Keywords: Optimization, Logistic planning
Abstract: In this paper a Vehicle Routing Problem of delivering and picking up crews for maintenance and servicing of machines in production systems is researched. Mixed hard-soft time windows and two optimization criteria relating to total travel distance and total weighted tardiness are assumed. A method of efficient computation of the goal function value via disjunctive graphs is presented. Two inexact solving methods: a greedy heuristic and a~Tabu Search local search metaheuristic are proposed. The effectiveness of the methods and the influence of the number of vehicles on solution quality is tested in a computer experiment using a number of problem instances.

Keywords: Equipment health prognostics, Machine learning, Machining process
Abstract: This article presents the application of machine learning methods to predict the surface roughness of materials processed by a milling cutter based on the analysis of measurement signals. An advanced measurement system equipped with vibration sensors, a microphone, and a current transformer was used to collect real-time data during the milling process of 1.7225 steel. The collected data were utilized to develop and compare three predictive models: Neural Network (NN), ElasticNet, and the k-Nearest Neighbours (k-NN) algorithm. The results demonstrated that the NN model achieved the highest prediction precision (MSE = 0.00280), while Elasticnet provided reliable performance in simpler scenarios and k-NN showed the lowest effectiveness. The study confirmed the potential of combining advanced signal processing with machine learning methods to optimise surface quality in machining processes.

Keywords: Process control, Inspection, Machine learning
Abstract: This study investigates the application of machine learning models - linear regression (LR), support vector machines (SVM) and neural networks (NN) - to predict product quality in a candle insert manufacturing process. The research analysed a six-month dataset containing key parameters of the candle insert production process. The NN model achieved the highest predictive accuracy with an RMSE of 0.028, R² of 0.99, and MAPE of 0.1%. The SVM model demonstrated strong performance with an RMSE of 0.038, R² of 0.98, and MAPE of 0.3%, offering a balance between accuracy and computational efficiency. The LR model was less effective, with an RMSE of 0.086, R² of 0.92, and MAPE of 0.5%, struggling to capture complex patterns in the data. Shapley analysis was conducted to determine the impact of each parameter on the models predictions, revealing that temperature before the second casting machine (x8) and filling time (x10) were the most influential parameters.

Keywords: Digital twin, Internet of Things (IoT), Smart system
Abstract: This paper presents the use of digital twins of 2 laboratory conveyor systems in a PLM (Product Lifecycle Management) tool environment, emphasizing the concept of Logistics 4.0. Two models of real conveyor systems were created and analyzed using Tecnomatix Plant Simulation software. The hardware interconnection via PLC and OPC interface creates active digital twins. The first model simulates a simple linear conveyor system with basic automation elements. In contrast, the second model depicts a more complex production line structure incorporating intelligent sensors and dynamic line control. The results show how digital twins enable efficient optimization of logistics processes, simulation of different operational scenarios and prediction of potential problems without disrupting real operations. The paper highlights the benefits of using digital twins in PLM tools, such as increased flexibility, reduced operating costs and real-time decision support. These insights provide a basis for implementing innovative solutions in Logistics 4.0.

Keywords: Decision making, Logistic planning
Abstract: The product as a service (PaaS) business model introduces a paradigm shift in the conventional boundaries of the production system and associated logistics. To enhance circularity, PaaS should be reinforced by value retention processes, such as remanufacturing. Remanufacturing is defined as an industrial process that aims to restore used products to a condition that is indistinguishable from new items. At present, remanufacturing in PaaS in business-to-consumer markets (B2C) is a niche practice. The aim of this paper is to propose a novel decision-making framework with a set of indicators that facilitates the scaling up of remanufacturing in circular business models such as product-as-a-service (PaaS). The proposed framework is created based on a systematic literature, supplemented by expert interviews and a longitudinal case study. The novelty of this paper results from the integration of a circular business model with a remanufacturing process perspective, complemented by a set of indicators that enables improved circularity of EEE producers

