Innovative Intelligent Industrial Production and Logistics

The current industrial landscape is characterized by the immense effort towards the digitalization of systems, networks and processes. Under the light of Industry 4.0, digital technologies such as Digital Twins (DT), eXtended Reality (XR), and Additive Manufacturing (e.g. Fused Deposition Modeling – FDM) are considered among other key pillars for completing the digital transformation of manufacturing and production systems/networks. Ultimately, the goal is to enhance system dependability, ensure precise process monitoring, and predict future system failures. Consequently, in this research work the design, development, and implementation of a Digital Twin based framework for monitoring and improving Fused Deposition Modeling (FDM). The focal point of the Digital Twin is to assess the quality of 3D printed components, calculate the experimental result and accordingly adjust the process parameters, via the seamless connection of the physical printer to its DT. As a result, a fundamental aspect of this study revolves around outlining the information flow among the previously mentioned components to optimize FDM process parameters, thus reducing the time and resources squandered due to human inaccuracies. The execution of the proposed framework hinges on the integration of appropriate communication protocols and channels between the physical and virtual counterparts in order to achieve seamless and continuous communication.

BLDC motors for demanding applications require a sensor system for electronic commutation to determine the current rotor position. Often, three Hall magnetic sensors, angularly displaced by 60º or 120º, are used for this purpose. Ideally, commutation should be performed precisely at the rotor position where the back induced electromotive force (BEMF) crosses the zero value. However, in reality, the deviation (phase shift between the voltage crossing the zero value and the HALL sensor switching) is different from 0º and depends on the mechanical tolerances of motor manufacturing, sensor soldering position, and rotor magnetization repeatability. To perform comprehensive final testing of motors, a dedicated measuring method had to be developed, as professional measuring instruments (Frequency counters) are not suitable for this purpose. This is due to winding voltage not being sinusoidal but pulse-width modulated. It is also not possible to rotate the motor during testing with an additional drive. Consequently, that is why the measurement can only be performed during the coast-down period, when the rotational speed exponentially decreases. In this paper, we present a solution, developed for the EOL (end-of-line) quality assessment production of electric motors (BLDCs) at Domel company. It consists of a PCB for signal preparation, a fast USB module for simultaneous acquisition of analog signals, software for data processing, result analysis and result presentation and company database for results storage.

Availability of computation capabilities and real-time machine data is one key requirement of smart manufacturing systems. Latency, privacy and security issues of cloud computing for Industrial artificial intelligence (AI) led to the edge computing paradigm, where computation is performed close to the data source. As on-premise edge deployments require companies to allocate budget and human resources to acquire and maintain the required information technologies (IT) infrastructure and equipment, they are not feasible for several companies. However, 5G can merge advantages of previous alternatives. Multi-Access Edge Computing (MEC) servers deployed at the edge of the 5G network close to the final user, offer security, privacy, scalability, high throughput and low latency advantages. MECs are suitable for industrial AI, while industrial companies do not face the burden of acquiring and maintaining servers and communication infrastructures. This paper proposes a real-time high-frequency data streaming architecture to deploy Industrial AI applications at MECs. The architecture has been successfully validated with data sent through a 5G network to a Kafka broker at the MEC, where different microservices are deployed in a Kubernetes cluster. The performance of the architecture has been investigated to analyze the capabilities of 5G and MEC to cope with the requirements of Industrial AI applications.

Data Science plays a crucial role in driving new approaches to process optimization. With the increasing complexity of internal logistics systems, data-oriented methods have become essential in addressing the challenges that arise. However, standardized process analytics frameworks are lacking due to the heterogeneity of the underlying processes and the resulting data. This article aims to address this complexity by presenting a categorization of internal logistics data, consolidating the current state of the art. The categorization takes into account both real-world and scientifically proposed data architectures, providing a comprehensive overview. It includes a classification of comparative data fields based on their importance, the associated internal logistics processes, and potential usage scenarios. This classification is designed to cater to different use cases, such as diagnostics or prescriptive analytics. By presenting this categorization, the article enables practitioners to effectively leverage generated process data in a more goal-oriented manner. It empowers them to conduct suitable analyses tailored to their specific needs and objectives, based on the provided data architectures. In summary, this article offers valuable insights into internal logistics data categorization, providing a framework for practitioners to make informed decisions and optimize processes using data-driven approaches.

