Real-time traffic signal control represents a key challenge in modern intelligent transportation systems, particularly under highly variable traffic flows and the presence of priority vehicles. This study proposes a hybrid framework for adaptive signal plan control at a signalized intersection. The framework integrates deep learning-based traffic prediction, surrogate-based performance evaluation, and reinforcement learning-based adaptive control. Short-term traffic flow is predicted using recurrent neural networks, providing anticipatory information for traffic control decisions. Based on predicted flows and generated candidate signal plans, a machine learning surrogate model enables fast estimation of key performance indicators, including average vehicle delay and queue length. Adaptive control is implemented using the Proximal Policy Optimization algorithm within the SUMO environment via TraCI, which enables real-time fine-tuning of signal phases. A dedicated priority and stability module ensures effective emergency vehicle preemption and adaptive public transport priority while preserving intersection stability. Simulation results show that the proposed framework reduces average vehicle delay by up to 35% compared with FT and by up to 15% compared with standalone RL, while also improving traffic flow efficiency and priority vehicle performance.

Production of three-dimensional parts in 3D printing process gains growing importance in various fields, such as: aviation and car industry, architecture, medicine, dentistry, etc. Mechanical performance is an important users' requirement for manufacturers of 3D printed parts. Furthermore, printed part highly depends on process parameters, position and orientation of the printed part, and performances of the 3D printer which prints the part. In this paper, based on experimental results, an artificial neural network has been used for modeling the dependence ofprocess parameters and object orientation during printing, on the one side, and tensile strength as very important mechanical performance, on the other side. After establishing abovementioned dependence, the developed neural network has been used as a fitness function for the genetic algorithm while the genetic algorithm has been created for the optimization process. The result of optimization process was a set of optimal process parameters and part orientation giving the maximum tensile strength. The results have shown acceptable potential of the developed methodology for optimizing the 3D printing process as a complex engineering problem.

The networking of production systems and the communication between production processes, machines, and personnel across various industries is being elevated to a significantly higher level. This advancement leads to the optimization of production processes and raises the overall industry standard. As data exchange and analysis play a pivotal role in optimizing and enhancing system productivity, the reliability and accuracy of the collected information are equally important. This paper presents an example of the successful integration of intelligent sensor systems with robotic solutions for the assembly of specific components in the automotive industry. By using smart sensors and actuators and their networking, parts are assembled more precisely and the possibility of errors is reduced to a minimum.

Today's manufacturing companies face numerous challenges in a dynamic and highly complex business environment. Effective planning, within the management of the production system, has a key role in achieving business success and achieving a competitive advantage for any company. The main setting of the research is the integration of three phases into one intelligent system. The first phase of the research consists of big data optimization of the planning model in the parallel connection of machines, the second phase of the experiment includes the application of different machine learning models, while the third represents the optimization of the planning model in the parallel connection of machines with stochastic processing times, which represents one of the more difficult NP problems of combinatorial optimization. The integration of machine learning models and job planning models in parallel machine connection under conditions of uncertainty is a big challenge. One of the reasons for the application of machine learning models is the influence of the input optimization parameters on the observed objective function. By choosing optimal optimization parameters, it is possible to solve the problem of parallel machine planning with stochastic processing times. The research in the paper aims to significantly improve the performance and reliability of machine planning in various industrial environments by proposing a robust and adaptive solution that can adapt to dynamic conditions and provide optimal results. The main purpose of the paper is the application and integration of an artificial intelligence model in a planning system in order to increase productivity, thereby increasing the competitiveness of small and medium-sized enterprises on the market. These tools can be relatively easily adapted to the needs of the company and would thus enable a better organization of business activities, as well as lower costs and greater business flexibility. The results of the experiment show the success of the proposed methodology.