Keywords: Product engineering
Abstract: Remanufacturing, a process that restores products to their original functionality. It plays a crucial role in within the circular economy. This study examines the feasibility of remanufacturing washing machine components, focusing on the presence of critical raw materials (CRMs) and the challenges associated with their recovery. The analysis is based on scientific literature and data from household appliance manufacturers. The theoretical analysis is compared with scanning electron microscope analysis (SEM) of selected components. Technical and economic aspects of the remanufacturing process are discussed, highlighting potential environmental benefits and barriers to practical implementation. Elements distribution analysis of selected components revealed the presence of critical materials, including gold (Au), silver (Ag), palladium (Pd), nickel (Ni), tin (Sn), and antimony (Sb). These findings emphasize the need for efficient recovery strategies to enhance resource sustainability and reduce dependence on virgin materials. The study contributes to the ongoing discourse on circular economy practices, offering insights into CRMs recovery in household appliance industry.

Keywords: Human factors, Human interaction, Design interfaces
Abstract: Systems Engineering (SE) plays a crucial role in the design of modern aerospace systems by ensuring the structured development, integration, and validation of complex technologies. Model-Based Systems Engineering (MBSE) enhances this process by enabling a model-driven approach to system design, improving traceability, reducing development time, and facilitating interdisciplinary collaboration. Literature analysis performed in the frame of this paper confirms the increasing adoption of MBSE in aerospace engineering and its benefits in managing system complexity. Job market analysis reveals that employers expect proficiency in MBSE concepts, modelling languages such as SysML, and tools like MATLAB/Simulink and Dymola, along with skills in teamwork, risk analysis, and decision-making. Moreover, a Human-Centric (H-C) approach is gaining importance, as it ensures that system design prioritizes usability, safety, and effective interaction between humans and machines. To address these needs, an interdisciplinary team project was proposed to develop MBSE-related skills. Bearing in mind that MBSE adoption is still developing, the project combines MBSE with H-C Design, fostering both technical expertise and soft skills. The proposed approach requires further validation through practical implementation in educational and industrial contexts; nevertheless it can serve as a model for engineering education, supporting the development of future engineers in the aerospace industry.

Keywords: Enterprise resource planning, Smart maintenance, Production planning
Abstract: Ensuring the reliability of the production process requires, among other things, a proactive approach to preventing failures and quickly restoring the full functionality of machines and equipment, minimizing disruptions in production. One way to support the effectiveness of maintenance activities is through the use of dedicated software or applications. This paper aims to present the effects of implementing the SAP Mobile Plant Maintenance (MPM) module in a manufacturing company. The paper provides an overview of SAP ERP systems and describes the main tasks involved in the implementation of the SAP MPM module. Based on a case study, the impact of the SAP MPM module on management processes and maintenance efficiency is presented. A reduction of over 85% in the average time to report failures has led to minimized production losses, while a faster response to failures has resulted in less complex and costly repairs. Additionally, the company observed benefits such as improved data quality, enhanced productivity, increased employee efficiency, and better availability of technical systems, all of which contributed to reducing overall company costs.

Keywords: Manufacturing execution control, Process control, Smart maintenance
Abstract: The low-pressure carburizing process is a key element in production engineering for many steel manufacturing technologies. The paper presents the results of research on the properties of ceramic insulation of the PIT vacuum furnace during low-pressure carburizing processes. A particularly important aspect was the analysis of the effect of the carbon deposit on the temperature distribution in the furnace insulation during thermochemical treatment processes. The experimental studies were carried out based on defined, practical low-pressure carburizing procedures (temperature, pressure, number of carburizing cycles).