The transition to Industry 4.0 requires the modernisation of legacy systems. This change poses challenges, in particular due to the different data formats and interfaces resulting from the heterogeneity of the components involved. To ensure seamless interoperability within Industry 4.0, a consistent data base is essential. The Asset Administration Shell (AAS), a standardised digital twin of assets, plays a key role in driving data-centric, interoperable solutions. The potential for efficient data integration means fewer errors and optimisation for the digitisation of legacy systems. This prompts an investigation into the necessary extensions and components within the AAS infrastructure to implement this process. Our paper outlines a strategy for integrating legacy systems into the AAS environment. We introduce new components that enable the modelling and embedding of information about assets throughout their lifecycle. The conceptualised architecture is designed to facilitate the assisted selection of relevant submodels, the reuse of existing resources and the automated generation of AASs. A service is proposed to verify the structure of the resulting AASs and the communication configuration. Using an articulated robot as a case study, we demonstrate the connection of data points via the OPC UA protocol. In addition, a prototype is presented that enables vertical data integration via the BaSyx DataBridge, highlighting the benefits of automated integration into AASs. The approach is versatile and is readily applicable to a variety of systems and scenarios as required.

Automation technology is undergoing a paradigm shift, manifested in the desire for flexible and adaptable systems. This trend is leading to an ever-increasing complexity of systems and their components. To manage this complexity, it is essential to promote interoperability between system components using Industry 4.0 components. Various software environments are available for their deployment. However, a test environment that supports essential test strategies for their verification remains to be developed. In particular, retrofitting and supporting legacy systems is a challenge. In this paper, we present a concept for a test environment based on the Asset Administration Shell (AAS), which incorporates established test methods from software engineering. This environment includes a multi-level testing process, which is able to evaluate both the structural conformance as well as the configuration and behaviour of the Industry 4.0 component. Test execution is facilitated by a ticket system that uses descriptors to enable test deployment without additional software development effort. The test environment can be seamlessly integrated into the existing AAS component landscape as an architectural component.

The increasing number of requests in the last-mile delivery has led to the introduction of advanced technological solutions to enhance couriers’ services. In addition, innovative strategies like crowd-shipping have been introduced in order to create synergies within the territory and involve ordinary people in the transportation activity with the aim of reducing operational costs and pollution as well as of increasing the service level. We refer to a company that manages a crowd-shipping platform and provides transportation services within time windows through its own fleet of vehicles and occasional drivers. The service requests correspond to pairs of pickups and deliveries. The objective of the company is to maximize the profit by balancing risks and benefits, from an economic perspective, associated with the involvement of ordinary people in fulfilling service requests. We extend the pickup and delivery problem with occasional drivers and regular vehicles, introducing risk and compensation considerations. The computational analysis conducted through an optimization model shows how the use of occasional drivers reduces overall costs. Moreover, a series of managerial insights is provided thanks to a sensitivity analysis on the risk and compensation parameters associated with crowd-shipping service.

Strategic Supply Chain Capacity Planning (SSCCP) is an essential activity for companies to prepare their future. However, since uncertainty became an essential factor to consider in this decision-making process, existing solutions to support this process do not fully satisfy their needs anymore. Especially in terms of uncertainty space coverage while exploring and assessing scenarios associated with uncertainty sources and decision options. Therefore, this paper introduces an approach to overcome the complexity of scenario exploration and improve the uncertainty space coverage, to better support SSCCP decision-making. This approach includes a bi-objective metaheuristic that first explores a probability-impact matrix to define a relevant subspace to consider in this uncertainty space, and then uses this subspace to explore the decision space and define a relevant subspace of this decision space to assess and display to decision-makers. Then, an implementation and experiment are described and discussed, and finally avenues for future research are suggested.

The growing operation complexity has led companies to adopt many indicators, making complex the evaluation of the overall performance of logistics systems. Among several studies about logistics management, this is the first one to determine an overall logistics performance indicator and evaluate its impact for logistics management. The proposed methodology encompasses four main phases: the first one defines the management scope and the indicator set, the second applies statistical tools reaching an initial model for indicators aggregation, the third one determines the global performance model, and the last phase implements the integrative indicator with its scale. The methodology is implemented in an outbound process from a Brazilian company with eight logistics KPI’s. Principal Component Analysis (PCA) is used to stablish the relationships among indicators and an optimization tool is applied to define the integrative indicator scale. The global performance (GP) provided by the integrative indicator has demonstrated that even if important indicators have not reached their goal, it is possible to attain a good global performance with improvements in other areas. Thus, the framework demonstrates to be a useful solution for logistics performance management.

