Advances in Intelligent Manufacturing and Service System Informatics

R&D project selection is one of the most important issues for an R&D center. Evaluating more than one project in terms of different criteria, selecting and implementing the most appropriate project is very critical for both the company’s profit and the sustainability of the project. The project selection process is handled by different processes in companies. Due to the importance of this issue, companies adopt a selection process in line with their own strategies. In this study, an application was carried out with the fuzzy TOPSIS method to evaluate alternative project ideas that will be an R&D project in the R&D center of an automotive company. 4 different criteria were evaluated by experts for 6 different project ideas. With the implementation realized as a result of expert evaluations, a priority order was obtained for 6 project ideas. In practice, as a result of the evaluation, the alternative project P5 with the highest value in the ranking is selected as the next R&D project to be started.

Standardized perceptions of beauty have always existed through bodies, which are expressions of our characters and identities. As societies have changed, these perceptions have changed shape along with societies. But in a globalizing world with social media, beauty standards also tend to go global. In this direction, the sense of beauty and the way of life of Western societies have been positioned as the goal sought to be reached in the whole world. Therefore, having slanted eyes or black skin has been declared ugly, beyond racism, because it does not fit the ideal perception of beauty.In this study, this ideal beauty, which is about the whole body and self, is evaluated through the face editor applications applied to the portraits in which identities and characters are revealed. In this context, FaceApp: Face Editor application, one of the most popular face-changing and editing applications, is taken as an example.Such applications based on machine learning and artificial intelligence take advantage of the user’s location, which is also data for them.The issue of protecting personal data, which is one of the biggest problems, and the possibility of it, as well as the fact that it is becoming increasingly difficult to distinguish between real and fake in a world centered on commodification, will be discussed.This study, in which the descriptive analysis method is used by examining the data, aims to draw attention to the current problems of the society that has become identical in the effort of differentiation and the causes of these problems.

Electro-erosion wear (EEW) is a significant problem in the mold steel industry, as it can greatly reduce the lifespan of electrodes. This study presents a machine-learning approach for predicting and modeling electrode and workpiece wear on an electrical discharge machining (EDM) machine. In the experimental design, EDM of CuCrZr and Cu electrodes of AISI P20 tool steel was carried out at different pulse currents and duration levels. In addition, CuCrZr and Cu electrodes used in the experiment were cryogenically treated at a predefined degree for multiple periods and then tempered. This study employed machine learning algorithms such as decision trees, random forests, and k-nearest neighbors to model the EEW of cryogenically treated electrodes made of mold steels. The results were compared according to the coefficient of determination (R2), adjusted R2, and root mean squared error. As a result, the decision trees outperformed the other algorithms with 0.99 R2 performance. This study provides valuable insights into the behavior of EEW in mold steel electrodes and could be used to optimize the manufacturing process and extend the lifespan of the electrodes.

The undesirable protrusions and roughness that occur on the surfaces of the parts during or after the production process are called burrs, and the process applied to remove these burrs is called “burring”. Burrs are mostly formed around the cut edges. A little more material than the measure is transferred to the mold so that the metal is not missing in a piece to be poured into the mold. This excess, when the molds are compressed, overflows from the joints and creates casting burrs. Post-production burr removal is an important issue in the manufacturing industry. It is necessary to clean the burrs formed after the production in the z-rod parts produced by the casting method. Burrs formed in the holes cause material problems and connection problems. Burrs cause stress at the hole corners due to high stress, reducing cracking resistance and fatigue life. In mating parts, burrs will enter the connector seat and damage the connector or assembly. Burrs in the holes will also affect the coating thickness on the rough surfaces, thus increasing the risk of corrosion. Burrs on moving parts increase unwanted friction and heating. Currently, this process is generally cleaned manually by the operator using a deburring tool (blaster) or it can be cleaned using high cost and special features robot arm integrated systems. Manual cleaning causes a significant loss of productivity in the labor factor. On the other hand, deburring operation and post-operation product control vary according to the operator’s competence and initiative. Therefore, the cleaning operation is not always at the same standards. With this study, a system was designed for deburring, apparatus suitable for the dimensions of the z-rod parts. And the deburring process was carried out in a standard way for mass production. At the same time, the process accuracy was instantly checked with the highest accuracy by using the image processing algorithms with the Python programming language, and the burr types were determined by detecting the burr form and size with the deep learning method. System cleaning efficiency was measured as 92% higher than manual cleaning.

The aim of this paper is to introduce a novel classification algorithm based on distance to class centroids with weighted Euclidean distance metric. Features are weighted by their predictive powers and in-class homogeneities. For predictive power, information value metric is used. For in-class homogeneity different measures are used. The algorithm is memory based but only the centroid information needs to be stored. The experimentations are carried at 45 benchmark datasets and 5 randomly generated datasets. The results are compared against Nearest Centroid, Logistic Regression, K-Nearest Neighbors and Decision Tree algorithms. The parameters of the new algorithm and of these traditional classification algorithms are tuned before comparison. The results are promising and has potential to trigger further research.

Recommendation systems are essential as they help users to discover new books and resources and increase their engagement and satisfaction, thus improving the overall learning experience. This paper presents a web-based intelligent book recommendation system for smart campus applications at Izmir Bakircay University. The system is designed as an intelligent hybrid tool that combines collaborative and content-based filtering techniques to recommend books to users with methodological differences. It considers the user’s reading history and preferences and integrates with other smart campus applications to provide personalized recommendations. The system is important for the digital transformation of smart campuses as it helps to make education more personalized, efficient, and data-driven. Also, it allows for the effective use of public resources. The effectiveness of the system was evaluated through user feedbacks, 22 users evaluated 148 books and the results showed that users responded positively to about 70% of the recommended books thus, it provided accurate and personalized recommendations.

Today, when technological developments are accelerating, the number of devices we use is increasing and household goods are equipped with smart functions and gain a more important place in our lives. Especially with the pandemic, the search for practical solutions that make housework easier has made online purchasing behavior indispensable. In this study, the problem of ranking the 6 best-selling new generation vacuum cleaner (NVC) types in a leading shopping site in Türkiye’s e-commerce market was discussed. For this purpose, the electronic word-of-mouth communication (e-WOM) for these products on a review platform in Türkiye where customers share evaluations of their experiences with the products or services was examined. In the first part of the study, the criteria of these products as specified on the review platform were evaluated by interviewing salespeople working in different electronics stores. Then, these criteria weights were obtained by using the Pythagorean Fuzzy Analytical Hierarchy Process (PFAHP) method. In the second part of the study, the most suitable NVC type was determined by the Pythagorean Fuzzy Technique for Order Preference by Similarity to Ideal Solutions (PFTOPSIS) method, taking the criteria weights into account obtained with PFAHP and customer satisfaction scores (CSS) on the review platform. Microsoft Excel 2010 program was used in the calculations. With the results obtained from the calculations, the type of vacuum cleaner that can adequately respond to the requests of the users and provide the highest satisfaction among the NVC with the highest sales has been determined. As far as is known, no study has employed hybrid PFAHP-PFTOPSIS method for the product ranking by using CSS. As a result, we think that this study provides a different perspective to the literature in this field.

Nowadays, the need for standardization of all physical and chemical processes for high quality products leads us to adopt fast and precise decision-making mechanisms, such as image processing. Image processing is preferred in various areas due to its ease of application and proven usefulness. In the food industry specifically, it has been utilized for quality control of products at each stage of production. In this study, the aim is to prepare an image processing algorithm that determines the homogeneity of almond spread and almond mass of the chocolate coatings topped with almond. 40 images were digitally produced to represent the chocolate coating with almond. Half of the images were designed to be examples of homogenously spread almonds while the other half were of non-homogeneously spread almonds on the chocolate coating. The produced images were formed in 4 different almond mass, which were uniformly distributed among the data. The images were then processed to determine the homogeneity of almond spread and almond mass on the chocolate coating. Co-occurrence-matrix was used to test the homogeneity of the spread. With a properly chosen offset value, co-occurrence-matrix correlation value of the image was observed to be able to determine the homogeneity of the samples. To determine the almond mass, the images were first converted to black and white form and then simply the white color ratio in the images were evaluated, which is observed to be directly proportional to the almond mass. Consequently, it is shown that the proposed image processing methodology can be successful at determining homogeneity of almond spread and almond mass. Furthermore, the proposed methodology may also be utilized for other quality control applications where determination of homogeneity and amount are of importance.

Quality is a measurement-based criteria that specifies the conformity of final products to certain rules and agreements. Monitoring product quality has always been critical, cost-effective and time-intensive process during manufacturing. Surface defects have major negative impacts on the quality of industrial products. Human inspection for visual quality control is challenging and less reliable due to the influence of physical and psychological factors on the auditor, including fatigue, stress, anxiety, working hours, and environmental conditions. Considering these risks, it is impossible for humans to deliver satisfactory stable performance over a long period of time and without interruption. Moreover, with the advancements in hardware and software, quality control can be done fast, reliable, efficient, and repeatable regardless of duration. Herein, we propose morphology-based image processing approach that enables detecting the scratch or alike defects with a width of 70 µm, which corresponds to 85 pixels of a 5181 × 5981 pixels image or 154 µm corresponding 225 pixels of a 7484 × 7872 pixels image. A scanner-based device is employed to capture the images and we combined dilation, closing, median filtering as well as gradient taking and edge detection, then contour finding. Our algorithm exceeds the performance of handcrafted feature-based methods on detecting tiny defects within very large images, which even outperforms the modern deep learning-based methods when there is not/enough training data thanks to not requiring training data. The inference time of our approach for an image is less than 1 s and consequently is capable of being exploited online surface defect detection applications robustly.