Keywords: Vision systems, Cobots (collaborative robots), Intelligent manufacturing
Abstract: In the era of Industry 5.0 Human Robot Collaboration (HRC) workstation should be human-centric. The main issue is to develop the Artificial Intelligence (AI)-based models to ensure work safety in the HRC based on vision system. In this study, based on the acquired real-life video recordings (VRs) at the research unit of HRC, the impact of the interferences in the VRs on the efficiency of collision detection within vision system was analysed. The different levels of the main interferences in the vision system were analysed as the 60 variants of the following parameters in VRs: (1) brightness levels, (2) analog noise, (3) chromatic aberration and (4) digital interferences. Next, the novel algorithm for determining the instantaneous collision risk at the HRC workstation based on the relative positions of objects detected in images when using the AI-based technique, namely Region-Based CNN (YOLOv8 Tiny) was developed. Applying this algorithm, the percentage-based quantitative comparison analysis for reference VRs with VRs with the determined interferences for each of the four scenarios was provided. The research results enable the determination of the risk of detecting a collision in the HRC workstation in the event of the interferences in VRs. Next, it evidences that the level of interferences in VRs causes a risk of collision to be underestimated or overestimated in the HRC workstation.

Keywords: Intelligent agent, Optimization, Monitoring
Abstract: In this research a novel extension of the Close Enough Dubins Orienteering Problem (CEDOP) is presented, dubbed the CEDOP with Angle Dependent Scores (CEDOPADS). In the CEDOPADS, the score associated with each target can only be collected from certain prespecified distances, observation angles, and headings, which are assumed to have an direct impact upon the score collected. A methodology for generating CEDOPADS instances from existing OP instances is presented, by incorporating obstructions and considering their impact on the observability of the targets. Finally a new dataset of problem instances, based on the OP instances presented in Chao (1993) is generated.

Keywords: Energy, Process planning, Genetic algorithm
Abstract: The following research paper presents the challenge of optimizing the serialization of production orders, particularly emphasizing the energy efficiency criterion. The serialization process primarily leveraged genetic algorithms, which were subsequently undergoing a comparative analysis against exemplar heuristic rules, as well as the fuzzy logic algorithm and Johnson algorithm. Evaluation of the prioritization was performed using additional criteria, including the overall processing time of all orders, total delay time, and the delay cost function. Furthermore, the investigation involved the simulation of two distinct production systems: one integrated with a prediction module and another operating without such a module. In line with the Sustainable Development Goals of the United Nations, the primary objective of this research is to identify and recommend optimal strategies for serializing production orders, aiming to effectively minimize and optimize energy consumption. The research findings indicate that optimal scheduling methods vary based on the optimization criterion, with genetic algorithms proving significantly more effective in minimizing energy consumption compared to heuristic rules and fuzzy inference methods. The study underscores the advantage of integrating genetic algorithms with predictive modules, as this combination demonstrates notable effectiveness in achieving energy-efficient scheduling.

Keywords: Scheduling, Production planning
Abstract: We deal with the problem of scheduling production jobs combined with the problem of scheduling cooperated automated transport system. The production sub-system is based on the flow of various jobs with unique but various technological routes, which is recognised as so-called job shop scheduling policy. The transport sub-system consists of a single automated guided vehicles running on predefined track. We model and analyse the problem on the ground of deterministic scheduling theory. We are seeking for the optimal optimal schedules of production and transport operations, which minimizes certain criteria. We consider chiefly two of them, namely the minimal cycle time and the makespan, being an upper bound of the minimal cycle time. We provide the formulation of the problem as an optimization task, the combinatorial formulation of the solution, the graph model, some properties of the problem, a solution algorithm, and an illustrative instance. The problem is dedicated for automated/intelligent systems without human personnel.

Keywords: Production planning, Assembly planning, Scheduling
Abstract: The growing automation and demand for customized products in modern manufacturing systems require innovative approaches to synchronize intralogistics with mixed-model assembly lines. This paper explores strategies to integrate production scheduling and sequencing to optimize the flow of materials and processes in flexible manufacturing systems. The study focuses on synchronizing component production subsystems with the main assembly line using Automated Guided Vehicles (AGVs) for material transport. We propose a two-stage approach: first, determining the minimum number of workstations required to maintain production continuity, and second, optimizing the number of AGVs to ensure seamless material flow. The research uses discrete event models and backward scheduling strategies to analyze system performance across various conditions. Results indicate that the proposed method effectively balances production and transport operations, minimizing delays and idle times. The study emphasizes the importance of integrating real-time data, predictive-reactive scheduling, and just-in-time principles to enhance the flexibility and efficiency of manufacturing systems. The work contributes to the advancement of production planning and control in complex, resource-constrained manufacturing environments.

