. The voltage produced is then fed to the operational amplifiers in the amplifying section. Afterwards, the output signal is fed to the filtering section where unwanted signals are eliminated. The analogue signal output from the active filter is then fed to the microcontroller section where it is converted to a digital signal through the analogue to digital converter (ADC). The ADC then presents the converted signal in a readable form to be displayed on the liquid crystal display (LCD). The developed equipment was calibrated (in µW/cm2) using an existing detector EMF DT 1130. With an average calibration coefficient value of 2.32, the detector was found to perform excellently well at both medium and low-frequency ranges.
Free
An Adaptive User Authentication Architecture for Drunk Driving and Vehicle Theft Mitigation
Scientific article
The high rate of vehicle theft and the loss of lives occasioned by drunk driving has caused irreparable losses to people and businesses, from a personal, commercial and reputation perspective. Existing systems deployed to mitigate against vehicle theft have all been breached by the ever-adaptive criminals. Drunk driving has been estimated to be a leading cause of deaths on highways and motorways, through preventable accidents. Technology has provided the tools that can aid in mitigating the vices aforementioned with the aim of provisioning lasting solutions. This paper proposes a new architecture for adaptive user authentication in order to mitigate drunk driving and vehicle theft. It considered user authentication in three (3) phases and proposed an authentication architecture for each identified phase, with a step by step description of the implementation method and tools for each phase. The architecture proposed in this study can aid in real time prevention of vehicular theft, unauthorized vehicular access and usage, while also utilizing the benefits of the latest technologies in machine vision and alcohol breadth analyzers to detect and prevent drunk driving, and the associated accidents it causes.
Free
Scientific article
In this paper we propose a low-rank alternating direction implicit (ADI) method to solve projected generalized continuous-time Sylvester equations with low-rank right-hand sides. Such equations arise in control theory including the computation of inner products and norms, and the model reduction based on balanced truncation for descriptor systems. The requirements of this method are moderate with respect to both computational cost and memory. Numerical experiments presented in this paper show the effectiveness of the proposed method.
Free
An Analysis on Price Dispersion in Online Retail Market Based on the Different of the Product Levels
Scientific article
This paper cares about the online price dispersion and analyses the online retail commodity as a complex. This paper analyses the attributes of the product layer of the online retail commodity and the characters of the online customers, and divided the online payment into two parts: the payment of the core value and the willingness of the added-value payment. And then this paper explains the online price dispersion by the attributes of online retail commodity’s dimension which affect the online customers directly.
Free
An Application of Rule-Based Classification with Fuzzy Logic to Image Subtraction
Scientific article
Surveillance camera is used as a new technology for security. In this research, the combination of OpenCV with image processing will be discussed. Saving the space in the hard drive by recording only video when here would be an image formed in the subtraction of the original image to the next image captured. With the use of Image Processing and Fuzzy logic, the research was enhanced by eliminating the recording of same image captured. After analyzing the background images, it can now determine when to start recording the video or when to stop recording a video by subtracting the images in the backdrop image and comparing the image if there was an object in motion using template matching. With the application of the project, memory storage saved up to forty-six percentage points.
Free
Scientific article
Researches on the relation of a loaded mechanical energy to the damaged banana maturation and its quality evolution has significance to its storage, but there are few researches at present. This paper studied on the effect of the loaded energy on banana to its internal maturation substance evolution and stored quality with time, including the change of maximum impact stress, re-enduring impact energy. The results showed there existed a significant correlation between energy transformation and increment of maturity substance in the damaged banana. Also, there existed a significant association between energy transformation and increment of maturity substance in the damaged banana in its storage time, and appeared a trends that the mechanical energy being transferred into fruit internal energy increased its maturity substance and decreased its stored quality, but individual maturity difference had less effect. Thus, decreasing the energy and the substance transformation in damage area are the important means of extending the physiology life of fruits, which is an urgent problem to be solved in their field of processing and storage at present.
Free
An Automated Model for Sentimental Analysis Using Long Short-Term Memory-based Deep Learning Model
Scientific article
A post, review, or news article's emotional tone can be automatically ascertained using sentiment analysis, a natural language processing approach. Sorting the text into positive, negative, or neutral categories is the aim of sentiment analysis. Many methods, including rule-based systems and machine learning algorithms, can be used to analyse sentiment, or deep learning models. These techniques typically involve analyzing various features of the text, such as word choice, sentence structure, and context, to identify the overall sentiment. Here long short-term memory-based deep learning is applied in this research for the model development purpose. Deeply interconnected neural networks are used in this method. Sentiment analysis can be used in many different applications, such as market research, brand reputation management, customer feedback analysis, and social media monitoring. It shows the use of sentiment analysis in a variety of fields and increases the need of technology to perform it on the existing machines.
Free
Scientific article
This study explores the conversion of English to Hindi, first to text, and subsequently to speech. The first part of the implementation is the text recognition from images, in which two approaches are used for text character recognition: a maximally stable extensible region (MSER) and grayscale conversion the second part of the paper deals with the geometric filtering in combination with stroke width transform (SWT). Subsequently, letter/alphabets are grouped to detect text sequences, which are then fragmented into words. Finally, a 96 percent accurate spell check is performed using naive Bayes and decision tree algorithms, followed by the use of optical character recognition (OCR) to digitize. The word Give our text-to-speech synthesizer (TTS) the recognized text to convert it to Hindi language using the text-to-speech model. Based on aspects such speech speed, sound quality, pronunciation, and clarity.
Free