In order for countries to have strong armies, they should attach importance to defense industry projects. Therefore, countries allocate high resources to defense industry investments in order to be a deterrent against their enemies. In recent years, the concept of sustainability has started to gain importance in the defense industry, as in many other sectors. Sustainability practices in the defense industry play an active role in the success of defense industry projects. One of the most important links of sustainability is sustainable supply chain management (SCM). Therefore, in this study, the problem of sustainable supplier selection (SSS) in the defense industry has been examined. Since multi-criteria decision-making (MCDM) methods are widely used in supplier selection, MCDM methods were also used in this study. In the study, first of all, sustainable supplier selection criteria (SSSC) in the defense industry were determined by making use of expert opinions and studies in the literature. The weights of the determined criteria were determined using the Analytical Hierarchy Process (AHP) method. In the last part of the study, a comparison was made between different suppliers working in the defense industry by using the Fuzzy Technique for Order Preference by Similarity to Ideal Solution (FTOPSIS) method and the weights previously determined by the AHP method, and the best supplier was determined as a result of the calculations.

Digitalization, which began with Industry 4.0 and has grown with great momentum due to the pandemic, is becoming significantly important to industries. Companies seeking to take their place in fierce competition with the global economy have begun to integrating digitalization into processes to increase their profitability and ensure operating excellence in this regard. In this study, simulation modelling was performed by using ARENA software, which accelerates decision-making processes completely objective, and allows determining and choosing the best possible scenarios without implementing any change in present system. The proposed method was applied to reduce the penalty costs in a company operating in the automotive sector by determining operational and strategic improvement with design of experiment. The main aim of this study is maximizing the profitability of a company by decreasing penalty cost and eliminate waste and bottlenecks. As distinct from literature studies, this study provides the opportunity to optimize any production process by following the steps explained elaborately, regardless of the sector. As a result of the study, the production factors were optimized by design of experiment and the recommendations regarding the investment decisions and production factors were given to the company to minimize penalty costs within the framework of the company’s constraints.

Digitalization can be defined as the transfer of activities performed in a field to digital environments. The application of digitalization in industry is revolutionary. The digitalization in industry can include applications such as collecting, analyzing, and managing company data with digital technologies, digitally monitoring and controlling the transfer of information between departments, and thus optimizing processes. In human resources management, digitalization can facilitate employee management in a variety of ways, increasing productivity and enabling better decisions. Human resources (HR) departments can develop more effective human resource management strategies by taking into account the amount of time employees are likely to work in the organization while making decisions such as incentives, bonuses, salary increases, and promotions. In this study, a decision support system is proposed to assist HR in determining the most appropriate departments for employees by predicting the potential working hours of current or new/to be hired employees in the organization. To estimate the potential work hours, we have used machine learning techniques that are widely used in the literature. We have adopted an assignment algorithm with work hour prediction to determine of the most suitable departments for employees. An application is carried out on a data set that has been published in the literature, and the results are discussed.

Predictive maintenance occupies a significant role to drop the operation and maintenance costs in production systems. Remaining useful life (RUL) prediction is one of the most preferred tasks in predictive maintenance decisions. Recently, deep learning techniques are extensively employed to accurately and effectively predict remaining useful life (RUL) by examining the past deterioration data of machinery and equipment failures. In this study, a deep learning approach that includes multiple separable convolutional neural networks (CNN), a bidirectional long short-term memory (Bi-LSTM) and fully-connected layers (FCL) are proposed to ensure more effective predictive maintenance planning. Separable CNN layers are applied to learn the nonlinear and sophisticated dependencies from the raw degradation data while the Bi-LSTM layer is employed to capture the long-short temporal characteristics. Besides, the dropout method and L2 regularization are used in the training stage of the proposed deep learning approach to achieve more accurate learning. The effectiveness of the proposed approach is verified by the popular FEMTO-bearing dataset presented by NASA. Finally, it is aimed that the experimental results provide better prognostic prediction compared with the benchmark models.

Internet of Things (IoT) is an innovative technology that enables physical objects used in daily life to exchange data among themselves over the Internet. Healthcare is one of them. The basic idea behind the Internet of Medical Things (IoMT) applications is to detect and process patient data without any restrictions, and to provide remote communication with smart devices. Especially chronic disease follow-up and early warning applications are some of the current and effective usage of this technology. In addition, applications such as remote monitoring, telerehabilitation, smart sensors and medical device integration can be given as examples. The fact that it has the potential to increase the effectiveness and efficiency of not only patients but also healthcare professionals in the diagnosis and treatment processes is an encouraging factor for the use of the Internet of Things in the field of health.When evaluated from the perspective of individuals, IoMT can enable better health self-management and personalization of the content of medical services. When evaluated from the perspective of the Ministry of Health and health institutions, it can reduce costs and ensure efficient use of the budget and increase the share allocated to other health services. For research institutions, smart healthcare can improve overall efficiency by reducing research costs and time.In this study, it is aimed to examine the use and examples of IoMT technology in the health sector, to discuss the advantages, potential gains, and possible difficulties in the application of this innovative technology that can change the way of health service delivery, and to offer suggestions.

In recent years, the cloth industry has faced fast fashion trends. This has resulted in Fast fashion as a supply chain model for clothing and accessories that is supposed to respond quickly to the latest fashion trends by frequently updating the products already available in the inventory. However, Fast Fashion has created serious challenges for the sustainability of the clothing industry. This paper investigates the use of social media analytics to understand fashion trends in the preowned fashion industry. The study aims to establish a link between environmental pollution and fast fashion by investigating the preowned fashion industry from both consumer and business perspectives. To achieve this, the study proposes a social analytics (SA) approach to analyze social media posts and predict preowned fashion trends. By using SA techniques, the study hopes to provide valuable insights into consumer behavior and preferences in the preowned fashion industry, which can be used to promote sustainable fashion practices and reduce environmental pollution. Overall, the study demonstrates the potential of social media analytics in understanding and predicting fashion trends, with the goal of promoting sustainable fashion practices.

The subject of the study is ordered flow shop scheduling problems which were first seen in the literature in the 1970s. The main objective is to get a fast and good solution. For the purpose of the study, firstly, the ordered flow shop scheduling problems were defined. Then, a heuristic method was suggested for the ordered flow shop scheduling problems and a sample problem was solved and discussed. A genetic algorithm (GA) based on the complexity property was developed. Full enumeration was applied to determine the optimal makespan values based on Smith’s rule because Smith had specified the conditions under which circumstances the permutation can be optimum.This study is one of the few studies in the literature to obtain optimum solutions up to 15 jobs ordered flow shop problems in a very short amount of time. The developed GA heuristic can also solve large size ordered flow shop scheduling problems very fast. The significant advantage of the proposed method is that, while Smith’s method rule not does not work with large size problems since it requires full enumeration to identify best solution to follow a convexity property where GA algorithm can find a good solution very fast.

Energy-saving technologies for heating and cooling systems in buildings have received much attention. Energy saving is even more important in the heating and cooling systems of intermittent use places such as mosques. Because the stop-start operation of the mechanical systems causes the overall efficiency to decrease even more, and the thermal comfort of the environment cannot be provided. In order to save energy and adapt to variable environmental conditions, this study investigates the fuzzy logic control structure of the intermittent radiant heating and cooling system. Fuzzy logic control can handle imprecise or uncertain information, making it an effective solution in complex systems where mathematical models are difficult to derive. The control strategies included actual ambient air temperature, outdoor temperature, return water temperature and system on/off timing. Expert knowledge and observations of system performance were used to create fuzzy logic rules. Studies were carried out on the radiant floor heating and cooling system with ground source heat pumps and thermal energy storage of a mosque.

Linguistic summarization, a subfield of data mining, generates summaries in natural language for comprehending big data. This approach simplifies the incorporation of information into decision-making processes since no specialized knowledge is needed to understand the generated language summaries. The present research employs linguistic summarization to examine the circumstances surrounding the Carbon Border Adjustment Mechanism, one of the most significant regulations confronting exporting nations to the European Union, and will be adopted to support sustainable growth. In this paper, associated with several attributes of the countries and product flow from exporting countries to European countries were defined as nodes and relations, respectively. Before the modeling phase, fuzzy c-means automatically identified fuzzy sets and membership degrees of attributes. During the modeling phase, summary forms were generated using polyadic quantifiers. A total of 1944 linguistic summaries were produced between exporting countries and European countries. Thirty-five summaries have a truth degree greater than or equal to the threshold value of 0.9, which is considered reasonable. The provision of natural language descriptions of the Carbon Border Adjustment Mechanism is intended to aid decision-makers and policymakers in their deliberations.

In the modern era, mobile robots are gaining special attention in various intralogistics operations such as warehousing, manufacturing, electrical high-voltage substations, roads, etc. These vehicles should be capable of effectively recognizing their routes, avoiding singularities and obstacles, and making a decision according to the environment. Hence an advanced control mechanism is required for path planning and navigation to work effectively in that dynamic environment. Keeping insight, into the challenges faced by the mobile robot this study aims to develop path planning using global (A* and Dijkstra) and local planners (Pure Pursuit) in a 2D navigation system, utilizing g-mapping for Simultaneous localization and mapping (SLAM) and Adaptive Monte Carlo Localization (AMCL) for probabilistic localization, to solve this issue. The system is designed to be compatible with the Robot Operating System (ROS) ecosystem. Path planning is carried out on a lower-resolution grid covering the navigable areas and the Pure Pursuit approach is enriched with a Proportional-integral-derivative (PID) controller. The results have shown that proposed schemes give superior performance in challenging obstacle-based warehouse systems, compared to publicly available ROS navigation planners.