Keywords: Decision making, Data science, Data-driven modeling
Abstract: This study explores the potential of machine learning models to enhance the forecasting accuracy of the spare parts in the automobile industry of developing countries. By evaluating various forecasting techniques and categorizing spare parts based on movement rates, the research demonstrates the significant advantages of predictive analytics by encompassing both statistical methods and machine learning approaches. Techniques such as Simple Exponential Smoothing, Croston, Random Forest Regression, and XGBoost Regression are compared to achieve the lowest forecasting errors. Overall, the research contributes to bridging the theoretical-practical by offering insights to drive operational efficiency and effective decision-making improvements in demand forecasting for the automobile sector in developing countries.

Keywords: Knowledge engineering, Evolutionary computation, Decision making
Abstract: Emergency vehicle accidents are increasing globally, thus we need to identify the underlying factors contributing to these incidents. Managing driver fatigue in emergency transportation systems is essential to ensure safety and operational efficiency. Therefore, this study introduces a multi-path emergency routing framework designed to mitigate driver fatigue through a modified genetic algorithm (MGA). Utilizing topic modeling techniques, key fatigue-related factors are extracted from extensive datasets sourced from the Web of Science. These insights inform the development of an MGA-based routing strategy that dynamically adjusts to drivers' fatigue levels, optimizing routes for both safety and performance. Fatigue levels are quantified using fuzzy logic, applying possibility and necessity approaches to assess the severity of driver fatigue based on the number of red signals crossed during the journey. A trade-off between travel time and fatigue severity indicates that the MGA improves route selection by reducing travel time by 67.1%, fatigue severity by 26.3%, and red signal encounters by 28.4%, emphasizing the importance of adaptive routing decisions based on operational conditions. Findings show peak fatigue (~2.7) at 09:00 AM, likely due to rush-hour congestion and commuter stress, demand further study into time-critical travel behavior. This framework offers valuable contributions to reducing fatigue-related risks and improving decision-making during critical emergency scenarios.

Keywords: Process engineering, Manufacturing execution control, Logistic planning
Abstract: The importance of adopting Climate-Smart Agriculture (CSA) strategies cannot be overstated especially as these strategies are fundamental for improving the overall resilience and sustainability of horticulture supply chains in the presence of climate change. This research aims to identify and prioritize major challenges preventing the successful implementation of Climate-Smart Agriculture (CSA) strategies in horticulture supply chains. It uses a literature review, and the Delphi technique was used to modify and confirm the barriers that were found. Primary data from horticulture farmers, and the Analytical Hierarchy Process (AHP) to classify these barriers from highly critical to least important. The goal is to improve the resilience and sustainability of horticulture supply chains in the face of climate change.

Keywords: Logistic planning, Case study of digitization or smart system, Autonomous system
Abstract: Floods are devastating disasters that often result in significant loss of lives. In such a scenario, delivering essential supplies to affected people is extremely challenging, especially when traditional transportation infrastructures are damaged by floods. Which makes it inaccessible to most of the affected areas. UAVs offer an effective means to overcome these barriers by facilitating the timely and efficient delivery of relief materials to inaccessible regions. However, limited UAV resources lead to inefficiencies in relief operations over large flood-affected areas. To address this challenge, this study develops a Multi-Criteria Decision-Making (MCDM) model employing the Analytic Hierarchy Process (AHP). The proposed model aims to rank flood-affected regions based on criteria such as the severity of the flood in the region, the urgency of humanitarian needs, and the population density of the regions. This ensures that UAV resources are allocated to priority areas with the highest demand. The proposed model was tested using an empirical case study from Sri Lanka. This study demonstrates how using AHP can significantly improve disaster relief efforts by optimizing UAV resource allocation and ensuring the timely delivery of essential supplies to the most affected flood zones.

