Search

The still prevalent use of paper conformity lists in the automotive industry has a serious negative impact on the performance of quality control inspectors. We propose instead a hybrid quality inspection system, where we combine automated detection with human feedback, to increase worker performance by reducing mental and physical fatigue, and the adaptability and responsiveness of the assembly line to change. The system integrates the hierarchical automatic detection of the non-conforming vehicle parts and information visualization on a wearable device to present the results to the factory worker and obtain human confirmation. Besides designing a novel 3D vehicle generator to create a digital representation of the non conformity list and to collect automatically annotated training data, we apply and aggregate in a novel way state-of-the-art domain adaptation and pseudo labeling methods to our real application scenario, in order to bridge the gap between the labeled data generated by the vehicle generator and the real unlabeled data collected on the factory floor. This methodology allows us to obtain, without any manual annotation of the real dataset, an example-based F1 score of 0.565 in an unconstrained scenario and 0.601 in a fixed camera setup (improvements of 11 and 14.6 percentage points, respectively, over a baseline trained with purely simulated data). Feedback obtained from factory workers highlighted the usefulness of the proposed solution, and showed that a truly hybrid assembly line, where machine and human work in symbiosis, increases both efficiency and accuracy in automotive quality control. [ABSTRACT FROM AUTHOR]

The pulp and paper industry, which is closely related to national economic and social development, is an important industry but also contributes high carbon emissions. Therefore, with the advent of the low-carbon economic era, ways to reduce the carbon emissions and to bring about a low-carbon industrial transition of the pulp and paper industry is becoming one of the important academic projects. A system for carbon footprint assessment, namely the Publicly Available Specification (PAS) 2050 methodology, is introduced in this paper. Based on the analysis and assessment of the carbon footprint (CO2 equivalent emissions) for the Coated Ivory Board production lines, it was used to provide a scientific basis and approach for reduction of carbon emissions and formulate the corresponding measures for carbon emissions reduction of China's pulp and paper industry. The business to business carbon footprint, for which steps of the life cycle are included in Coated Ivory Board production, was analyzed and calculated. The results showed that there were 888 kg of CO2 equivalent emissions per metric ton of Coated Ivory Board, in which the largest part, accounting for 57.5%, was associated with purchased electricity, followed by fuel oil at 40.2%, and others accounted for 2.3% of the CO2 equivalent emissions. [ABSTRACT FROM AUTHOR]

This paper considers the optimisation of Mixed-model assembly lines with continuous paced line control. The two minimisation goals have a mixed-integer linear multi-objective dispute. The paper proposes a criterion-space method to define the Pareto front for this class of problems. The method combines Pareto fronts obtained from integer solutions and gradually refines them until the instance's global front is determined. Comparing paced to unpaced line controls can be challenging, since they can produce the same cycle time given sufficiently long line lengths or buffers. Hence, determining Pareto fronts between cycle time and line length for paced lines allows meaningful comparisons between line controls. An industrial case study shows that line length acts as continuously distributable buffers for paced lines, leading to weaker diminishing returns. This result suggests that paced lines are more efficient than unpaced ones for lower cycle time ranges. [ABSTRACT FROM AUTHOR]

This paper discusses a digital twin-based approach for designing and redesigning flexible assembly systems. The digital twin allows modeling the parameters of the production system at different levels including assembly process, production station, and line level. The approach allows dynamically updating the digital twin in runtime, synthesizing data from multiple 2D–3D sensors in order to have up-to-date information about the actual production process. The model integrates both geometrical information and semantics. The model is used in combination with an artificial intelligence logic in order to derive alternative configurations of the production system. The overall approach is discussed with the help of a case study coming from the automotive industry. The case study introduces a production system integrating humans and autonomous mobile dual arm workers. [ABSTRACT FROM AUTHOR]

Industrial manufacturing is a core component of a nation's economic wealth. It is driven by technology, and its paradigm has shifted from craft production to mass production to lean production to mass customisation (MC). We identify how emerging technologies offer a new way to shift MC research focusing on the scientific communities that have been driving MC research, changes in MC throughout the years, and the technological advances that influence future research trends of MC. We then identify MC paradigms. We analyse 1,947 MC papers from 1992 to 2019 and discover the research attention changes from the perspectives of the communities and themes using topic modelling. Our result indicates that MC was explained in terms of the business and marketing strategy (BMS) and the operations management (OM), in accordance with technological development. In BMS, MC topics moved from demand uncertainty to industry transformation. In OM, topics changed from assembly line for product differentiation to additive manufacturing. We discuss the future research agenda according to the technological advances in computer-aided design, additive manufacturing, machine learning, identification technologies, virtual reality, and human-robot collaboration. Not only researchers but practitioners and policymakers can utilise our approach to analyse and formulate MC strategies considering new technologies. [ABSTRACT FROM AUTHOR]