The digital twin technology coordinates digital and physical spaces in order to improve the current and future actions in the system based on the real-time data. It allows organizations to follow and optimize their systems in a virtual environment before performing actions in reality. Digital twin can play a role in labor planning within an organization as well as in smart manufacturing environments such as performing the simulation of different scenarios, helping to determine the most efficient use of multi-skilled workers. In case of unexpected absences or changes in labor resource during a shift, organizations need to reconsider labor assignments to reduce downtime and inefficiencies. Traditionally, these actions are performed by the shift supervisor. In industry 4.0 concept, we design a digital twin-based decision support system with simulation capabilities for dynamic labor planning. The proposed system allows the unit to adapt to new conditions and also provides performance measures for the future state of the system. Additionally, the operator can simulate different scenarios and evaluate their performances. We present the results and performance of the proposed system on a case example.

The increase of digitalization has enhanced the collection of time series data using sensors in various production and service systems such as manufacturing, energy, transportation, and healthcare systems. To manage these systems efficiently and effectively, artificial intelligence techniques are widely used in making predictions and inferences from time series data. Artificial intelligence methods require a sufficient amount of labeled data in the learning process. However, most of the data in real-life systems are unlabeled, and the annotation task is costly or difficult. For this purpose, active learning can be used as a solution approach. Active learning is one of the machine learning methods, in which the model interacts with the environment and requests the labels of the informative samples. In this study, we introduce an active learning-based approach for the time series classification problem. In the proposed approach, the k-medoids clustering method is first used to determine the representative samples in the dataset, and these cluster representatives are labeled during the initialization of active learning. Then, the k-nearest-neighbor (KNN) algorithm is used for the classification task. For the query selection, uncertainty sampling is applied so that the samples having the least certain labels are prioritized. The performance of the proposed approach was evaluated using sensor data from the production and healthcare systems. In the experimental study, the impacts of the initialization techniques, number of queries, and neighborhood size were analyzed. The experimental studies showed the promising performance of the proposed approach compared to the competing approaches.

Hands are the most used parts of the limbs while performing complex and routine tasks in our daily life. Today, it is an important requirement to determine the user’s intention based on muscle activity in exoskeletons and prostheses developed for individuals with limited mobility in their hands due to traumatic, neurologic injuries, stroke etc. In this study, 5-finger movements were classified using surface electromyography (EMG) signals. The signals were acquired from forearm via the 8-channel Myo Gesture Control Armband. EMG signals from three participants were analyzed for the movements of each finger, and the activity levels of the channels were compared according to the movements. Following, movement classification was performed using the Gaussian mixture network, a statistical artificial neural network model. According to the experimental results, it was seen that the model achieved an accuracy of 73.3% in finger movement classification.

Sectoral growth is increasing day by day and the competition market is growing with it. At the same time, customer awareness is also increasing. As customer awareness increases, the quality of service provided should also increase. One of the ways that companies will apply in order to maintain their existence in this competitive environment and to prevent customer loss is to make the lean manufacturing philosophy a corporate culture. It is a production approach that does not contain any unnecessary elements in the lean manufacturing structure, minimizes waste and aims to increase efficiency in production. When moving to the lean manufacturing philosophy, it is of great importance for companies to draw a correct road map. This study was applied to the product/product group produced in a foundry. Value stream mapping (VSM), which is one of the lean manufacturing techniques for the determined product/product group, was made and the current situation value stream map was created. With value stream mapping, bottlenecks and losses in the process were determined and a future situation value stream map was created. Lean manufacturing techniques were applied at these determined points, problems were eliminated and productivity increase was achieved in production. Fuzzy logic was used to clearly determine the productivity increase. Fuzzy logic creates numerical models by imitating the human mind many vague, non-numerically expressed terms that we use daily. With fuzzy logic, the efficiency rate was modeled numerically and the contribution of the lean manufacturing techniques applied to productivity was determined.

Transportation of perishable food in cold chain logistics systems is crucial in order to preserve the freshness of the products. Due to the extended traveling times and frequent stops, planning the distribution operations in cold chain logistics plays a vital role in minimizing the deterioration cost of the products. In order to minimize the total cost of cold chain logistics activities related to the purchase of perishable products, the route and procurement operations have to be well-planned. In this context, this paper addresses the well-known traveling purchaser problem (TPP) and extends the TPP by considering the procurement of perishable products. This is called the traveling purchaser problem in cold chain logistics (TPP-CCL). In the TPP-CCL, the demand for a number of perishable products is provided from a number of markets, where the products purchased at markets are transported by a temperature-controlled vehicle. In addition to the transportation and procurement cost, the deterioration cost of the products is taken into account in the problem. The problem is formulated as a non-linear mixed-integer programming model in which the objective is to find the best procurement and route plan for the purchaser that minimizes the total cost. Considering the complexity of the problem, a simulated annealing (SA) algorithm is proposed to solve the TPP-CCL. The SA is formed by using a number of local search procedures, where the procedures are randomly selected to find a new solution in each iteration. The proposed SA is performed for a TPP-CCL problem set that includes different-sized instances. The results of the SA are compared to the GUROBI solver results. A better result is obtained by the SA for most of the instances. The computational results show that the proposed SA outperforms the GUROBI results by finding better results in shorter computational times.

In automatic feeding systems, feeding of characteristic workpieces by mechanical tools causes accuracy and cost difficulties. For this reason, in systems where special workpieces are fed, image processing applications are necessary to obtain characteristic features of a product. In this study, a novel image processing algorithm is proposed for feeding a workpiece which has characteristic geometrical structures. The proposed algorithm is based on obtaining geometrical and rotational properties of the product and the gradient-based analysis as follows. The first step is to extract features from the shape of the workpiece, this step includes noise reduction, filtering, and edge detection operations. The gradient values of the edge information are used to create the angle-length vector pair in the second step. The workpiece rotation information is derived from length values indexed with angle information. The last step involves determination of the workpiece position in the 2D coordinate system. The coordinate information is used to determine the position of the gripper holder. The coordinates and angle are transmitted to the feed control. The proposed algorithm is applied on the 800 images that are collected from manufactured products. Rotation angle of the workpiece is determined by a tolerance of 1.5°. It is seen that results have sufficient accuracy for industrial applications.

Making investment decisions for infrastructure projects, monitoring and controlling investments is one of the critical issues that policy makers need decision support. Successful integrated project management is needed to evaluate the projects structurally and economically from a holistic perspective. It is necessary to design a framework to manage factors such as financing, technical capacity, contract management, equipment planning, known as limited resources in infrastructure projects. A roadmap guide is needed to manage these projects with limited resources. The use of digital systems in this regard provides an important advantage. Especially in public administration, monitoring the planned, ongoing, and completed projects on a portal with digitalisation opportunities are possible. GIS- based UYS developed by The Ministry of Transport and Infrastructure for digital monitoring of infrastructure projects is described in this article. Within the scope of system engineering, the UYS system has modules such as contract, performance, and planning. Program and cost performance indices can be calculated using the UYS methodology. Thanks to the UYS system, it’s possible to keep track of the goals set for contract management and constantly monitor the most critical tasks. The developed data health approach provides to the monitoring of the current data condition on the UYS platform. In addition, other projects related to the mega-project that followed can be monitored via UYS with integrated management. Approaches to the development of these modules are discussed in this paper. With the concept of earned value, the planned projection and actual follow-up for the projects are carried out, and activities with an increased risk level can be closely followed. It is also explained as a case study through the work of “Ankara-Sivas HSR Project”, a mega-project in which the earned value approach of the UYS system is followed.

This research aims to compare two well-known solution methodologies, namely Geographical Information Systems (GIS) and Discrete Event Simulation (DES), which are used to design, analyze, and optimize the solid waste management system based on the locations of the garbage bins. A significant finding of the study was that the application of the simulation methodology for a geographical area of a size of 278 km2 was challenging in that the addition of the geographical conditions to the developed model proved to be time-consuming. On the other hand, the simulation model that was developed without adding geographical conditions revealed that the number of bins could be reduced by 60.3% depending on the population size and garbage density. However, this model could not be implemented since the required walking distance was higher than 75 m, which is greater than the distance the residents could be reasonably expected to travel to reach a bin. Thus, using a cutoff value of 75 m, the total number of bins can be reduced by 30% on average with regard to the result obtained from the GIS-based solution. This can lead to an annual cost reduction of 93.706 € on average in the collection process and carbon dioxide release reduction of 18% on average.

As the intelligent technologies progress, the control systems and designing of efficient process systems has become more important. Besides that, analysis of data in the production lines play crucial role in the process. One of the essential parts of the systems in electric vehicles is battery systems and cooling of these batteries can be implemented by thermal isolators. In this study, an industrial imagining control system using profiler laser sensor for thermal mastic surface mapping was developed. For this, 3D image data were acquired and analyzed by a Laser Profiler Sensor that attached on a robot. In that case, the size of the mastic lines was measured and checked whether the sizes remained within tolerances which are defined by the car producers. Location data are sent to the robot and sealing process begins as defined. When processing is done, the operator is visually informed about the results. The operator can see mistakes on the mastic lines that can’t be seen by eyes. Thus, applications that out of tolerances can be determined.