Keywords: Human factors, Human interaction, Decision making
Abstract: Value creation (VC) can be defined as a process of producing goods, services, experiences, or solutions that satisfy specific requirements or desires, thereby generating benefits, utility, or value for individuals, organizations, and society. In this paper we try to understand the value that the Industry 5.0 paradigm can offer to consumers that remains largely underexplored. We followed the research design of conceptual model building and used Technology-Organization-Environment framework and the Socio-Technical Systems theory as theoretical lenses. Following this design, we have synthesized the RESET framework of Value 5.0 comprising of five constructs: (1) Resilience Value, (2) Environmental Value, (3) Social Value, (4) Economic Value and (5) Technological Value. Being the first of its kind, this framework has many practical and theoretical implications. Business entities can use the framework to evaluate the value that can be generated by adopting Industry 5.0, which could act as a major incentive for them to adopt the new revolution. Policymakers can use the conceptual pathways outlined in this framework to align their strategies and promote Industry 5.0 adoption effectively. Moreover, the researchers can build a new stream of research to further elaborate the VC process within an Industry 5.0 ecosystem. We conclude this paper by stating the need for empirical grounding for the proposed framework due to its conceptual nature.

Keywords: Logistic planning, Optimization, Smart system
Abstract: Agriculture plays an important role in supporting national economies and community livelihoods, with growing demand for food posing considerable challenge to agricultural productivity. Traditional approaches often fall short in satisfying these requirements. The use of pesticides is a key part of agricultural operations; nevertheless, traditional application methodologies have poor efficiency and expose agricultural workers to danger and therefore become impractical for widespread use. Unmanned Aerial Vehicles (UAVs), also known as drones, have been utilized for pesticide application, but still less accuracy, causing healthy crops to become contaminated by spraying. To prevent this problem, geofencing technology can be utilized to restrict pesticide sprays to target regions. In addition, Google OR-Tools utilizes the Capacitated Vehicle Routing Problem (CVRP) for improving efficiency in routing drones. In this work, both methodologies have been incorporated together to maximize efficiency in pesticide application with a view to minimizing energy consumption.

Keywords: Logistic planning, Optimization, Digital manufacturing
Abstract: Determining the most effective picking method that reduces travel time, distance, and operational costs has become a critical aspect of decision-making in omnichannel warehouse management. This study investigates the picking operation in an omnichannel warehouse focusing on how picking strategies, specifically automation, can improve the efficiency of picking operations. This study constructs multiple models to simulate picking operations within an omnichannel warehouse setting. It starts by evaluating current manual picking processes and subsequently integrates automation technologies into the same models for comparative purposes. The results underscore the effectiveness of automation in enhancing picking efficiency and offer practical insights for industry professionals aiming to optimize warehouse performance.

Keywords: Data-driven modeling, Enterprise resource planning
Abstract: The implementation of the concept of sustainability poses challenges for organisations to measure their environmental, social and economic impacts. A key element of this process is sustainability indicators, which are used to monitor, evaluate and report on the effects of sustainability activities undertaken by companies and institutions. SD indicators are tools or metrics that allow the effectiveness of sustainability activities to be assessed. They can be quantitative or qualitative in nature. They cover a variety of aspects, such as energy efficiency, water resource management, social responsibility and financial aspects. These indicators not only allow progress to be tracked, but are also a key tool in making informed strategic decisions. The complexity of the indicators results in the auditing process that results in a report on the company's SD activities becoming time-consuming and costly. The aim of this article is to present a research concept for the creation of a proprietary, intelligent and rapid method for assessing the sustainability of Industry 4.0 and 5.0 companies using data from ERP systems. The paper presents the results of a survey of managers responsible for sustainability and ESG reporting in manufacturing companies. The results provided an understanding of the practical aspects of implementing SD standards and the challenges associated with this process.