Purpose: This study aims to investigate how to identify and address production levelling problems in assembly lines utilising an intensive manual workforce when higher productivity levels are urgently requested to meet market demands. Design/methodology/approach: A mixed-methods approach was used in the research design, integrating case study analysis, interviews and qualitative/quantitative data collection and analysis. The methodology implemented also introduces to the literature on operational performance a novel combination of data analysis methods by introducing the use of the Natural Language Understanding (NLU) methods. Findings: First, the findings unveil the impacts on operational performance that transportation, limited documentation and waiting times play in assembly lines composed of an intensive workforce. Second, the paper unveils the understanding of the role that a limited understanding of how the assembly line functions play in productivity. Finally, the authors provide actionable insights into the levelling problems in manual assembly lines. Practical implications: This research supports industries operating assembly lines with intensive utilisation of manual workforce to improve operational performance. The paper also proposed a novel conceptual model prescriptively guiding quick and long-term improvements in intensive manual workforce assembly lines. The article assists industrial decision-makers with subsequent turnaround strategies to ensure higher efficiency levels requested by the market. Originality/value: The paper offers actionable findings relevant to other manual assembly lines utilising an intensive workforce looking to improve operational performance. Some of the methods and strategies examined in this study to improve productivity require minimal capital investments. Lastly, the study contributes to the empirical literature by identifying production levelling problems in a real context. [ABSTRACT FROM AUTHOR]

For manual picking processes in warehouses and assembly line feeding, collaborative robots – cobots – hold a potential to support operators and, thereby, enhance performance. However, studies focused on the economic aspects of cobot applications have been scarce. The present paper aims at modelling a new picking system, in which human operators collaborate with cobots during picking (or kitting) activities. It is considered the case of having a cobot on board of the picking trolley or cart, working in parallel with the picker by performing sorting of the picked items. The paper focuses on the economic aspects of this application and presents a model of the relative cost difference between a manual and a cobot-supported process, accounting for the costs associated with operators, equipment, and quality. The relative cost difference is analysed in a numerical example. The findings suggest that the decision to use a cobot sorting mode is robust when there is considerable sorting work to carry out, for example, when there are higher yearly order volumes to handle, and when higher order commonality among orders processed as a batch can be achieved. Future research should focus on implementation challenges and safety issues associated with cobot-supported picking processes. [ABSTRACT FROM AUTHOR]

Feasibility checking is a step in manufacturing system engineering for verifying the normalization and effectiveness of a manufacturing system associated with a specific configuration of resources and processes. It enables factory operators to predict problems before operational time, thus preventing equipment and machinery accidents and reducing labor waste in physically organizing the shop floor. In Industry 4.0, feasibility checking becomes even more critical since emerging challenges, such as mass personalization, require reconfiguring work cells quickly and flexibly on demand. Regarding this need, digital twin technologies have emerged as an ideal candidate for practicing feasibility checking. Indeed, they are tools used to implement digital representations of manufacturing entities that can constitute a digital environment and context. Factory operators can test a manufacturing process within a digital environment in different contexts before the execution with physical resources. This approach currently receives significant attention from the manufacturing community; however, there is still a lack of sharing experiences to implement it. Thus, this paper contributes a methodology to engineer a digital environment and context for a manufacturing work cell using AAS digital twins and physics-based 3D digital twins technologies. Technically, this methodology is a specific case of N-DTs, a general methodology for engineering heterogeneous digital twins. The product assembly line case study, also presented in this paper, is a successful experiment applying the above contributions. The two methodologies and the case study can be helpful references for both public and private sectors to deploy their feasibility-checking frameworks and deal with heterogeneous digital twins in general. [ABSTRACT FROM AUTHOR]