The development of photoacoustic spectroscopy is being driven by the growing demand for precise, efficient, and reliable detection methods that can be used for in situ measurements and real-time monitoring. Along with rapid technological progress, photoacoustic spectroscopy became an ultra-sensitive, selective, cost-effective technique that can meet the demanding requirements for environmental monitoring, industrial safety, and medical diagnostics. This paper highlights how continuous improvements in photoacoustic technologies, including the use of appropriate laser sources as well as sensing elements, and machine learning methods, are pushing the limits of gases and solid analysis and providing critical tools for addressing modern scientific and industrial challenges.

Planetary gear trains offer numerous advantages over traditional gear systems, including high efficiency, the ability to handle large torque loads, and significant reductions in mass and size for the same torque capacity. However, their relatively complex design necessitates the use of optimization techniques to identify the most suitable configurations for specific applications. A key requirement for effective optimization is a mathematical model that accurately captures the essential operational characteristics of the system. Moreover, the optimization process must account for multiple, often conflicting, objectives. This paper focuses on the multicriteria optimization of a three-stage planetary gear train intended for use in a road vehicle winch. The development of the optimization model involves defining the objective functions, decision variables, and constraints. Optimization criteria were based on the following characteristics: overall volume, mass, transmission efficiency, and the production costs of the gear pairs. In addition to identifying the group of solutions that are Pareto optimal, the model employs the weighted coefficient method to select a single optimal solution from this set. The selected solution is then analyzed through simulation to assess potential gear failure scenarios. By combining optimization techniques with simulation and contact analysis, this study contributes to improving the reliability of planetary gear transmissions.

This paper presents a cognitive robotic system for detecting, classifying, and grasping elongated and deformable objects, such as bananas, carrots, and other produce, which requires precise gripper alignment and orientation. The system integrates deep learning-based object detection using CNNs for real-world positioning and pose estimation to enable adaptive grasping. A 5-DOF robotic arm equipped with a soft-gripping end effector is employed to execute grasping tasks, ensuring gentle handling while minimizing damage. The vision system detects objects, determines their spatial coordinates, and computes an optimal grasping pose based on object shape and orientation. The proposed method leverages real-time image segmentation and contour analysis to assess the gripping width, while an orientation-aware approach enhances grasp stability. The results demonstrate the effectiveness of combining cognitive perception and soft robotics for handling delicate and non-rigid objects in real-world agricultural and industrial applications, improving efficiency and reducing manual intervention.

The topic of piezoelectric active thin-walled structures has attracted a great deal of attention over the previous couple of decades. Lightweight structures with piezoelectric material based active elements, sensors and actuators, offer numerous advantages over their passive counterparts. This explains the motivation of authors to dedicate their work to this enticing research field. Accurate and reliable numerical tools for modeling and simulation of this type of structures is still a hot topic in the research community. This paper offers an isogeometric finite element formulation for shell type of structures made of composite laminates including piezoelectric layers characterized by the electro-mechanical coupling. The shell kinematics is based on the Mindlin-Reissner assumptions, thus including the transverse shear effects. A few examples selected from the available literature are considered to demonstrate the applicability of the developed numerical tool and assess its performance.

As energy efficiency becomes a significant performance indicator in mobile machines, power losses are recognized as an important criterion in the design and optimization of these systems. This paper analyses the loads and power loss due to friction in the revolute joints of the manipulator drive mechanisms during all phases of the loader manipulation task, based on dynamic simulations of the loader model with different variants of Z-kinematics manipulator drive mechanisms, using the MSC ADAMS 2020 software. The analysis is based on a general dynamic mathematical model of the loader, which enables the assessment of the influence of the parameters of the manipulator mechanisms on the functional, structural, and tribological characteristics of the revolute joints within the manipulator’s kinematic chain. Based on the analysis results, a minimum power loss criterion was defined as part of a multi-criteria optimal synthesis procedure for the manipulator drive mechanisms, with the objective of maximizing energy efficiency by minimizing power loss caused by friction in the revolute joints of the manipulator drive mechanisms.