Keywords: Smart maintenance, Internet of Things (IoT), Decision making
Abstract: Contemporary organizations operate in a dynamically changing social and economic environment. Globalization, digitalization, climate change and growing social expectations force companies to be flexible, innovative and able to respond quickly to disruptions. Disruptions in global supply chains, such as those caused by the COVID-19 pandemic or geopolitical conflicts, emphasize the need for reliable internal processes. Maintenance minimizes the risk of downtime in key operations, which allows organizations to become independent from external problems. In this context, maintenance becomes a key tool supporting the resilience of organizations. Moreover, modern society expects companies to act in a pro-ecological and socially responsible manner, and one of the aspects of such approach is maintenance activity, prolonging equipment lifecycle and reducing waste. Solutions changing the approach to maintenance is the development of technologies such as the Internet of Things (IoT), artificial intelligence (AI) and big data analysis, as they enable collection and processing of data that can be used in predictive maintenance to predict failures in advance, which increases efficiency, reduces cost and enhances organizational resilience. To sum up, maintenance in the era of contemporary socio-economic challenges is not only the technical foundation of an organization's operations, but also a strategic element supporting their adaptability, sustainable development and resilience in a changing market. The goal of the study is to explore the links between maintenance and resilience and use the results of the analysis to develop the model of management that benefits from the positive impact of maintenance activities on resilience. The study is organized as follows: Section 2 outlines the research methodology, including indication of data sources, presentation of research methods and tools. In Section 3 research results are presented, Section 4 presents developed model. Finally, Section 5 concludes the study with a summary, implications of the research, its limitations and new opportunities for future research.

Keywords: Assembly planning, Human factors, Decision making
Abstract: Assembly lines are crucial in manufacturing, enabling efficient task organization and resource allocation. While automation has advanced in many industries, sectors producing large-sized products or small series often rely heavily on human labor due to low automation levels. This reliance highlights the need to optimize work organization and ergonomic conditions to enhance both production performance and worker well-being. This paper addresses the Ergonomic Assembly Line Balancing Problem with parallel workstations (ErgoALBP-PS), a challenging optimization problem that combines traditional objectives, such as minimizing Time Smoothness Index, with ergonomic metrics like physical and mental workload. A novel mathematical model is proposed to balance these competing objectives, accounting for the complexities of parallel workstations. The model’s effectiveness is demonstrated through real-world use case, showcasing its ability to achieve balanced task assignments, reduce both physical and mental risk, and improve overall efficiency. The results underscore the importance of integrating ergonomic considerations into assembly line design, particularly in low-automation settings.

Keywords: Autonomous system, Assembly planning, Machine learning
Abstract: This research evaluates statistical models to predict travel times, focusing on XGBoost for estimating maneuver durations in batching robot planning with a delta picker. Using simulated motion data from an industrial testbed, XGBoost excels in adapting to environmental changes, modeling non-linear behaviors, delivering quick results, and providing accurate, explainable predictions. The study highlights the growing trend of data-driven solutions in manufacturing and production systems.

Keywords: Human interaction, Case study of digitization or smart system, Digital manufacturing
Abstract: This paper investigates the application of the CERAA Glove, a developed smart glove system equipped with advanced force sensors, in assessing hand action forces during industrial tasks. The study evaluates the effectiveness of the system in reducing physical strain on workers through the implementation of technical corrective measures by innovative glove system. Initial measurements revealed significant exceedances in grip force limits during manual handling tasks. The introduction of corrective measures, including a modified grip technique and changes to task dynamics, resulted in a marked reduction of exerted forces. Post-implementation measurements demonstrated grip forces consistently below the critical threshold, highlighting the device's utility in improving workplace ergonomics and worker comfort. The findings underscore the importance of integrating such ergonomic technologies in industries to mitigate physical strain and promote sustainable working conditions. The work brings a new way to accurately detect and reduce excessive forces when working with hands using a smart glove, thereby solving the problem of worker overload.