Classical assembly line balancing (ALB) models assume constant cycle times during production. However, this assumption oversimplifies the actual situation, especially in small batch production of up to a few hundred units, since employees can significantly improve their performance thanks to the learning effect, causing task times to decrease. Several researchers have realised the importance of the effect of learning in ALB. However, only a limited number of papers have so far addressed this issue. This is problematic, since ignoring the learning effect in ALB may lead to inaccurate results and by extension misleading conclusions. This study summarises the main contributions in the field of ALB that focus on the learning effect. First, assembly lines (ALs) and ALB problems are characterised. Next, the importance of the learning effect in ALB is highlighted, and the main learning curve (LC) models are introduced. Finally, an exhaustive review of the main contributions in the field of ALB and learning effect is provided. The results highlight that many problems in this area need to be investigated further, in relation to both conceptual model building and the development of algorithms for solving practical size problems. [ABSTRACT FROM AUTHOR]

Collaborative robots (cobots) are important components of the Industry 4.0 paradigm and smart manufacturing. Cobots are known for their ability to interact with the operators in a shared workspace. Due to their spread in the last decade, cobot research proliferated. However, most individual studies focused on specific aspects of cobot deployment, and only scant attention was given to their evaluation (mostly not based on productivity criteria). Thus, better support is needed for cobot acquisition and deployment decisions. This paper answers this need by presenting a summary of the major considerations related to cobots acquisition and deployment, and providing a productivity analysis procedure that supports cobot acquisition and deployment decisions. Defining the cobots' required characteristics and capabilities, effectively narrows the possible selection of cobots. However, it does not give information as to the economic value of acquiring and deploying a specific cobot. So, in addition to cobots' characteristics and capabilities, this paper presents computational techniques to analyse and support this decision for a single workstation per se, and for a station in an assembly line. The difference between these two cases is discussed and analysed, and corresponding models are presented with computational examples. [ABSTRACT FROM AUTHOR]

Purpose: This paper proposes a new metric, named Operational Space Efficiency (OpSE), intended to diagnose and quantify the inefficient use of floor space for stocking materials in industrial workstations. OpSE presents a formulation analogous to the well-known Overall Equipment Effectiveness and can be obtained as the product of three distinct indicators: Standard Compliance Effectiveness, Standards Selection Effectiveness and Design Space-usage Effectiveness. Design/methodology/approach: This indicator scrutinizes how usefully floor space in workstations is used to temporarily stock materials in the form of raw materials, semi-finished products, parts and components. It is suited for analyzing fixed-position layouts as well as product layouts typical of repetitive manufacturing settings, such as assembly lines in the automotive sector. The proposed indicator leverages an appropriate loss structure that features those factors affecting floor space utilization in workstations with regard to supplying and stocking materials. Findings: An Italian manufacturer in the field of electro-technology was used as an industrial case study for the application of the methodology. The application shows how the three indicators work in practice, the effectiveness of OpSE and the methodology as a whole, in diagnosing floor space usage inefficiencies and in properly addressing improvement actions of the internal logistics in industrial settings. Originality/value: The paper scrutinizes some important Key Performance Indicators (KPIs) dealing with space usage efficiency and identifies some significant drawbacks. Then it suggests a new, inclusive structure of losses and a KPI that not only measures efficiency but also allows to identify viable countermeasures. [ABSTRACT FROM AUTHOR]

This paper provides a literature review and an analysis of the studies related to workforce reconfiguration strategies as a part of workforce planning for various production environments. The survey demonstrates that these strategies play a crucial role in the resilience and flexibility of manufacturing systems since they help industrial companies to quickly adapt to frequent changes in demand both in terms of volume and product mix. Five strategies are considered: the use of utility, temporary, walking, cross-trained workers, and bucket brigades. They are analysed in the context of mixed and multi-model manual assembly lines, dedicated, cellular, flexible, and reconfigurable manufacturing systems. The review shows that most of the researches on these reconfiguration strategies focus on multi- or mixed-model assembly lines. At the same time, few studies consider workers team reconfiguration in flexible and reconfigurable manufacturing systems. Finally, this paper reveals several promising research directions in workforce reconfiguration planning, namely, the use of both machine and workforce reconfigurations, consideration of the ergonomic aspects, the combination of multiple workforce reconfiguration strategies, the study of workforce reconfiguration in human-robot collaborative systems, and the use of new technologies in human-machine industrial environments. [ABSTRACT FROM AUTHOR]