The advantages of track & trace solutions to enterprises that move physical goods are widespread. Yet, track & trace solutions are not a standard component of an enterprise’s system landscape. While mostly being triggered by customer requests, track & trace initiatives are often siloed. Seldomly, the entire enterprise and even more rarely the enterprise’s network are considered. The reasons for a siloed implementation of track & trace software are manifold. Different business priorities, adverse data sharing policies, lack of knowledge, and a lack of structured approach are a few to mention. Starting with the structured approach of the digital enterprise architecture (DEA) is a first remedy to these issues. From a technological perspective, embracing a blockchain subnet gives enterprises access to solutions for data ownership, validity, and integrity impediments. The blockchain-based subnet is compatible with a bimodal enterprise architecture. This allows enterprises to connect to their supply chain partners’ logistics systems differently than to their own one. Several steps need to be considered to achieve a functional supply-chain-network-wide track & trace solution. From finding consensus among the blockchain peers to designing the track & trace block, academic input is combined to render the implementation of a blockchain-based track & trace solution possible. From a business perspective linking the requirement catalogue with the purposeful use of the latest technology follows an end-to-end approach.

The main job of a virtual imaging try-on is to transfer a garment to a specific area of an individual’s body part. Trying to deform said garment so that it fits in a part of the desired body. Despite some research, the vast majority use a low-quality image resolution of 192\(\,\times \,\)256 pixels, limiting the visual satisfaction of online users. Analyzing this visual limitation, we find that the vast majority of the algorithms use these mentioned measures to obtain better performance and optimization during their training, since while the number of pixels is smaller, in the same way, their execution time will be less in the generation. of segments or masks of garments and body parts. Despite having better performance and optimization, quality and pixel size are also of the utmost importance, since it is in the final resolution that the result for the user is appreciated. To address this challenge, we propose a super-resolution extension module, added to the ACGPN model. Such a module gets the resulting image from the ACGPN model, and then with the help of computer vision aims to increase the resolution to 768\(\,\times \,\)1024 pixels with minimal loss of quality. For this, a comparison of models that perform this task of increasing the resolution is made. Finally, it is quantitatively shown that the proposal obtains better results.

This paper presents the development of a mathematical model for product assignment in a Vertical Lift Module (VLM), which are increasingly employed in the industrial sector due to their advanced technology and efficient parts-to-picker process. Mathematical modelling plays a crucial role in addressing the complexity of these problems and providing intelligent and innovative solutions. Despite being a tactical problem with medium-term implications, the competitive nature of the industrial environment demands quick adjustments driven by the mass customization paradigm. This requires continuous evaluations to reconfigure the supermarket accordingly, which can be efficiently accomplished through the rapid application of artificial intelligence using advanced mathematical methods.

The proposed integer linear programming, which is based on the well-known transportation problem, features a simple objective function aimed at minimizing the number of trays in the VLM. Additionally, five constraints are included to ensure the applicability of the model to real-world scenarios. The simplicity of the AMPL implementation of the mathematical model is emphasised. Experimental computation using real data validates the proof of concept and assesses the impact of introducing new rules for product assignment. This research also explores the potential for optimising warehouse operations and suggests avenues for further investigation.

The emergence of Industry 4.0 (I4.0) technologies has significant implications for small and medium-sized enterprises (SMEs) in the manufacturing sector. Current research highlights the benefits of I4.0 technologies but often overlooks the unique challenges and needs of SMEs, particularly in the transition from implementation to routinization of data analytics (DA) in the context of I4.0 initiatives. Our paper addresses this gap by introducing a prototype of an integrated data analytics model (iDAM) specifically designed to help SMEs incorporate DA as part of I4.0. Our model was developed based on a comprehensive review of existing frameworks and methodologies. It covers three key areas (the project situation, the organization’s maturity level, and the application landscape) and proposes a situational process model to bridge the implementation-routinization gap. We demonstrate and evaluate our approach using a practical, real-world use case of a multi-stage manufacturing process in an SME. The iDAM provides a structured and tailored approach to guide SMEs in operationalizing DA based on their individual maturity level and promote the use of DA methods.

Office work has become the most prevalent occupation in contemporary society, necessitating long hours of sedentary behavior that can lead to mental and physical fatigue, including the risk of developing musculoskeletal disorders (MSDs). To address this issue, we have proposed an innovative system that utilizes the NAO robot for posture alerts and camera for image capture, YoloV7 for landmark extraction, and an LSTM recurrent network for posture prediction. Although our model has shown promise, further improvements can be made, particularly by enhancing the dataset’s robustness. With a more comprehensive and diverse dataset, we anticipate a significant enhancement in the model’s performance. In our evaluation, the model achieved an accuracy of 67%, precision of 44%, recall of 67%, and an F1 score of 53%. These metrics provide valuable insights into the system’s effectiveness and highlight the areas where further refinements can be implemented. By refining the model and leveraging a more extensive dataset, we aim to enhance the accuracy and precision of bad posture detection, thereby empowering office workers to adopt healthier postural habits and reduce the risk of developing MSDs.