Keywords: Intelligent manufacturing, Autonomous system, Digital manufacturing
Abstract: Germany/Europe and China are both leaders in smart manufacturing and the application of industrial AI but differ in their approaches to market regulation and standardization. China uses a centralized, government-led model that aligns domestic priorities with international standards. Germany and the European Union, on the other hand, follow a decentralized, stakeholder-driven process that emphasizes collaboration and interoperability to meet both domestic and global needs. This paper explores the opportunities and challenges of harnessing artificial intelligence for manufacturing by comparing the two approaches. The analysis highlights the key role of standardization in linking innovation to market adoption, ensuring interoperability and promoting sustainable growth. Challenges such as divergent regulations, market fragmentation and limited resources require increased international cooperation. The findings highlight the value of early integration of research findings into standardization processes to adapt to regulatory requirements and accelerate industrial adoption. This approach strengthens competitiveness while promoting ethical and sustainable AI practices. The paper provides recommendations for bilateral cooperation to drive AI innovation, sustainability and resilience in manufacturing systems.

Keywords: Intelligent manufacturing, Human interaction, Machine learning
Abstract: Within the domain of Industry 5.0, the “Healthy operator” concept aims at using biometric sensors to enhance the physical and mental wellbeing of the operators. However, a sizable portion of the tools and the methodologies involved are borrowed from fields such as neuroscience and medical science where experimental context and conditions are significantly different from the typical manufacturing plant. In the current work we explored if the method for recording baselines of biometric signals used in clinical studies should be directly transferred to the shop floor. We measured baselines in sitting and standing positions and compared these two methods by measuring the performance of Stress state prediction model during standing assembly tasks.

Keywords: Data science, Big data, Machine learning
Abstract: The vast amount of information generated by big data, combined with modern computational capabilities, enables numerous applications in the automotive industry, such as anomaly detection, electric vehicle range prediction and optimal operational strategy. Vehicle sensors and network technologies record multivariate time-series data. However, these data are characterized by chaotic, noisy and unstructured properties. To address these challenges, this work develops models for extracting insights, learning representation, removing redundancy and structuring the data with labels. The approach integrates various autoencoder architectures for dimensionality reduction and representation learning, different clustering algorithms for grouping and organizing data and visualization techniques for the interpretation of the results.

Keywords: Design engineering, Digital manufacturing, Smart factort
Abstract: Demonstrators play an important role in manufacturing by enabling research into problems and testing of new technologies. Despite the many manufacturing demonstrators that have been created, no clear design approach or method is often presented. This research outlines the design and development process used to create a novel multi-location demonstrator. A list of 20 key requirements was gathered from industry scenarios, key projects and research labs involved and used for the design. All 20 requirements are met or at least partially met by the final demonstrator design. The final usefulness of the demonstrator is highlighted by its use in several studies. Other organisations could replicate the design method outlined in this study to build demonstrators based on their existing resources and requirements.

Keywords: Digital twin, Smart maintenance
Abstract: The paper aims to assess the state of the art on adopting Digital Twins in maintenance and asset management. The research subject is scientific publications and patents in the indicated fields of interest, which have been included in publicly available databases such as Scopus and Google Patents. The paper aims to identify the current state of research and assess the trend regarding the growth rate of publications and patents from 2017 to 2024. The research method includes collecting relevant data and fitting them to S-curves using Fisher-Pry and Gompertz models. In addition, the collected publications are evaluated in terms of a general overview of topics and country of origin. The results provide high-level information about the current status of Digital Twin technology; in particular, they indicate that the technology diffusion is nowadays completing the slow growth phase (i.e., the initial stage) and is gradually entering the acceleration phase (i.e., the fast growth stage). The results should be extended by additional analyses in future research: the content analyses of topics in publications and patents should be detailed to assess the functionalities of Digital Twins in maintenance and asset management that could be promising for applications to innovate the industrial practice.

Keywords: Process control, Process engineering
Abstract: In aluminum castings the proper thermal management of the die is of crucial importance for the quality of the resulting cast parts. However, nowadays it is in many casting processes still common practice to neglect temperature information of the die during serial production and to entirely rely on pre-defined process settings. In casting processes with a significant variation in cycle times this often leads to a high variability of process conditions and, as a result, in elevated scrap rates. The current paper investigates the potential of feedback die temperature control for the manufacturing of aluminium e-mobility components in the Rotacast process. For this purpose a novel PI-based cycle-to-cycle adaptation of process settings is proposed. The applicability of this proposed control scheme is demonstrated with data from a real casting campaign. In this pilot application it is shown how the variability in die temperatures can be decreased by half (change in inter-quartile range of start temperatures from 19°C to 8°C).