Requirements regarding security, trustworthiness and low latency in modern industrial control and robotic systems have led to implemented machine learning solutions considerably shifting from cloud computing to edge computing. In this context, the emerging concepts of Edge Artificial Intelligence and Tiny Machine Learning are becoming essential for implementing machine learning based industrial control solutions using resource-constrained devices. In this paper Tiny Machine Learning is used to solve the industrial control problem of apple classification using two small edge devices. Several versions are compared while presented experimental results and comparisons with our previously published results demonstrate that Tiny Machine Learning can provide a viable solution even for complex industrial grade control problems in the food industry.

The paper proposes an optimization procedure for maximizing the resistance of composite plates exposed to impact loads. For a composite plate with a predefined composite material, number, and thickness of layers, the set objective is to find the optimal solution in terms of the layer orientation so as to withstand the impact test. The fiber orientation angle is treated as a continuous design variable within the context of the problem. The commercially available finite element software package Abaqus is used to model a Kevlar 49/Epoxy composite plate and simulate its mechanical behavior when exposed to an impact load. As this deals with a highly dynamic process that involves significant nonlinear effects, an explicit time-integration scheme is selected. Prediction of the plate damage based on its maximum stress failure criteria is used as the objective function for optimization, whereas the penetration analysis is based on the Hashin criteria and implemented in an Abaqus VUMAT subroutine. The obtained results are expected to be of interest to ballistic vest manufacturers to develop passive protection solutions.

Headway variability has a negative impact on the public transport passengers' perception of service quality. However, most of the existing methods aimed at improving the headway regularity operate in real time and require precise vehicle location data, making it difficult to implement them in practice. On the other hand, planning-level methods can be used to increase the resilience of public passenger transport (PPT) to the accumulation of headway disturbances. As this is typically done from the operator's perspective, the passengers' perspective tends to be overlooked, motivating the current work. In this article, an optimisation procedure for evaluating the viability of diametrical line splitting in terms of passenger travel time and headway regularity is proposed. The aim is to increase the robustness/resistance of the PPT system to the propagation of headway disturbances without reducing the service quality. The developed optimisation procedure was validated by applying it to real data pertaining to an urban PPT line. The results show that there is a positive correlation between the transport demand and the effects of the optimisation procedure, whereby an increase in the primary headway disturbance increases the sensitivity of the optimisation procedure to the transport demand.

Convolutional Neural Networks (CNNs) have established themselves as a cornerstone in object detection and classification, delivering exceptional performance across many applications. However, the efficacy of CNNs is heavily dependent on the meticulous design and optimization of their architecture and hyperparameters, a process that is often labor-intensive and computationally demanding. Genetic Algorithms (GAs), inspired by the principles of natural evolution, present a viable solution to automate and enhance the optimization of CNNs. This survey reviews the integration of genetic algorithms into the development and refinement of CNN architecture for object detection and classification tasks. We explore a range of GA-based approaches, including architecture optimization, hyperparameter tuning, and ensemble methods, highlighting how these techniques improve CNN performance. The survey also delves into application areas of medical image analysis and agricultural monitoring, demonstrating the versatility and effectiveness of GA optimization strategies. By synthesizing findings from recent studies, this paper highlights key trends, identifies prevalent challenges, and outlines future research directions in the convergence of genetic algorithms and deep learning. This survey aims to provide researchers and practitioners with a consolidated understanding of how genetic algorithms can be leveraged to advance the performance and applicability of CNNs in object detection and classification.