In this study, a novel technique was developed to solve ergonomic problems that were encountered in a robotic production line. Ergonomic is essential in production lines both for physical health of operators and for the improvement of the cycle time. In this process of tailgate in the robotic production line, the operator in the field bends over more than it should. This situation eventually impairs the health of the operator. The fixture must be horizontal for this sealing process, and it must have a safe angle for the operator. The operator starts the process after positioning a part and a sealer robot starts sealing process. This ergonomic problem was solved by designing a piston and bearing mechanism. The main problem is the positioning of the piston in the design. For this, position of the piston was calculated and determined. Then an appropriate CAD design was developed. Designed parts were manufactured and assembled with standard equipment. The ergonomic analysis was successfully completed and assembled to the production process.

In today's world artificial intelligence studies are applied in many sectors. There are companies in the insurance field that use data-driven and innovative technologies. The insurance industry depends on the ability to be competitive in the market with the right analysis of customer data and the right products. The insurance industry is a challenging area where customer data needs to be analyzed correctly. Artificial intelligence-supported analysis of customer behavior provides a competitive and more profitable structure. All steps from actuarial calculation to customer claim transactions can become much more effective with artificial intelligence. In the research, artificial intelligence products used by today's insurance companies were examined. An alternative model based on real practice data, much more detailed than other studies in the literature, is recommended in the research. The research offers innovative approaches to artificial intelligence and digital transformation studies.

The rapid increase in electrical appliances and systems today means an increasing load on the main power grid. This rising will increase much faster in the near future, since especially electric vehicle load can create a local or global peak load on the grid. This will cause power quality problems such as frequency or voltage instabilities on the main power grid. However, topologies that will reduce the load that the electric vehicle will create on the grid have been suggested by researchers recently. To solve this problem, it is recommended to use the batteries of electric vehicles connected to the grid. In this context, the operation of V2G (Vehicle-to-Grid), V2H (Vehicle-to-Home) or V2V (Vehicle-to-Vehicle) topologies using vehicle batteries has begun to be seen as an opportunity. The most critical issue in the operation of these topologies is to control the energy flow in a coordinated manner, considering the losses. For this reason, a smart charge control algorithm is needed for the operation of this charge topologies. In this study, a rule-based control algorithm has been developed for these topologies, which also considering user satisfaction. This developed algorithm was simulated in real time according to a scenario prepared in the Matlab/Simulink program and the results were analyzed.

There is a remarkable increase in the number of electric vehicles (EV) with the increase in the demand for renewable energy sources. The integration of EVs into the grid has become an important issue with the widespread use of EVs. The grid integration of EVs has detrimental effects on power quality. The charging topologies such as vehicle-to-grid energy transfer (V2G), grid-to-vehicle energy transfer (G2V), vehicle-to-vehicle energy transfer (V2V) have been developed in order to overcome this problem. In this study, a microgrid using V2G and G2V topologies has been designed for a building and the EVs in this building. In this designed microgrid, 120 kw energy flow is provided for the loads in the building and the charging of EVs. When an extra load is added to the grid in the building, the energy above 120 kw is supplied from the EVs in the microgrid (V2G) to support the grid. The LCL filter design has been carried out for the grid connected inverter used in the V2G topology in the microgrid. After determining the output power, switching frequency, busbar voltage, etc. values of the three-phase inverter for the designed V2G topology, the LCL filter parameters have been calculated. The total harmonic distortion (THD) has been determined according to the calculated parameter values. It has been observed that the THD performance of the LCL filter is better than the THD performance of the LC and L filters. The designed microgrid simulations and LCL filter analyses have been carried out in the MATLAB 2020b Simulink program. With the simulations, G2V and V2G topologies have been analysed and LCL filter design has been carried out for the grid connected inverter used in the V2G topology.

Although aviation has been hit hard by the pandemic in the last years, air transport is expected to continue to increase, even if some companies go into crisis. Although some companies went bankrupt during the pandemic, new companies were established, and the planes in the bankrupt companies started to fly in other airlines. The number of employees in the aviation industry is also increasing. Road, maritime, and rail transport cannot keep up with the increase in air transport. By 2037, the annual passenger capacity will be predicted to exceed 8 billion or even 9 billion. There are unique airline connections between more than 20 thousand city-pairs worldwide. Despite this, many passengers fly via transit as there are no direct flights from many cities to many cities or because direct flights are sometimes not economical or common. Since, naturally, passengers do not like to wait in transfers, passengers are lost to other flights and airline companies if the airline does not have a suitable flight at the appropriate time for the transfer. Therefore, synchronization and optimization of international and domestic flights are important at major hub airports and major airlines. In this study, we will try to solve this problem by using the genetic algorithm to maximize the number of total passengers for Istanbul Airport, an important global hub airport. The genetic algorithm is a search heuristic that was inspired by Charles Darwin’s theory of natural evolution. This algorithm reflects the process of natural selection, where the fittest individuals are selected for reproduction to produce offspring of the next generation.

The transition towards Generative Artificial Intelligence (GAI) is rapidly transforming the digital realm and providing new avenues for creativity for all humanity. In the past two years, several generative models have disrupted worldwide, including ChatGPT and DALL-E 2, developed by OpenAI, which are currently receiving significant media attention. These models can generate new content, respond to prompts, and automatically create new images and videos. Nevertheless, despite this progress of GAI, research into its application in business and industry is still in its infancy. Generative AI is bringing ground-breaking innovations that go beyond the limitations of conventional Contextual AI. This new type of AI can generate novel patterns in human-like creativity, encompassing various forms of content such as text, images, and media. It transforms how people communicate, create, and share content, taking organizations by surprise. Unfortunately, these organizations were not fully prepared as they were focused on the advancements and impacts of Contextual AI. Given the significant organizational-societal opportunities and challenges posed by generative models, it is crucial to comprehend their ramifications. However, the excessive hype surrounding GAI currently makes it difficult to determine how organizations can effectively utilize and regulate these powerful algorithms. In research, the primary question is how organizations can manage the intersection of human creativity and machine creativity, and how can they leverage this intersection to their advantage? To address this question and mitigate concerns related to it, a comprehensive understanding of GAI is essential. Therefore, this paper aims to provide technical insights into this paradigm and analyze its potential, opportunities, and constraints for business and industrial research.

Due to the increasing number of disasters in the world, the number of studies in the field of disasters is increasing. The planning, implementation, management and coordination of disaster management activities are very important. Search and rescue, humanitarian assistance, evacuation operations, etc. special infrastructures must be provided to manage situations. One of the most important issues in disaster management is establishing due diligence immediately after the disaster occurs. Quick detection of debris, especially during earthquakes, is extremely important and necessary to reduce the number of casualties. As infrastructures such as the internet, telephone and power lines are damaged in major disasters such as earthquakes, due diligence becomes difficult. Identifying wreckage locations is essential to support search and rescue efforts. In this study, an arc routing problem is considered to determine the condition of buildings in a region affected by an earthquake. The objective is to determine the locations as quickly as possible by checking every road and path in the disaster area at least once. For disaster management, however, it is of great importance to obtain a quick solution to arc routing problems rather than optimal results. Therefore, the solution was sought by the heuristic method of nearest neighbor search, which is widely used in the literature, and the results were recorded.

Process planning, scheduling, and due date assignment functions are the three fundamental manufacturing functions. Traditionally, these functions were examined independently in production systems. The integrated communication of these functions is one of the most efficient ways to ensure high customer satisfaction, nonetheless, in the technological and competitive climate of today. Although these functions have been integrated in academic research over the past few decades, in practice, they are still generally performed sequentially and independently. Although limited research exists that integrates the three functions, this study introduces delivery as a fourth function. The objectives of this study are to make a significant contribution to the literature by demonstrating the integrated nature of four functions in manufacturing systems, with a view to increasing efficiency compared to traditional solutions. The study also seeks to investigate the impact of incorporating the delivery function. Customers are not viewed as being equal, as is the case with many of the other integrated studies that can be found in the literature; rather, each client is given special consideration. Each of the four job shops is unique and has a different number and location of customers. The study solves the complex problem by utilizing simulated annealing and evolutionary strategies algorithms. Both method-based and job shop-based comparisons are made, and the results show which methods perform better in each job shop. The results demonstrate that the integrated system offers an improvement of approximately 50% compared to independent systems. Furthermore, the study found that, across all four job shops, the evolutionary strategies (ES) outperformed simulated annealing (SA) in terms of results.

The study’s main objective is to develop a ROS compatible local planner and controller for autonomous mobile robots based on reinforcement learning. Reinforcement learning based local planner and controller differs from classical linear or nonlinear deterministic control approaches using flexibility on newly encountered conditions and model free learning process. Two different reinforcement learning approaches are utilized in the study, namely Q-Learning and DQN, which are then compared with deterministic local planners such as TEB and DWA. Q-Learning agent is trained by positive reward on reaching goal point and negative reward on colliding obstacles or reaching the outer limits of the restricted movable area. The Q-Learning approach can reach an acceptable behaviour at around 70000 episodes, where the long training times are related to large state space that Q-Learning cannot handle well. The second employed DQN method can handle this large state space more easily, as an acceptable behaviour is reached around 7000 episodes, enabling the model to include the global path as a secondary measure for reward. Both models assume the map is fully or partially known and both models are supplied with a global plan that does not aware of the obstacle ahead. Both methods are expected to learn the required speed controls to be able to reach the goal point as soon as possible, avoiding the obstacles. Promising results from the study reflect the possibility of a more generic local planner that can consume in-between waypoints on the global path, even in dynamic environments, based on reinforcement learning.