The solution to the assembly line balancing with the hierarchical worker assignment problem (ALBHWP) provides the optimal allocation of workers and tasks to the stations that minimise the total worker cost. In the ALBHWP, tasks differ in terms of the qualification requirements of workers, and the qualification levels of workers are hierarchical. In the hierarchical workforce structure, a lower qualified worker can be replaced by higher qualified ones with higher costs, while the vice versa is not applicable. This problem has only been studied for straight assembly lines so far. In this paper, we introduce the ALBHWP for U-shaped assembly lines. We developed integer and constraint programming models for solving the ALBHWP and compared their effectiveness using an extensive set of benchmark instances. We solved the ALBHWP for straight and U-shaped assembly lines comparatively. Constraint programming models have been statistically proven to provide better quality solutions faster than integer programming models. Besides, the CP model outperforms the only available metaheuristic in the literature for the S-ALBHWP in almost all problem sizes. Another observation is that a U-shaped line design is more cost-effective than a straight line design, but solving the ALBHWP for U-shaped lines is more difficult regarding computational complexity. [ABSTRACT FROM AUTHOR]

Assembly processes play a big role in the current business context as global supply chains depend on many subcomponents to produce a single finished product. Previous studies have shown contrasting results regarding the effect that supply variability (the variability of feeding stations) has on the performance of assembly systems, as opposed to the variability of the station matching and assembling the components. This paper aims to close this gap by studying the behaviour of simple assembly systems with differing degrees of variability allocation among the stations through an experimental simulation study. Results suggest that a reduction in feeding station variability results in higher throughput, even in systems where the variability of one of the feeding stations increases while the other decreases. Furthermore, in scenarios with high total variance, the highest throughput is reached by transferring both variance and work from one of the feeding stations to any other station, whereas in low variance systems symmetrical work transfer to the feeding stations results in the highest throughput, as previously shown. Finally, reducing feeding station variability decreased the time spent in the assembly station (waiting time for component matching plus time for the assembly operation) only in experiments with high total variance. [ABSTRACT FROM AUTHOR]

In an era where countries worldwide are emphasizing the process of educational informatization, traditional experimental teaching methods encounter inherent limitations, such as constraints related to the availability of experimental sites and the amount of experimental equipment. Consequently, it becomes a challenge to provide students with an unconstrained, open, and collaborative experimental environment. The digital twin, as a pivotal technology for achieving interactive mapping of the physical world and the information world, possesses attributes characterized by real time interactivity and the seamless fusion of virtual and tangible elements. As a result, various virtual simulation experimental teaching platforms grounded in digital twin technology are now extensively integrated into talent development initiatives. In the context of the "Production Planning and Control" course at Harbin Institute of Technology (HIT), the virtual simulation experimental teaching method is examined, leading to the creation of a virtual simulation instructional platform that blends the virtual and tangible realms, thanks to the foundation of digital twin technology. The virtual simulation experimental teaching platform detailed in this paper, specifically designed for optimizing scheduling within assembly line workshops, represents a pioneering model practice in the integration of digital twin technology into workshop-based experimental instruction and assumes an influential role in enhancing students' grasp of theoretical knowledge and their capacity for practical innovation. [ABSTRACT FROM AUTHOR]

Purpose: Various approaches and algorithms have been proposed since the 1950s to solve the assembly line (AL) balancing problem. These methods have established an AL configuration from the beginning. However, a prebalanced AL may have to be rebalanced in real life for many reasons, such as changes in the cycle time, production demand, product features or task operation times. This problem has increasingly attracted the interest of scientists in recent years. This study aims to offer a detailed review of the assembly line rebalancing problems (ALRBPs) to provide a better insight into the theoretical and practical applications of ALRBPs. Design/methodology/approach: A structured database search was conducted, and 41 ALRBP papers published between 2005 and 2022 were classified based on the problem structure, objective functions, problem constraints, reasons for rebalancing, solution approaches and type of data used for solution evaluation. Finally, future research directions were identified and recommended. Findings: Single model, straight lines with deterministic task times were the most studied type of the ALRBPs. Eighteen percent of the studies solved worker assignment problems together with ALRBP. Product demand and cycle time changes were the leading causes of the rebalancing need. Furthermore, seven future research opportunities were suggested. Originality/value: Although there are many review studies on AL balancing problems, to the best of the authors' knowledge, there have been no attempts to review the studies on ALRBPs. [ABSTRACT FROM AUTHOR]

This paper proposes new algorithms for the assembly line balancing problem with hierarchical worker assignment (ALBHW). The ALBHW appears in real industrial contexts, where companies deal with a multi-skilled workforce. It considers task execution times that vary depending on the worker type to whom the task is assigned. Qualification levels among workers are ranked hierarchically, where a lower qualified worker costs less but requires larger execution times then a higher qualified one. The aim is to assign workers and tasks to the stations of an assembly line, in such a way that cycle time and precedence constraints are satisfied, and the total cost is minimised. In this paper, we first present a mathematical model and improve it with preprocessing techniques. Then, we propose a constructive heuristic and a variable neighbourhood descent that are useful to solve large instances. Extensive computational experiments on benchmark instances prove the effectiveness of the algorithms. [ABSTRACT FROM AUTHOR]