In food industry automation, the integration of machine learning with robotic systems can introduce improved efficiency and precision in various tasks such as fruit and vegetable handling and manipulation. This paper introduces an innovative approach to robotic grasping, specifically customized for the challenge of identifying and handling apples and oranges within a dynamic and unstructured environment. Utilizing the YOLO (You Only Look Once) object detection algorithm, the system consists of a camera mounted on the robot's end effector, enabling real-time identification and localization of fruits and vegetables. Implemented on the NVIDIA Jetson platform, this solution presents the combination of machine learning techniques with hardware acceleration, ensuring optimal performance in real-time object recognition and manipulation tasks. Our experimental setup tests the robustness of our object recognition and localization approach under varying lighting and orientation scenarios and evaluates the system's overall performance in a real-world, cluttered environment typical of the food industry. The findings of this study underscore the potential of integrating ML-powered vision systems with hardware-accelerated robotics for food handling and manipulation applications. By demonstrating high levels of accuracy in recognition, classification and grasping, our research contributes to the broader field of robotic automation, offering insights into the scalability and adaptability of such systems across different sectors of the food industry.

The paper deals with the isogeometric analysis (IGA) of active composite laminates with piezoelectric layers.IGA is a special formulation of the finite element method (FEM) that aims at seamless integration of geometric and finite element modelling.NURBS basis functions are employed to develop isogeometric shell formulation based on the Reissner-Mindlin kinematics.Piezolayers characterized by electro-mechanical coupled field effects enable active behavior of the considered structures.The electric field acts across the thickness of the piezolayers and is coupled to the in-plane strains.In addition to a number of advantages that NURBS modelling provides, defining the surface normal vector at the points of the control polygon, which are generally not located on the surface, creates certain difficulties.A method of determining the surface normal vectors at the points of the control polygon based on the Greville's points is discussed.In order to demonstrate the applicability of the developed formulation, a benchmark case is computed and the results are compared with those obtained by means of classical FEM formulation, which are available in the literature.

Photoacoustic spectroscopy achieves high sensitivity and selectivity using high-power lasers. Variations of pa-rameters such as laser beam spatial profile and fluence phi can alter precision of photoacoustic measurements. Numerous commercial instruments aren't usable for high phi values measurement, due to possible harmful effects. To estimate high phi values from photoacoustic signals in time domain we applied computational intelligence method: adaptive-network-based fuzzy inference system (ANFIS). Experimental photoacoustic signals are generated in two different gas mixtures: SF6 + Ar and C2H4 + Ar, for phi values ranging from (0.2-1.4) J center dot cm -2. Obtained results indicate that different absorption characteristics of examined molecules and various signal intensities don't influence ANFIS prediction, due to its adaptation ability and error tolerance in dealing with imprecise and noisy data. Furthermore, robustness, high learning capability and self-correction, make this technique computational effective and recommendable for in situ photoacoustic measurements. Aside from many advantages, limitations of the proposed method are also discussed.

Innovative lightweight smart structures incorporating piezoelectric material-based active elements, both as sensors and actuators, have been identified to present manifold advantages over traditional passive systems. Such structures have become intrinsically integrated into smart mechatronic systems, necessitating advanced design, testing, and control techniques. Real-time simulation of shell-type deformable objects, especially when employing the finite element method for non-linear analysis and control, has been challenging due to the extensive computational demand. Presented herein is an efficacious implementation leveraging machine learning with the isogeometric finite element formulation. This implementation focuses on shell-like smart mechatronic structures crafted from composite laminates comprising piezoelectric layers, which are characterised by electro-mechanical coupling. The foundation for the shell kinematics is derived from the Mindlin-Reissner assumptions, effectively incorporating transverse shear effects. While the inclusion of machine learning facilitates real-time efficient operations, the isogeometric finite element analysis (FEA) introduces pronounced advantages over conventional finite element method (FEM), also serving as a valuable source of offline data crucial for the training phases of machine learning algorithms. A piezo-laminated semicircular arch has been analysed to exemplify the effectiveness and performance of the presented methodology. Explorations into further machine learning techniques and intelligent control schemes are also contemplated.