“Made in Italy” products and Italian manufacturing are worldwide recognized for their quality. Nonetheless, businesses and societies are evolving, affected by structural transformations. To maintain their competitive advantage, Italian companies are asked to move towards a transformation aligned with global call for actions addressing critical issues, such as climate change. The transition of manufacturing companies, in particular Small and Medium Enterprises (SME), towards circular economy should be supported by adequate investments. To answer a national call, the Extended Partnership (EP) “Made in Italy Circolare e Sostenibile” was established. The EP aims to provide research and innovation resources to enable circular manufacturing practices in Italian companies, developing best practices to be adopted by SMEs. One of the main themes that the EP is investigating is the one of Product Service Systems (PSS), which appear as a viable path to achieve environmental sustainability. Nonetheless, resources and researches to support manufacturing companies in the path of servitization are still required. This paper aims at presenting a project, created in the context of the EP, to support companies in the development of circular PSS business models, in particular leveraging the opportunities offered by digital technologies.

This paper is giving an overview on different industrial application fields for LiDAR sensors beyond the field of autonomous driving. With insights to three specific use cases, different approaches to process the LiDAR point cloud data are described and referring results and findings of the developed applications are summarized. The application fields of the described use cases are the surveillance of industrial process environments of automated fenceless cells, the provision of visual assistance and location information for crane operators and the monitoring of the storage space occupancy in a port terminal. Based on the information from the three use cases and further general LiDAR related background an initial morphological box is drafted to enable the classification of industrial LiDAR use cases. The paper concludes with a brief overview on future work.

The impact COVID-19 generated on people routine linked with the rapid digitalization has led the educational approach to the need of a fundamental shift. In response to these evolving circumstances, SMILE (Smart Manufacturing Innovation, Learning-Labs, and Entrepreneurship) has undertaken a comprehensive initiative in order to move from the traditional learning labs towards digital ones. This digital transformation aims to enhance students’ learning experiences and problem-solving capabilities by leveraging technologies and innovative approaches. Digital Learning Nuggets are central to this paradigm shift, which are small units of interactive and engaging educational content. These nuggets have proven instrumental in augmenting students’ comprehension and retention of subject matter while fostering a more personalized learning journey. Moreover, the integration of MIRO, a collaborative software platform, further drives students’ learning by facilitating interactive discussions and fostering teamwork in a virtual environment. SMILE’s digital learning lab (SMILE DLL) serves as the foundation for the upcoming Hackathon, an event characterized by the collaboration between academia and industry. In collaboration with various companies, this experiential learning initiative offers students the opportunity to tackle problems posed by these organizations. Participants gain invaluable insights into real-world applications of their knowledge and are better equipped to address complex issues in a professional context. The Hackathon represents the link between academia and industry, fostering a dynamic environment where students can apply theoretical concepts in order to solve real problems. This immersive learning experience not only fosters their critical thinking and analytical skills but also nurtures creativity, adaptability, and teamwork, paramount attributes for today’s competitive job market.

Stockouts present significant challenges for Fast-Moving Consumer Goods (FMCG) companies, adversely affecting profitability and customer satisfaction. This research investigates key drivers causing Case Fill Rate (CFR) to fall below target levels and identifies the best model for predicting future CFR for the sponsor company. By utilizing feature importance techniques including Shapley additive explanations (SHAP) value plots, we conclude demand forecast error is the most critical driver influencing CFR. Machine learning classification and regression techniques were deployed to predict shipment cut quantity. To improve longer-term forecasts, a combination of models should be incorporated, along with extended historical data, promotions data, and consideration of exogenous variables. In conclusion, companies should prioritize forecasting accuracy and optimize inventory policy to improve CFR in the long run.

A continuous review policy (CRP) was used to simulate the inventory of finished product, raw material and packaging material at a warehouse. The ordering points were simulated based on the minimum order quantity and the production forecast. The safety stock levels did not impact the ordering points in this scenario because the demand was known and did not deviate. The simulation results showed that the average weekly number of pallets received at the warehouse was 341, with a standard deviation of 115 pallets. The aggregated ordering volume was fairly volatile, but the warehouse inventory levels were fairly uniform over time. The overall bin occupancy remained below 1,700, less than 20% of the total warehouse capacity. This suggests that the CRP is a good inventory management policy for products with known and stable demand. However, the volatile ordering volume could be a problem if there are constraints on processing incoming deliveries to the warehouse.