Keywords: Self awareness, self diagnosis,prognosis
Abstract: This systematic review examines the application of Federated Learning (FL) in Prognostics and Health Management (PHM). Through a structured analysis of all relevant literature using the PRISMA methodology, we analyze research from major academic databases to investigate the current state of FL frameworks, their application to prognostics tasks, and associated implementation challenges. To the best of our knowledge, this is the first review dedicated to exploring the intersection of FL and PHM. Our findings highlight a growing interest in this domain and reveal that applications remains underexplored. This review aims to provide clear insights into the current landscape and guide future research and practical applications in industrial contexts.

Keywords: Web-based system, Internet of Things (IoT), Smart system
Abstract: Monitoring systems for vital signs are needed to detect ongoing changes and to analyse the data later to detect diseases, but recording such data has been expensive until now. Currently, popular IoT modules allow basic parameters to be measured. The aim of this article is to verify whether IoT modules provide reliable data. Tests were carried out using universal solutions (UDP protocol, ESP32 microcontroller) and then the recorded measurement results were compared to those obtained with several IoT devices. The results of the comparison are sufficient to build a vital signs monitoring system.

Keywords: Case study of digitization or smart system, Vision systems, Virtual reality
Abstract: Properly planned escape routes are very important for employee safety in large buildings. The development of an escape route in old buildings should be preceded by simulations to help identify bottlenecks, i.e. places where congestion can potentially form. This paper presents the process of simulating an escape route using the Unity environment, which allowed the identification of bottlenecks. In addition, using the Unity game engine, we created a dynamic simulation model of an existing building layout, enabling realistic visualization and analysis of crowd movement under emergency scenarios. Unity's physics and agent-based movement systems allowed for a study of how people may interact with their environment during an evacuation. Our results identified specific structural features that contribute to congestion, such as turns and stairs. This work demonstrates the practical applicability of game development tools like Unity in safety planning and provides a scalable method for evaluating evacuation procedures in complex architectural settings.

Keywords: Digital twin, Manufacturing execution control, Production planning
Abstract: The article provides a short review on the use of digital twin technology for modeling and simulation in continuous production processes (coal, raw materials, metallurgy, energy). In this area, the deployment of advanced Industry 4.0 technologies is less common. The article analyzes information systems at coal processing plants in the Czech Republic and the possibilities of connecting them with digital twin. This case study was chosen because of the author's long-term involvement in the design and implementation of these systems. Three areas appear suitable for the possible deployment of the digital twin simulation - simulation of technological processes in coal processing, monitoring of machine downtime and logistics of processing plant output products.

Keywords: Bio manufacturing, Augmented reality, Human factors
Abstract: Snapshots of a working day, during which monitoring focuses on all the work shift and assesses absolute times and time percentages of individual actions within the work shift, are compiled for operators of harvesters. The highest share is taken up by operation time (unit work time) which constitutes 74.8 %. It is followed by maintenance time constituting 9.2 % and time necessary for harvester repairs which constitutes 5.9 % of the shift time. Other times, i.e. ration and shift time, represent time for the preparation and completion of work, time for work instructions, time for technical service, time for biological breaks and rest, time losses due to technical and organizational losses and other unnecessary time. The coefficient of ration and shift time for harvester operation can be defined as 1.258. This coefficient increases operation time by the time necessary for the harvester operation.

Keywords: Human interaction, Web-based system, Mathematical programming
Abstract: The design of embedded systems, including CNC machines, now uses object-oriented languages and component structure, where the user interface unit is separated from the execution unit (which directly controls motors, etc.). This structure of systems can lead to errors that are difficult to detect during testing. This paper presents an attempt to detect errors in a development version of CNC machine software. The detected error belongs to the class of race condition and could cause the machine not to work properly. In order to detect the error, a model of the source code was built and a modified RD-L method was used, in which conditions are defined, the fulfillment of which will indicate the occurrence of the error.