This paper investigates the game-changing paradigm in modern manufacturing, known as the "smart factory," from an organizational perspective. Factories get smarter—or smartized—by becoming more data-intensive and more tightly integrating operations with management. We design a unique survey to measure this intangible organizational capital accumulation process for 939 Korean manufacturing plants. Observing heterogeneous levels of smartization across factories, we explore their relationship with three performance dimensions: productivity, cost efficiency, and product variety. The drivers of smartization are also identified. Our analysis uncovers a strong association between smartization and heightened productivity, with varying interactions with other performance metrics based on the specific manufacturing process (e.g. lower defect rate only in assembly line process). Importantly, we find that investments in digital technologies drive smartization only when complemented by structured incentive management and proper chief executive officer (CEO) leadership. [ABSTRACT FROM AUTHOR]

This paper provides a novel methodology for human-driven decision support for capacity allocation in labour-intensive manufacturing systems. In such systems (where output depends solely on human labour) it is essential that any changes aimed at improving productivity are informed by the workers' actual working practices, rather than attempting to implement strategies based on an idealised representation of a theoretical production process. This paper reports how worker position data (obtained by localisation sensors) can be used as input to process mining algorithms to generate a data-driven process model to understand how manufacturing tasks are actually performed and how this model can then be used to build a discrete event simulation to investigate the performance of capacity allocation adjustments made to the original working practice observed in the data. The proposed methodology is demonstrated using a real-world dataset generated by a manual assembly line involving six workers performing six manufacturing tasks. It is found that, with small capacity adjustments, one can reduce the completion time by 7% (i.e., without requiring any additional workers), and with an additional worker a 16% reduction in completion time can be achieved by increasing the capacity of the bottleneck tasks which take relatively longer time than others. [ABSTRACT FROM AUTHOR]

An important and highly complex process in the automotive industry is the balancing of the assembly lines. Optimally distributing jobs among the lines in order to obtain the highest efficiency is mostly done manually, taking a lot of time. This paper aims to automate the process of line balancing fora real-world test case. Automotive assembly lines are highly complex, and multiple factors have to be considered while balancing the lines. All factors relevantin a case study at VDL Nedcar are considered, namely, mixed-model production, sequence-dependent setup times, variable workplaces with multiple operators and multiple assignment constraints. A Genetic Algorithm (GA) is proposed to solve the formulated balancing problem and to act as a decision support system. Results on newly proposed benchmark instances show that the solution is dependent on the relation between the takt time and processing time of jobs, as well as the setup times. In addition, results of a real-life case study show that the proposed GA is effective in balancing a real-world assembly line and that it can both increase the efficiency of the line and decrease the variance in operating time between all model variants when compared to current practice. [ABSTRACT FROM AUTHOR]

Assembly lines are production flow systems wherein activities are organised around a line consisting of various workstations through which the product flows. At each station, the product is assembled through a subset of operations. The assembly line balancing problem (ALBP) consists of allocating operations between stations to maximise the system efficiency. In this study, a variable-depth local search algorithm is proposed for solving simple assembly line balancing problems (SALBPs), which are the most widely studied versions of the ALBP. Although the state-of-the-art techniques for solving the SALBP consist of exact enumeration-based methods or heuristics, this paper proposes a local search-based heuristic using variable-length sequences that allow the solution space to be efficiently explored. The proposed algorithm improves the best solution known for multiple instances reported in the literature, indicating that its efficiency is comparable to those of the state-of-the-art method for solving the SALBP. Moreover, the characteristics of the instances for which the proposed procedure provides a better solution than previously reported construction procedures are investigated. [ABSTRACT FROM AUTHOR]

As an enabling technology and means to practice the concept of intelligent manufacturing, digital twin has attracted extensive attention from both industry and academia. In view of the lack of concrete implementation methods of digital twin robot assembly lines, and the complex implementation process, and long development cycle using existing methods, the implementation of digital twin robot assembly lines is a significant research problem. In view of this, this paper proposes an architecture as well as an implementation method of a digital twin system for a robot assembly line. Taking a robot assembly production line as an example, a digital twin architecture and modular development approach are proposed, and three key technologies for implementing the digital twin system, namely virtual model construction, real-time data acquisition and accurate virtual-real mapping, are elaborated. The architecture and implementation method can quickly and effectively facilitate building a digital twin system of the robot assembly line, and realizing real-time monitoring and intelligent analysis of all elements of the physical robot production line. [ABSTRACT FROM AUTHOR]