Digital twins have emerged as a key technological concept in the manufacturing industry. They form an information hub for industrial equipment and interact with dedicated applications in the operational manufacturing network. Digital twins consume and deliver information from machines to basically all connected applications. This results in complex integration requirements. The paper builds on previously designed semantic interoperability concepts for data-driven digital twins. It gives an overview of semantic data integration standards and provides insights into the current implementation of semantic integration patterns. Based on the challenges of the underlying research project “i-Twin”, semantic integration patterns provide standardized communication channels for operational management systems and connected assets. They build on the services of the semantic data integration middleware and use semantic connectors to bridge the proprietary data objects with an I4.0 compliant information model based on the asset administration shell (AAS). Semantic integration patterns will reduce the integration effort for equipment manufacturers and software providers, thereby accelerating automation and digitalization processes.

Digital transformation emerged as a strategic imperative for organizations in the era of Industry 4.0. As disruptive technologies reshape industries and business models, organizations must navigate this evolving landscape to remain competitive and relevant. Maturity models directed to Industry 4.0 offer a valuable framework to assess an organization’s current state, identify gaps, and develop a strategic roadmap for successful digital transformation. Digital transformation is a complex and multifaceted journey that requires a well-defined roadmap to guide organizations through leveraging digital technologies and capabilities. Therefore, a practical and sound approach to developing a roadmap is the basis for achieving a successful digital transformation strategy. A structured approach, based on building blocks, can lead to a systematic plan to implement digital initiatives. In this article, we will explore creating a roadmap to digital transformation using building blocks.

Under the background of the integration and development of next-generation information technology and the manufacturing industry, production equipment has evolved into Cyber-Physical System (CPS), which seamlessly combine embedded computing, network communication, and high-performance control. In this context, the information model of production equipment has gradually become a pivotal aspect supporting the visualization of the physical operation status of production equipment, optimizing decision-making in the digital realm, and enabling communication and interaction mapping in the cyber-physical space. Based on the model-based systems engineering (MBSE) principle, a unified information model for the whole manufacturing process of production equipment is proposed to solve the problems of diverse information types, diverse expression forms and complex model heterogeneity in the field of production equipment information modeling. From the perspectives of communication, knowledge, and geometry, the management activities of production equipment are analyzed, function modules are assigned, and information elements are defined. Consequently, a holistic information model for production equipment is formulated to support the integration of multidimensional models and realize the optimization and control of heterogeneous production equipment, which providing valuable reference for developers and practitioners of the production equipment information model.

The paper express results about a project on hyper-connected ecosystem of industrial networks and especially the services infrastructure developed during the project. A specific aspect is the concept and related feasibility study to establish an autonomous and distributed data management using the web presents of potential network partners. The question of the approach was “How to allow the accessibility of required information on demand everywhere and independent from a specific platform or cloud infrastructure for every partner or company within a business network”. It starts with the accessibility of information as a major asset for decision making. During the project, industry partners insist on reducing the preparation effort for participation in the partnership and therefore data provision had to be supported with services. The paper provides insides of the work done, the findings and results. The briefly described demonstrator illustrates the usage of the service prototypes and early application cases.

Human-Robot Collaboration combines the reliability of robots with human adaptability. It is a prime candidate to respond to the trend of Mass Customization which requires frequent reconfiguration with variable lot sizes. But the close contact between humans and robots creates new safety risks, and ergonomic factors like robot-induced stress need to be considered. Therefore we propose a human-centric Digital Twin framework, where information about the human is stored and processed in a dedicated Digital Twin and can be transmitted to the robot’s Digital Twin for human-aware adaptations. We envision and briefly discuss three possible applications. Our framework has the potential to advance collaborative robotics but inherits technical challenges that come with Digital Twin based approaches and human modelling.

The pharmaceutical industry continually seeks advancements to improve manufacturing processes, ensuring product quality, regulatory compliance, and operational efficiency. Lyophilization, a critical process for preserving pharmaceuticals and biological materials, necessitates innovative solutions for optimization and risk mitigation. This research investigates the cost-benefit analysis of implementing a Digital Twin for a pharmaceutical lyophilization plant, focusing on system resilience. The research addresses disruption scenarios affecting the manufacturing process and evaluates the potential benefits and costs associated with Digital Twin integration. The methodology encompasses data collection, system resilience assessment, and cost-benefit analysis. The outcomes indicate the transformative potential of Digital Twin technology in enhancing operational resilience and reducing disruption rates.