Compliance with delivery times is crucial for businesses in the logistics sector. Numerous research has been conducted to improve distribution performance. Many of these studies touch on lean production as well. The strategies used in lean manufacturing are often employed by businesses and have a positive impact on performance.This study focuses on the overseas shipping department of a textile company’s logistics center. Workflow starts with product acceptance from manufacturers and ends with shipment to customers abroad. After a thorough examination, some bottlenecks that increase delivery times are observed.Value Stream Mapping (VSM), which is a lean manufacturing technique, is chosen as the main method to be used. It aims to determine value added and non-value-added activities, resulting in minimizing or eliminating the non-value-added ones.Initially, necessary data are gathered through workshops and interviews, and observations on Current State VSM are made. During these workshops, various improvements are proposed and evaluated together with the company’s engineers. After takt time and cycle time calculations, label change station is identified as the bottleneck.In the next step, Kaizens are suggested for the stations, and some lean techniques are employed to solve different workflow problems. Finally, short-term applicability of proposed improvements is discussed, and Future State VSM is drawn.It can be concluded that significant improvements are achieved especially in lead time, changeover time, productivity rate and production speed. By reducing or eliminating non-value-added activities and identifying deficiencies that slow process flow, a standard, sustainable and developable process is proposed to the company.

The mass production of identical products with high precision and accuracy heavily relies on dies and molds. In particular, the project-based die manufacturing process is crucial for introducing new products to the market. To produce a car body, an average of 1000–1250 unique dies are required, with no backups available. Simultaneously commissioning these dies is one of the most expensive and critical processes in automotive and other forming industries. However, with increasing technology and customer expectations, designers must design products quickly and efficiently to beat competitors to the market. Traditional production planning software is not well-suited for project-based work, such as die manufacturing, and often results in increased project duration due to a lack of integration and variability. To address this issue, this study proposes the use of Robotic Process Automation (RPA) with artificial intelligence methods to enable automatic data transfer from design to production. The findings of this study provide recommendations for efficient project management in the single manufacturing industry.

The advent of Industry 4.0, characterized by the integration of digital technology into mechanical and electronic sectors, has led to the development of autonomous vehicles as a notable innovation. Despite their advanced driver assistance systems, these vehicles present potential security vulnerabilities, rendering them susceptible to cyberattacks. To address this, the study emphasized investigating these attack methodologies, underlining the need for robust safeguarding strategies for autonomous vehicles. Existing preventive or detection mechanisms encompass intrusion detection systems for Controller Area Networks and Vehicle-to-Vehicle communication, coupled with AI-driven attack identification. The critical role of artificial intelligence, specifically machine learning and deep learning subdomains, was emphasized, given their ability to dissect vehicular communications for attack detection. In this study, a mini autonomous vehicle served as the test environment, where the network was initially scanned, followed by the execution of Man-in-the-Middle, Deauthentication, DDoS, and Replay attacks. Network traffic was logged across all stages, enabling a comprehensive analysis of the attack impacts. Utilizing these recorded network packets, an AI system was trained to develop an attack detection mechanism. The resultant AI model was tested by transmitting new network packets, and its detection efficiency was subsequently evaluated. The study confirmed successful identification of the attacks, signifying the effectiveness of the AI-based model. Though the focus remained on autonomous vehicles, the study proposes that the derived methodology can be extended to other IoT systems, adhering to the steps delineated herein.

The amalgamation of information technologies and progressive wireless communication systems has profoundly impacted various facets of everyday life, encompassing communication mediums, occupational procedures, and living standards. This evolution, combined with enhanced wireless communication quality, has culminated in an exponential rise in interconnected devices, including domestic appliances, thereby birthing the Internet of Things (IoT) era. This proliferation, facilitated by cloud computing enabling remote device control, concurrently intensifies cybersecurity threats. Traditional Information and Communication Technology (ICT) architectures, characterized by a hub-and-spoke model, are inherently vulnerable to illicit access and Man-in-the-Middle (MITM) intrusions, thereby endangering information confidentiality. Leveraging Artificial Intelligence (AI) can ameliorate this scenario, enhancing threat training and detection capabilities, enabling precise and preemptive attack countermeasures. This research underscores the criticality of addressing the security implications accompanying technological advancements and implementing protective measures. Deploying AI algorithms facilitates efficient passive attack identification and alleviates network device burdens. Specifically, this study scrutinized the ramifications of an MITM attack on the system, emphasizing the detection of this elusive threat using AI. Our findings attest to AI’s efficacy in detecting MITM attacks, promising significant contributions to future cybersecurity research.

The Internet of Things (IoT) encompasses a vast network of interconnected devices, vehicles, appliances, and other items with embedded electronics, software, sensors, and connectivity, allowing them to collect and exchange data. However, the growing number of connected devices raises concerns about IoT cybersecurity. Ensuring the security of sensitive information transmitted by IoT devices is crucial to prevent data breaches and cyberattacks. IoT cybersecurity involves employing various technologies, standards, and best practices, including encryption, firewalls, and multi-factor authentication. Although IoT offers numerous benefits, addressing its security challenges is essential. In this study, a flood attack, a significant threat to IoT devices, was executed to assess the system’s impact. A reference model without the attack was used to analyze network traffic involving single or multiple attackers. To prevent additional load on the operational system, network packets were mirrored via the cloud and transferred to artificial intelligence (AI) and forensic analysis tools in real-time. The study aimed to ensure continuity, a vital aspect of IoT system cybersecurity, by detecting the attacker using AI and analyzing real-time data with forensic analysis tools for continuous network monitoring. Various AI algorithms were evaluated for attacker detection, and the detection process proved successful.

Big Data hype is increasing extremely fast. A few quadrillion bytes of data in various formats are generated almost daily.These data are the subject of extensive data analysis, especially visual data analytics. New applications and technologies for processing and visualizing Big Data are constantly developing and evolving daily. Therefore, being well-informed and up to date with Big Data processing and visualization software is of very high importance, yet inevitable in contemporary data analysis, providing better decisions and solutions for both businesses and managers. This research is based on a thorough analysis of reviews of current Big Data visualization tools made by the world’s most popular authorities. The study’s main objective is to describe a new visualization tool based on algorithms, primarily aiming to improve existing Big Data visualization software capabilities. Considering the visualization aspects of Big Data and the previously noted criteria, we present a custom-made application named DataViz, developed using Python. The DataViz application is simple and easy to use since it has an intuitive user interface to serve various users, including those without enhanced computer skills. Regarding the analysis of current visualization tools, the DataViz application considers and implements several important criteria, including Accuracy, Empowering, Releasing, and Succinct. The development of such an application fills the gap among different commercially available Big Data visualization tools delivering enhanced visualization capabilities and optimization. As such, it can provide a solid basis for further improvement and transformation into a fully functional software tool.

The production of several materials on the same production line is very important for efficiency. Electrified Monorail System (EMS) is a product transport system with rails and this system can reach every production plant in the manufactory. If EMS is designed and used effectively, the production will be fast and efficient. In this study, an existing EMS was improved on a vehicle production line.EMS carries the parts of vehicles to different areas in the factory environment. When the production type varies, then the transportation of the parts will be complicated. In other words, the increasing of the production variant makes the operation of the system be difficult. EMS lines are needed to develop solving transportation problems that are mismatch buffer, new lines design, flexible using of the carriers of the body parts and smart flexible automation. In this project, an enhanced EMS study related to flow algorithm of the production line was carried out to increase the efficiency of the manufacturing. Thus, the motion algorithms and coordination algorithms of the system were developed. Carriers that adapt flexible manufacturing system were designed to carry different car parts.

The Three-Index Assignment Problem (3-AP) is well-known combinatorial optimization problem which has been shown to be NP-hard. Since it is very difficult to find the best result in polynomial time, many heuristic methods have been proposed to obtain near optimal solutions in reasonable time. In this paper, a modified Bacterial Foraging Optimization Algorithm (BFOA) is proposed to solve 3-AP. BFOA is inspired by the social foraging behaviour of Escherichia coli (Ecoli). Algorithm imitates the behaviour of the foraging bacteria Ecoli and aims to eliminate those bacteria that have weak foraging methods and maintaining those bacteria that have strong foraging methods. The Hungarian method (most known method for solving the classical linear two-dimensional assignment problem) is integrated to BFOA algorithm at repositioning phase to swim farther and faster to find the best solution. Proposed algorithm has been tested and benchmarked with other algorithms in literature and results show that the new algorithm outperforms other heuristics in literature in terms of solution quality.

In this study, the importance of quality management practices in achievement of operational results and customer satisfaction in logistics is handled. The objective of this study is to propose a quality management framework through multi criteria decision making (MCDM) based on the criteria utilized in the European Foundation for Quality Management (EFQM) for assessment of suppliers. In this regard, this study proposes EFQM based integrated methodology using Analytic Hierarchy Process (AHP) and Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS) method for supplier selection problem.