In the mixed-model assembly line balancing and sequencing problem (MALBSP), workstations are assumed to be constantly available. The failure of any workstation will make the entire assembly line stop working. Preventive maintenance (PM) is a way to maintain the workstation before its failure, reduce unexpected downtime, and prolong its useful life. Previous studies have considered PM scenarios (PMS) in the simple and U-shaped assembly line to improve production efficiency and smoothness effectively, but not in the mixed-model assembly line. This paper fills this research gap, and the MALBSP considering PMS (MALBSP_PMS) is studied in this paper. A mixed-integer linear programming model is proposed to minimize makespan and task alteration. A migrating birds optimization algorithm is improved (IMBO) to obtain well-distributed Pareto frontier solutions. This algorithm designs a restart mechanism and an intra-population crossover operator to avoid falling into the local optimal and enhance its searchability. Experimental results demonstrate the effectiveness of two improvements and the IMBO algorithm. In addition, a real-world case study is introduced to illustrate the importance of considering PM scenarios in MALBSP. [ABSTRACT FROM AUTHOR]

For a company to make a profit and satisfy customers' needs, it must have the correct individual processes - the production process is one of the most important. This paper proposes a solution to eliminate the problem based on identifying the cause of the blocking - idle times in several workplaces of the production line. Another goal was to determine whether increasing productivity and the number of products on the given line is possible. At the beginning of the research, a system analysis of production operations was carried out. A simulation was chosen as a tool for solving the goals. Tecnomatix Plant Simulation creates the simulation model. The model was verified by a simulation experiment that simulated the current state, and the data obtained by the analysis were used. The experiment was performed for three simulated times: 8, 16 and 48 hours. After the verification of the model, experiments were performed on the models. The paper presents the results of three experiments for a simulated time of 48 hours. Based on the experiments, it was found that in the case of shortening the work cycle at critical workplaces No.5 and No.15, it is possible to make the process more efficient - to equalise the workload of individual workplaces and increase production by 25%. [ABSTRACT FROM AUTHOR]

Mixed and multi-model assembly line sequencing problems are more practical than single-product models. The methods and selection criteria used must keep up with the constantly increasing level of variability, synchronize flows between various—often very energy-intensive production departments—and cope with high dynamics resulting from interrupted supply chains. The requirements for conscious use of Earth's limited natural resources and the need to limit energy consumption and interference in the environment force the inclusion of additional evaluation criteria focusing on the environmental aspect in optimization models. Effective sustainable solutions take into account productivity, timeliness, flow synchronization, and the reduction of energy consumption. In the paper, the problem of determining the sequence of vehicles for a selected class of multi-version assembly lines, in which the order restrictions were determined taking into account the above criteria, is presented. Original value of the paper is the development of the Grey Wolf Optimizer (GWO) for the mixed-model assembly lines sequencing problem. In the paper, a comparative analysis of the greedy heuristics, Simulated Annealing and GWO for a real case study of a mixed vehicle assembly line is presented. The GWO outperforms other algorithms. Overall research performance of the GWO on the sequencing problem is effective. [ABSTRACT FROM AUTHOR]

In the field of industry, especially in the production areas, it is particularly important that the monitoring of assembly efficiency takes place in real-time mode, and that the related data-based estimation also works quickly and reliably. The Manufacturing Execution System (MES), Enterprise Resource Planning (ERP) and Customer Relationship Management (CRM) systems used by companies provide excellent support in data recording, processes, and storing. For Overall Equipment Effectiveness (OEE) data showing the efficiency of assembly lines, there is a regular need to determine expected values. This paper focuses on OEE values prediction with Multiple Linear Regression (MLR) as supervised machine learning. Many factors affecting OEE (e.g., downtimes, cycle time) are examined and analyzed in order to make a more accurate estimation. Based on real industrial data, we used four different methods to perform prediction with various machine learning algorithms, these were the cumulative, fix rolling horizon, optimal rolling horizon and combined techniques. Each method is evaluated based on similar mathematical formulas. [ABSTRACT FROM AUTHOR]