Background: Previous research has shown different effects of hematological malignancies on the outcome of patients with COVID-19 infection depending on the type of disease and the treatment received. This research was aimed at examining the clinical outcome of COVID-19 infection in positive patients with chronic myeloid leukemia treated with tyrosine kinase inhibitors. Methods: We collected retrospective information on chronic myeloid leukemia patients who were treated and monitored in our institution during the pandemic period. Within this cohort, we recorded COVID-19 positive symptomatic patients and analyzed their basic characteristics, symptoms, severity, and outcome. Results: In the study cohort when COVID-19 was diagnosed, 86.7% of patients were on first-generation tyrosine kinase inhibitors therapy—imatinib. At the time of infection, 70% of patients were in molecular remission, 23.4% in complete cytogenetic remission, and 3.3% in complete hematological response. Most patients had symptomatic disease. Within the analyzed group, 56.7% of patients had asymptomatic/mild COVID-19 infection, 23.3% of patients had moderate symptoms which did not require hospitalization, and 20% of patients had severe/critical symptoms that required admission to the intensive care unit. More than half of the patients interrupted treatment with tyrosine kinase inhibitors temporarily during COVID-19. There were no deaths due to COVID-19 infection. Conclusions: In compliance with other larger clinical studies, analysis of the clinical outcome of COVID-19 infection in patients with chronic myeloid leukemia on tyrosine kinase inhibitors therapy in this study showed that they do not have an increased risk for COVID-19 infection and that they have a mild course of the disease with recovery.

Photoacoustic spectroscopy is a powerful, non-destructive, ultrasensitive technique that covers a wide range of applications including atmospheric monitoring, industrial, environmental, and biomedical practice. In this paper, our attention is focused on the area of artificial intelligence implementation in photoacoustic spectroscopy of gases. Artificial intelligence has been proven as a very successful, effective, and promising method for the accurate and real-time determination of photoacoustic signal parameters, related to relaxation, thermal and other physical properties of various media (i.e., for solving the inverse photoacoustic problem). To improve the sensitivity and selectivity of the photoacoustic method feedforward multilayer perceptron network is applied for real-time simultaneous determination of photoacoustic signal parameters: vibrational-to-translational relaxation time, and radius of the laser beam. Also, to solve the problem of finding optimal values of these photoacoustic parameters, metaheuristic algorithms, genetic algorithms and simulated annealing are used. The performance of artificial intelligence methods has been tested on a set of experimental signals generated in the (SF6+Ar) gas. The potential advantages and disadvantages of those methods are discussed.

Abstract BACKGROUND: In this era of target therapies, novel data on the correlation between response endpoints and survival outcomes in multiple myeloma have arisen. OBJECTIVE: To determine the impact of quality of response on clinical outcomes, using first-line treatment, and identify risk factors influencing progression-free survival (PFS) and overall survival (OS) among myeloma patients. DESIGN AND SETTING: Retrospective analysis on myeloma patients who were treated at the Clinic of Hematology and Clinical Immunology, University Clinical Centre, Niš, Serbia, over a four-year period. METHODS: A total of 108 newly diagnosed patients who received first-line therapy consisting of conventional chemotherapy or novel agent-based regimens were included in this analysis. RESULTS: The quality of response to first-line therapy for the whole cohort was classified as follows: complete response (CR) in 19%; very good partial response (VGPR) in 23%; partial response (PR) in 38%; and less than PR for the remaining patients. After a median follow-up of 25.4 months, the three-year PFS and OS for the entire study population were 47% and 70%, respectively. Achievement of CR was the main factor associated with significantly prolonged PFS and OS, in comparison with patients who reached VGPR and PR. Likewise, addition of the new drugs bortezomib and thalidomide to standard chemotherapy led to considerably extended PFS and OS, compared with conventional therapy alone. CONCLUSIONS: This analysis demonstrated that the quality of response after application of first-line treatment using novel agent-based regimens among multiple myeloma patients was a prognostic factor for PFS and OS, which are the most clinically relevant outcomes.