Deep learning is a machine learning approach that has been widely used in many different fields in recent years. It is used in agriculture for various purposes, such as product classification and diagnosis of agricultural diseases. In this study, we propose a deep-learning model for the classification of rice species. Rice is an agricultural product that is widely consumed in Turkey as well as in the world. In our study, a rice data set that contains 7 morphological features obtained by using 3810 rice grains belonging to two species is used. Our model consists of three hidden layers and two dropouts (3H2D) added to these layers to prevent overfitting in classification. The success of the model is compared with Logistic Regression (LR), Multilayer Perceptron (MLP), Support Vector Machine (SVM), Decision Tree (DT), Random Forest (RF), Naïve Bayes (NB), K-Nearest Neighbors (KNN), Ada Boost (AB), Bagging (BG), and Voting (VT) classifiers. The success rates of these methods are as follows: 93.02%, 92.86%, 92.83%, 92.49%, 92.39%, 91.71%, 88.58%, 92.34%, 91.68%, and 90.52% respectively. The success rate of the proposed method is 94.09%. According to the results obtained, the proposed method is more successful than all of these machine learning methods.

Today, the use of heterogeneous robot teams is increasing in military operations and monitoring the environment. However, single 2D autonomous navigation systems fail in rough terrains. In this study, a solution to this problem is proposed using 2.5D navigation that takes into account the slopes on the elevation map. Using ROS and Gazebo, we coordinate drones and ground vehicles to process terrain elevations. The simulation world used in the study reflects a real-world rough terrain and also some urban artifacts are added to the simulation world. Husky simulation model is used as the ground vehicle utilizing 3D LIDAR, GPS and 9 DOF IMU sensors, which can output 3D map and 3D localization using 3D SLAM. Using the SLAM localization, a 2.5D map is created on the ground vehicle. Drone simulation model, similarly equipped, follows the ground vehicle with a GPS-based waypoint navigation and can create a 2.5D map using its sensors. A global plan is created for the ground vehicle by cooperative effort of both robots, using the map information from ground vehicle where available and using the map information from the drone where ground vehicle’s map is insufficient. 2.5D navigation of the ground vehicle is carried out by the local planner taking into account the calculated cooperative global path. Proposed method results in shorter routes and fewer path planning issues. This is shown by the comparative analysis where the ground vehicle or the drone is used alone.

The demand for renewable energy to mitigate CO2 emissions is continuously growing due to the risk of environmental degradation. In this regard, agrivoltaic systems emerged to provide a potential solution for environmental deterioration. Agrivoltaic systems are the common use of land for crop and energy production. These systems are designed to efficiently share the same land between photovoltaics and agriculture so that food and energy can be produced conjointly. In this study, we investigate the potential of agrivoltaic systems in Turkey. With this aim, three provinces with the highest agricultural production are selected, and their capabilities are investigated by using GIS (Geographic Information System) with a fuzzy analytical hierarchy process (FAHP) method. First, the literature is reviewed to determine the criteria for agrivoltaics installation. Then, criteria weights are determined using FAHP in order to capture the associated uncertainty in decision-maker’s (DMs) judgments and preferences. Finally, suitability analysis is conducted using the criteria weights and the GIS data to reveal potential sites for agrivoltaics development. The study results contribute to practitioners deciding the target locations to sustainably generate solar energy while allowing crop production.

Logistics has become one of the most significant parts of the process in many business areas. For an efficient logistics system, each stage of the operation needs to be designed carefully. Logistics approaches are also applied in the financial sector such as Automatic Teller Machines (ATM) cash management. ATMs provide efficient service for a financial institution to its customers through a self-service, time-independent, and simple-to-use mechanism. For daily financial transactions, ATM is the fastest, safest, and most practical banking tool. Many challenges have risen in the network design of the cash and these problems can be solved by using the optimized solution. This solution aims to satisfy the customer at the ATM and at the same time, minimize loss for banks. This paper states the solution for the replenishment of ATMs. Firstly, data is analyzed by using different policies from several approaches after then an efficient metric system is applied to compare the results of it. In the end, the method selected has the appropriate results according to metric. Furthermore, to avoid bottlenecks and become quicker in the procedures, inventory control connects the supply of cash and customer demand in the ATMs. The replenishment policy starts with forecasting cash withdrawals by applying various methods such as statical methodologies (ARIMA and SARIMA) and machine learning methods (Prophet and DNN). By creating a decision support system, several methods are applied in order to visit ATMs by using different inventory control methodologies.

Human involvement in manufacturing is indispensable. The performance and reliability of the human element directly impacts the success of a system. The complex interplay between fatigue, workload, and human performance necessitates recognition of the importance of human factors in manufacturing to achieve optimal outcomes. The efficiency of employees in production systems can be influenced by a number of factors, including exhaustion, cognitive and physical abilities, and the amount of time available for tasks. In general, these factors can have an impact on human reliability, which in turn can affect the reliability of production systems. A key objective of workload research is to identify and analyze potential performance declines, along with their associated factors. The pursuit of successful outcomes in an operation is not enough, as the negative impacts of such outcomes on the system or personnel must also be minimized. Once a possible issue or decrease has been determined, corrective measures are put in place to address such instances. Such foresight serves to prevent the costs that may arise from making process modifications, and to optimize resource utilization. This study conducts a comprehensive review of related literature, focusing on simulation integration. A taxonomy of fatigue and workload is used to contextualize and assess these factors. Methodologies and techniques are described, and evaluations of specific assessment techniques are provided, along with discussions of available information. This research makes a persuasive case for the importance of evaluating fatigue and workload among operators.

Social media usage has significantly increased, allowing people to freely express their wishes and complaints through these platforms. Consequently, this has caused a significant increase in data, providing information about users, companies, services, and products. However, making sense of this data with human effort is impossible, necessitating various methods. Sentiment analysis is one such method that helps us understand customers’ thoughts on products and companies. In this study, the emotions of e-commerce site users were analyzed using Turkish Twitter data. Text mining techniques were applied to Twitter data, which were analyzed and classified as positive and negative. It was found that 66.9% of the tweets were negative, and 33.1% were positive. The classification results were evaluated using precision, Recall, and F1 criteria. As a result of the evaluation, the sensitivity criterion for negative comments was 94%, and the precision criterion for positive comments showed the highest value with 86%. When looking at the F1 score, 85% for negative comments and 69% for positive comments were calculated. The accuracy rate of the model was found to be 79%.

Today, scarce resources or many unpredictable factors such as demand have increased the importance of the supply chain. The motivation of this study is the need for the design and analysis of the dynamic supply chain model that will shed light on the companies. In the study, four different dynamic nonlinear supply chain models, which will be an example for companies to reveal their structures, are summarized and a new chaotic supply chain dynamic model developed for citrus production from perishable products, which has not yet been studied in the literature, is presented. The chaotic structure of this new model is demonstrated with time series, phase portraits, bifurcation diagrams, and Lyapunov exponents. In addition, with the active control technique, in which control parameters are added to all the equations of the supply chain system, the chaotic structure of the system was brought under control and synchronous operation was ensured with a different system. Thus, the production amount, demand, and stock data, which are the supply chain status variables of a company’s factories in a different area, can have similar values with an error close to zero.

In recent years, digitalization has become widespread in many fields, including manufacturing systems. Following the deployment of Industry 4.0 technologies in manufacturing, significant improvements have been noted in many production areas, such as predictive maintenance, production planning, dynamic scheduling, and more. The digital twin is one of the technologies used in this area. The objective of this paper is to develop an automated and executable digital twin process within the concept of industrial artificial intelligence. For this purpose, hyperparameter optimization is used, a method to find the best external parameters that should be set by machine learning algorithm developers. It is used to automate the process and increase its accuracy by selecting the optimal parameters that are appropriate for the dataset. In this study, a digital twin was created using the random forest method on a dataset obtained from a CNC machine for a predictive maintenance application. Hyperparameter optimization is integrated into the machine learning process and the random forest hyperparameters such as bootstrap, maximum depth, maximum features, minimum sample leaf, minimum sample split, and the number of estimators is optimized. This study is a digital twin application under the concept of industrial artificial intelligence. A machine learning algorithm with integrated hyperparameter optimization is proposed, which is more efficient and faster in the field of predictive maintenance in production. For this purpose, the Hyperopt library in the Python programming language is used.

Work-life balance is an approach that aims to enable employees to balance their work, family, and private lives. It is seen that the factors in the work-life balance are not relevant to work and family, considering the activities that one wishes for oneself, friendships and social life. For this reason, it has become mandatory for working people to devote enough time to their business life, family life and private life to protect their physical and mental health. This is particularly important in health-care sector, where the peace of mind of workers influences the outcoming service significantly. Within the scope of this study, a framework (also implemented as a software) for health-care workers has been developed in order to make weekly and monthly scheduling suitable for work-life balance. Three population-based heuristic algorithms for scheduling, namely genetic algorithm, ant colony and particle swarm optimization algorithms, are integrated into the system to optimize the schedules. The system aims to show optimal schedules for both which personnel is to work in which time zones of the hospital administrators and in which time zones the individual personnel is planned to work. The proposed approach allows doctors, as the most flexible workers, to opt the working hours and periods in a hospital flexibly. This study provides comparative results to demonstrate the performance of the three algorithms integrated to optimize the generated personal schedules optimized with respect to one's own preferences. Furthermore, it identifies the boundaries of the parameters affecting the success with Taguchi Method.

According to estimates, the world population is expected to be around 9 billion by 2050. Related to this, the number of people suffering from hunger is increasing day by day. Unconscious use of agricultural lands, climate changes, and the effect of increasing population, the problem of food needs increase the pressure on agriculture. In order to meet the need for food effectively, studies in areas such as sustainable land/forest management, improvement of cultivation areas, and agricultural policies should be carried out urgently. Increasing productivity by improving agricultural activities is important in terms of realizing the potential of agricultural lands.In this study, a plant production plan is created according to the relevant data on the agricultural lands in a city. The aim is to create a sustainable production plan and to present a model that will increase the economic return that maximizes the return that was established. In the study, a mixed-integer model that maximizes the return is developed and it is aimed to determine which plants will grow in which region and to make a production plan accordingly. The distinctive value of the study is to propose a new decision support system for planning studies in Turkey in the field of agricultural production. As a result of the success of the project in the desired way, a system that will help the decision-makers in the relevant field will be gained, as well as contribute to the literature.