Assembly lines (conveyors) are traditional means of large-scale and mass-scale productions. An assembly line balancing problem is needed for optimizing the assembly process by configuring and designing an assembly line for the same or similar types of final products. This problem consists of designing the assembly line and distributing the total workload for manufacturing each unit of the fixed product to be assembled among the ordered workstations along the constructed assembly line. The assembly line balancing research is focused mainly on simple assembly line balancing problems, which are restricted by a set of conditions making a considered assembly line ideal for research. A lot of published research has been carried out in order to describe and solve (usually heuristically) more realistic generalized assembly line balancing problems. Assembly line designing, balancing and scheduling problems with not deterministic (stochastic, fuzzy or uncertain) parameters have been investigated in many published research works. This paper is about the design and optimization methods for assembly and disassembly lines. We survey the recent developments for designing, balancing and scheduling assembly (disassembly) lines. New formulations of simple assembly line balancing problems are presented in order to take into account modifications and uncertainties characterized by real assembly productions. [ABSTRACT FROM AUTHOR]

A newly emerging mass-individualisation concept has attracted increasing attention in recent years. However, this concept increases the complexity of manufacturing systems within organisations. In such systems, one of the main challenges is the sequencing problem, especially in dynamic environments where unpredictable events demand new constraints. In this context, the ability to use real-time data to make efficient, quick decisions has become one of the main priorities of managers. In this paper, based on a real-world case from Fiat Powertrain Technologies, we define a dynamic mixed-model assembly line sequencing problem with walking workers. In this context, each worker is assigned to a product for all assembly operations and moves from one station to another. A mathematical model is proposed to minimise production time. Since the problem is NP-hard, a hyper-heuristic is also developed to solve the problem. Moreover, a simulation-optimisation model is developed using FlexSim software to solve a real-world problem in a dynamic environment. Comparison of the results illustrates the effectiveness of using the simulation approach to dynamically solve such problems, especially in real-world cases. Finally, a thorough description of managerial insights is provided to indicate the applicability of the proposed approach. [ABSTRACT FROM AUTHOR]

Since global warming and the needs for sustainable production models, this paper focuses on a Just-in-Time (JIT)-based sustainable material handling scheduling problem (JSMHSP) with time window and capacity constraints for mixed-model assembly lines in the automobile industry. A novel Hybrid-load Automated Guided Vehicle (H-AGV) is proposed to fulfil material handling tasks between supermarkets and assembly lines. The motivation is to minimise the total line-side inventory and the total energy consumption, which corresponds to JIT and environmental objectives. Due to the NP-hard nature of the proposed scheduling problem, a Deep Q network and Non-dominated sorting-based Hyper-Heuristic (DN-HH) algorithm is presented to solve the bi-objective scheduling problem, which benefits from the synergy of the Deep Q Network (DQN) and Hyper-Heuristic (HH). In the DQN, the states and rewards are designed according to the characteristics of the scheduling problem. To improve the performance of DQN, the experience pool (EP) and the target network are presented to improve the convergence speed. Computational results reveal that the proposed DN-HH algorithm outperforms the other two compared algorithms in both solution quality and convergence speed and the performance of the H-AGV is better than that of the other two types of AGVs. [ABSTRACT FROM AUTHOR]

In advanced manufacturing, optimizing mixed-model synchronous assembly lines (MMALs) is crucial for enhancing productivity and adhering to sustainability principles, particularly in terms of energy consumption and energy-efficient sequencing. This paper introduces a novel approach by categorizing sequence-dependent setup times into bipartite categories: workpiece-independent and workpiece-dependent. This strategic division streamlines assembly processes, reduces idle times, and decreases energy consumption through more efficient machine usage. A new mathematical model is proposed to minimize the intervals at which workpieces are launched on an MMAL, aiming to reduce operational downtime that typically leads to excessive energy use. Given the Non-deterministic Polynomial-time hard (NP-hard) nature of this problem, a genetic algorithm (GA) is developed to efficiently find solutions, with performance compared against the traditional branch and bound technique (B&B). This method enhances the responsiveness of MMALs to variable production demands and contributes to energy conservation by optimizing the sequence of operations to align with energy-saving objectives. Computational experiments conducted on small and large-sized problems demonstrate that the proposed GA outperforms the conventional B&B method regarding solution quality, diversity level, and computational time, leading to energy reductions and enhanced cost-effectiveness in manufacturing settings. [ABSTRACT FROM AUTHOR]