The evolution of science and technology has led to increasingly complex cyber security threats, with advanced evasion techniques and encrypted communication channels making attacks harder to detect. While encryption has improved privacy and confidentiality for users, it has also provided a new avenue for attackers to exploit. Traditional intrusion detection systems, which transitioned from signature-based to behavior-based approaches, have struggled to keep up with these challenges. To address this issue, researchers have turned to continuous system monitoring and network traffic packet analysis. However, this method can be resource-intensive and time-consuming, particularly when analyzing encrypted packets. In this study, the JA3 fingerprint infrastructure was examined as a potential solution for quickly detecting attacks conducted over encrypted sessions while minimizing system downtime and damage. The results demonstrated that the JA3 infrastructure effectively detected attacks carried out via encrypted channels. Although Windows 10 and Kali 2020.4 operating systems were used as the victim and attacker systems respectively, the methodology can be applied to other operating systems and network hardware by following the outlined steps. This research is expected to make a significant contribution to the field of encryption-based attack prevention.

With the rise of carbon footprint awareness and the circular economy, the need to improve supply chain traceability and visibility has increased. Traditionally, companies have used database-driven platforms to improve traceability, reuse, and reduce waste. In practice, lateral transshipment among stakeholders also aims to improve reuse and reduce waste by redistributing unsold products. Although the optimization issues are well-studied in this field, there are many challenges in sharing data about ingredients, processes, delivery stages, and the lifecycle of a product. Manufacturers, retailers, and consumers have only limited access to this information. Blockchain technology can help to manage the complexities of managing a circular supply chain, especially in the context of lateral transshipment. This study presents a framework for blockchain and smart contract-enabled lateral transshipment among fast fashion retailers. There are three main components of the framework: inventory management system by the organizer, introducing retailers’ excess inventory and orders into the inventory sharing pool, delivery with execution of smart contracts. The proposed framework not only contributes to the visibility and traceability of the system but also increases responsiveness, flexibility and efficiency.

This study focuses on predictive maintenance, a critical maintenance policy that benefits from the development of the Digital Twin (DT) philosophy. To implement predictive maintenance, it is essential to predict potential failures. In this study, machine learning algorithms are used to detect failure conditions. Five different types of failures are classified by examining parameters such as air temperature, process temperature, rotation speed, torque, and tool wear. The study utilizes Automatic Machine Learning (AutoML), which runs machine learning algorithms and returns the best method, its hyperparameters, and many outputs, such as accuracy and performance metrics. The literature on machine learning algorithms in predictive maintenance has focused on finding the best algorithm by applying selected methods. However, this study aims to contribute to the literature by finding the algorithm that provides the best results among all methods using AutoML in predictive maintenance.

Simulation technology is being used to reduce the costs of development and testing processes in autonomous driving studies. Autonomous systems gaining driving experience in simulation is faster and more cost-effective than real-world work. However, systems developed through simulation are becoming increasingly distant from reality, and it is uncertain how these systems will respond to situations that may occur in the real world. Therefore, having a simulation environment that is close to reality in which an autonomous driving system will be developed will contribute to the more efficient operation of the systems developed in the simulation environment in the real world. To create a more realistic driving simulation, a large amount of driving data from the real world is needed. In this study, a smart system has been developed that analyzes various driving videos and produces the necessary data for the simulation. The proposed system calculates the position change of the vehicles in each frame of the video. Depending on the position change, the vehicle's speed, acceleration, total displacement, and maneuvering style are revealed. This allows for data on different driving styles to be obtained, transferred to the simulation environment, and enables the simulated vehicles to behave like real drivers. In addition, the vehicles and stationary objects in the video are detected by an intelligent system, and the vehicles in consecutive frames are matched and scaled to made frames identical. By this way we eliminate the effects of extrinsic parameters such as camera transformation, angle change, and zoom in/zoom out to produce the required data.

A stochastic bilevel model for an industrial symbiosis (IS) network in an eco-industrial park (EIP) is presented in this paper. Companies in IS networks produce final products using mostly by-products rather than raw materials. Companies in these networks are known to share a variety of synergies, including wastes and by-products. They mostly negotiate on the price and amount of the by-products. This negotiation is critical as by-products should be available on time at the right amount for smooth production. Moreover, the use of these by-products should provide savings compared to the use of raw materials so that companies could benefit from IS. However, the fluctuations in the final product demand affect both the EIP management (the leader) and companies (followers) in the network. Therefore, we proposed a stochastic bilevel model where the lower-level problem is a two-stage stochastic programming model. Our findings show that the demand uncertainty has an impact on both parties.

The implementation of Industry 4.0 enabling technologies to optimize supply chain processes has become an essential aspect of modern manufacturing. Supply chain optimization (SCO) through additive manufacturing technology (AMT) is one area where Industry 4.0 is having a considerable impact. This study, therefore, explores the aspects of Industry 4.0 enabled by the incorporation of AMT into the production system that contribute to supply chain optimization. Ten such features have been identified and analyzed in this study using the novel Grey Influence analysis (GINA) technique. GINA methodology in this research helps to circumvent the restrictions of contemporary causal modeling methods such as the Decision-making trial and evaluation laboratory (DEMATEL) and Interpretative structural modeling (ISM) with the cumulative integration of responses, thus mitigating data loss. The results of the investigation demonstrate that cloud manufacturing, sustainable manufacturing, on-demand manufacturing, and distributed manufacturing are the predominant features of Industry 4.0 enabled by AMT. The findings of this study can be utilized by businesses to take informed decisions concerning incorporation, investment, and optimal application of AMT as well as to determine the significance of AMT to optimize their supply chains.

Project management is a critical process for every institution and/or organization. This process should be managed as best as possible in order to manage resources and time well and at the same time achieve successful results. One of the reasons that make project management difficult is the increase in project proposals with the increase in reading rates and incentives. The evaluation process of the project proposals includes the assignment of an expert who will evaluate the project. This stage is solved with reviewer assignment problems in the literature. When the literature is examined, it is seen that the general aim of reviewer assignment problems is to maximize the degree of reviewer-project match. This study, in addition to the literature, it is aimed to minimize the evaluation time of the reviewers’ project proposals and to ensure a balanced distribution among the reviewers. When we look at the results of the test problems for this study, which has two objectives, maximum matching degree and minimum evaluation time, it is seen that the objectives have been met. In this way, the assignment, which was made manually and caused a waste of time, was completed in a fair and reliable way.

In today’s technology age, digitalization, which aims to enable institutions to carry out their processes independently of time and space, can be briefly summarized as transferring physically carried out processes to electronic environment, monitoring and managing from there. In this study, a model has been proposed for the Student Affairs Unit, which operates within the Faculty of Economics and Administrative Sciences of Dokuz Eylül University, to move its communication with the students to the digital environment. The proposed model is web-based and responsive, which requires user authentication for the current students. The model includes two different functions. The first function is to create and categorize the frequently answered questions (FAQs) for the faculty and present them to the students with a user-friendly design. The other one includes the process where the student can submit a request as a ticket for the subject to consult for the situations other than the FAQs, and the request is answered by the relevant personnel in the student affairs unit. The proposed model provides business analytics information in the form of response times for requests, requests according to departments, and requests according to categories, in order to be beneficial for Managers as well.

In Industry 5.0, humans and machines work together, using advanced technologies like Artificial Intelligence (AI), the Internet of Things (IoT), and automation to improve efficiency, productivity, and quality while also supporting sustainable practices and human values. There is a growing interest in learning about the challenges of Industry 5.0 and exploring these technologies to promote sustainability and responsible business practices. We need a hybrid decision model to strike a balance between technical progress, human values, and sustainable practices as we move toward Industry 5.0, which presents enormous challenges in the areas of technology, the environment, society and ethics, and business and economics. Through a literature analysis guided by the PRISMA technique and the Delphi method, the study highlighted challenges in the areas of technology, the environment, society and ethics, and business and economics, as well as solution measures to address them. The weightage of the challenges was determined using the Best Worst Method, and the ranking of the potential solutions was prioritized using the Elimination and Choice Expressing Reality method.

In this study, instance selection was made using the fuzzy rough set based instance selection method, which is the main indicator of heart disease risk, with the finding of a certain narrowed main cardiovascular number and some other medical findings. Then, with the help of machine learning algorithms, a heart disease risk estimation model was developed over two different size and structure data sets. The heart disease dataset, formed by combining 5 different heart disease datasets, was taken from the IEEE dataport website [1]. In order to eliminate noisy instances, the 12-variable data of 1190 patients was reduced to 836 instances by fuzzy rough set based instance selection method. Qualitative variables used in the analysis are age, sex, chest pain type, resting bps, cholesterol, fasting blood sugar, resting ecg results, maximum heart rate, exercise induced angina, oldpeak, ST slope. Then, the data set was divided into two as 70% training and 30% test data sets, two-class averaged perceptron, two-class Bayes point machine, two-class logistic regression, two-class support vector machine, two-class neural network, two-class locally deep support vector machine and two-class boosted decision tree models were trained. As a result of the validity analysis carried out, the use of fuzzy rough set based instance selection method improved the prediction performance of all models. While the Two-Class Boosted Decision Tree method gave a higher accuracy than other methods, it gave an accuracy result between 89% and 93% in other methods.

In this study, olive seed activated carbon (OSAC), which is primarily agricultural waste, was used as an adsorbent. Response surface methodology (RSM) approach and Artificial Neural Network (ANN) was used to optimize/model the cationic dyestuff removal by adsorption. RSM was first applied to evaluate the process using four controllable operating parameters, namely the amount of OSAC, initial pH (pH initial), mixing time and dyestuff concentration, and optimal conditions for decolorization were determined. In the optimization method, initial dye concentration (50–150 mg/L), adsorbent dosage (0.1–0.5 g), pH (3–9) and contact time (10–90 min) were used as independent variables, percent removal efficiency was chosen as the dependent variable. Also, the value of R2 (R2 = 0.9714) shows that regression can predict the response for the adsorption process in the studied range. It shows that it is possible to optimize/estimate the dyestuff removal process using the RSM approach and shows that adsorption to OSAC can be used to remove colour from dye effluent.

The Balanced Scorecard (BSC) performance system application has been generally accepted and is frequently applied in evaluating and measuring company performance. This performance evaluation system envisages a system in which the company’s performance is assessed in four dimensions, the indicators determined according to these four dimensions are weighted, and the performance calculation is made consequently. It is a performance system in which the performance of each dimension is independently examined and evaluated.In this study, the “red area performance evaluation system” applied to Kayalar Pres company, which produces installations and fasteners, is modeled with fuzzy logic. The results are compared with the actual application results.The application of this study in a performance system with actual data enables the accuracy of the performance indicator weights determined by the management to be compared. The data obtained on the performance system were formed as fuzzy sets based on expert opinions, and a fuzzy set matrix with 81 rules was developed with these sets. The created fuzzy set membership rules were made for all four indicators, and the “company performance score” was calculated. The 81 rules and output results determined in this direction were defined in the MATLAB fuzzy logic function, and the system was modeled on MATLAB accordingly. The results of the fuzzy logic function were obtained and interpreted by comparing the actual measurement results of the company.

Today’s brutal competition environment has made it a necessity for businesses to evaluate their performance in order to maintain their existence and gain sustainable competitive advantage. The current environment does not allow businesses that do not perform performance evaluations a chance to survive. Therefore, corporate performance evaluation is a multi-criteria, complex and uncertain real-life problem that is vital for businesses. Corporate performance should be measured multi-dimensionally, but performance indicators of these dimensions may not always be expressed with a numerical value or may have uncertainty. In this case, the closest results are obtained by using fuzzy logic. Based on the current importance of the subject, a fuzzy model has been developed for the evaluation of corporate performance in this study. The outputs of the model are total enterprise performance, corporate reputation and financial output. Although the number of inputs used for each output varies, the inputs used in the model are the environment, customers, society, government, competitors, suppliers, business processes and sustainability. The application of the model is demonstrated with a case study using the fuzzy logic toolbox of the MATLAB program. Performance values for “Total Enterprise Performance”, “Corporate Reputation” and “Financial Output” were obtained as 16.9, 16.9 and 20.8, respectively.

This paper presents a study on the application of reverse engineering in electroless coatings, specifically in the optimization of bath parameters for the production of user-defined Ni–B-P coatings. Electroless coatings are widely used for their unique properties such as hardness, thickness, and corrosion resistance. However, the optimization of bath parameters for producing coatings with specific properties can be a complex and time-consuming process. In this study, reverse engineering is used to analyze the structure and properties of existing coatings and to determine the optimal bath parameters for the production of user-defined Ni–B-P coatings. The results show that the use of reverse engineering in the optimization of bath parameters can significantly reduce the time and cost associated with the development of coatings with specific properties. This study provides valuable insights into the application of reverse engineering in the optimization of electroless coatings and demonstrates its potential for enhancing the performance and functionality of coatings.

Inflation is caused by the growing gap between the amount of money actively involved and the sum of products and services available for purchase. It is an economic and monetary process that manifests itself as a constant rise in prices, a fall in the current value of money. Inflation is a subject that keeps itself constantly updated in our country and around the world. The main purpose of the central banks, which are dependent on countries in the world and continue their activities, on the economy is to ensure price stability permanently. In recent years, artificial intelligence techniques have been used more and more in order to consistently predict the value of inflation in the future and to make future studies with the forecasts obtained. The aim of this study is to estimate inflation in the Turkish economy with time series analysis by using LSTM (Long Short Term Memory) model, which is one of the artificial neural networks types, on a python computer program. With this study, the estimation made by the LSTM model showed result when compared in terms of MAPE and MSE statistical analyses. It has been observed that the irregular increase in the inflation value within the country in the recent periods directly affects the success level of the models.

Computer vision applications are a branch of science that aims to extract meaning from digital images. It is used in many fields such as automatic inspection, quality control, security, efficiency improvement, autonomous vehicle technology and health sector. Especially in quality control applications, the reliability of product measurement processes with traditional methods depends on the human factor. This study aimed to measure the product dimensions automatically and accurately with computer vision for the shower tray products of a company operating in the ceramic sanitary ware sector. In addition, low cost, high accuracy and high efficiency are targeted with this study. Object size measurement with computer vision consists of two main components: software and hardware. The choices related to these 2 components can directly affect the results of the application. The application stages can be listed as; calibration of the camera that is fully aligned with the product area, calculation of the pixel/centimeter ratio using the calibrated camera and ArUco marker, background segmentation on the collected image data, edge finding applications and measuring the product dimensions on the found edges. As a result of the application, for 79 shower tray images; For 43 data within 70% frame area, average 2.86 mm, for 36 shower tray images outside 70% frame area, average 4.44 mm and for all test data, measurement was made with an error of 3.67 mm on average. This error rate is below the company’s 1 cm tolerance value in manual measurement and gives successful results.

The aim of this study is to draw attention to Circular Economy (CE) approach, to address a future trend and to create a compilation of some important CE studies. In this context, we provide information to guide businesses in the transition to CE and to provide ideas for future studies. This study examines some studies about the concept, structure, principles, business models, benefits, and barriers of CE implementation, moreover differences between Linear Economy (LE) and CE. The failure to solve problems such as climate change, loss of biodiversity, overpopulation, and resource scarcity in the current LE Model has shown that sustainable economy models can be a successful solution. Artificial intelligence (AI), on the other hand, is an emerging technology with enormous potential to impact CE. The use of AI in integrating the Circular Economy into product lifecycle processes was investigated. This study collects many studies examining CE from different perspectives in a single study and also shows the usability areas of AI in applying CE in production. Another contribution of this study to the literature is, proposing the use of AI regarding CE’s the 9R framework. This study has shown that, through artificial intelligence, the stated barriers to the transition to the circular economy can be removed or reduced. By using the business models outlined in the study, using Artificial Intelligence and implementing CE strategies, the stated benefits of the circular economy can begin to be achieved. The study guides the implementation of the CE concept.

Feasibility studies in assessing the viability of renewable energy investments involve conducting economic analyses to evaluate crucial performance criteria such as return on investment and payback period. In these analyses, it is essential to calculate the profitability derived from the electricity generated and sold, considering the investment and operational costs. At this point, electricity prices emerge as a determining parameter. However, accurately predicting electricity prices is challenging due to the need to account for production costs, as well as regulations and policies within the framework of the free market mechanism. Moreover, compared to other traded commodities, electricity prices are further complicated by electricity’s inability to store, the necessity of instantaneous balancing of production and consumption, and the high seasonality of demand for domestic, industrial and commercial electricity. On the other hand, precise forecasting of future electricity prices is a critical factor that enhances the accuracy of economic analyses. This study employs the Prophet Algorithm, which utilizes time series analysis and considers historical electricity prices to make periodic predictions of future electricity prices. The Prophet algorithm is specifically designed to capture the evidence of deterministic trend and seasonality, and at the same time the effects of an econometric shock, providing reliable results in forecasting electricity prices. Our study has examined and compared the forecasting performance of Python’s Prophet Algorithm and Excel Estimator. Although the Excel Estimator did not achieve the same level of precision, it produced results within an acceptable range.

Vehicle routing problems are one of the real-life problems studied extensively in the literature, especially in the logistics and transportation sectors, and consist of various constraints and parameters. Vehicle routing problems, the primary purpose of which is cost minimization, are solved with heuristic or metaheuristic methods within the scope of their content.In this study, the problem is to plan the routes for delivering white goods from a main warehouse to homes or dealers in Ankara and surrounding cities, considering the delivery time window constraint. Deliveries can be made before or after the time window, but if there is a delay, it will incur penalty costs. Therefore, the problem examined is in the class of “Vehicle routing with flexible time windows” problems. The main focus in solving the problem is to minimize cost and deliver within the time window. A two-stage method based on “cluster-first route-second” approach has been proposed. Products to be delivered are divided into two groups regarding the size and product group-based placement constraint added to the DBSCAN clustering method. If there is capacity in the vehicle, the DBSCAN algorithm was revised to include the next point in the cluster. In the second stage, clusters with a high occupancy rate and a minimized number of vehicles are routed to deliver under time window constraints with the Ant Colony Algorithm approach. The results of the study are compared with the previous planning results, financially and operationally. The proposed approach achieved a 30% improvement in the number of vehicles. The vehicle occupancy rates have been increased to an average of 94.89%.