A 1 millisecond (1‐ms) vision system that guarantees high efficiency and timely response for tomato defect detection is essential for factory automation. Because of various defect appearances, recently many existing researches focus on CNN based defect detection, but few of them attempt to reach high processing speed to adapt to the factorial assembly line. This paper proposes a global to multi‐scale local based parallel architecture with hardwired CNN for tomato defect detection. This architecture breaks down image‐wise detection into pixel‐wise localization and block‐wise classification. The pixel‐wise localization utilizes tomato‐aware information as constraints for localization performance. The block‐wise classification uses a fully pipelined network structure to obtain the classification result for each block as the pixel stream moves through the network. The classification network has a six‐layer lightweight network structure with quantization for hardwired type implementation on FPGA. The experiment results show that the proposed architecture processes 1000 FPS images with 0.9476 ms/frame delay. And for detection performance, this architecture keeps f1−score$f_{1}-score$ at 80.18%, only 1.31% lower than ResNet50 based detection system. [ABSTRACT FROM AUTHOR]

As the take-off of China's macro economy, as well as the rapid development of infrastructure construction, real estate industry, and highway logistics transportation industry, the demand for heavy vehicles is increasing rapidly, the competition is becoming increasingly fierce, and the digital transformation of the production line is imminent. As one of themost important components of heavy vehicles, the transmission front andmiddle case assembly lines have a high degree of automation, which can be used as a pilot for the digital transformation of production. To ensure the visualization of digital twins (DT), consistent control logic, and real-time data interaction, this paper proposes an experimental digital twin modeling method for the transmission front and middle case assembly line. Firstly, theDT-based system architecture is designed, and theDTmodel is created by constructing the visualization model, logic model, and data model of the assembly line. Then, a simulation experiment is carried out in a virtual space to analyze the existing problems in the current assembly line. Eventually, some improvement strategies are proposed and the effectiveness is verified by a new simulation experiment. [ABSTRACT FROM AUTHOR]

Predicting anomalies in manufacturing assembly lines is crucial for reducing time and labor costs and improving processes. For instance, in rocket assembly, premature part failures can lead to significant financial losses and labor inefficiencies. With the abundance of sensor data in the Industry 4.0 era, machine learning (ML) offers potential for early anomaly detection. However, current ML methods for anomaly prediction have limitations, with F1 measure scores of only 50% and 66% for prediction and detection, respectively. This is due to challenges like the rarity of anomalous events, scarcity of high-fidelity simulation data (actual data are expensive), and the complex relationships between anomalies not easily captured using traditional ML approaches. Specifically, these challenges relate to two dimensions of anomaly prediction: predicting when anomalies will occur and understanding the dependencies between them. This paper introduces a new method called Robust and Interpretable 2D Anomaly Prediction (RI2AP) designed to address both dimensions effectively. RI2AP is demonstrated on a rocket assembly simulation, showing up to a 30-point improvement in F1 measure compared to current ML methods. This highlights its potential to enhance automated anomaly prediction in manufacturing. Additionally, RI2AP includes a novel interpretation mechanism inspired by a causal-influence framework, providing domain experts with valuable insights into sensor readings and their impact on predictions. Finally, the RI2AP model was deployed in a real manufacturing setting for assembling rocket parts. Results and insights from this deployment demonstrate the promise of RI2AP for anomaly prediction in manufacturing assembly pipelines. [ABSTRACT FROM AUTHOR]

Many of the reports created at assembly lines, where all components of an aircraft are installed, frequently indicate that errors threaten safety. The proposed methodology in this study evaluates error prediction and risk mitigation to prevent failures and their consequences. The results linked to a typical electrical harness manufacture of a military aircraft estimated reductions of 93% in time and 90% in error during the creation of engineering manufacturing processes using AI techniques. However, traditional risk assessments methods struggle to identify and mitigate errors effectively. Thus, developing an advanced methodology to ensure systems safety is needed. This paper addresses how innovative AI technology solutions can overcome these challenges, mitigate error risks, and enhance safety in aerospace. Technologies, such as artificial intelligence, predictive algorithms, machine learning, and automation, can play a key role in enhancing safety. The aim of this study is to develop a model that considers the factors that can potentially contribute to error creation, through an artificial intelligence (AI) approach. The specific AI techniques used such as support vector machine, random forest, logistic regression, K-nearest neighbor, and XGBoost (Python 3.8.5) show good performance for use in error mitigation. We have compared the modeled values obtained in this study with the experimental ones. The results confirm that the best metrics are obtained by using support vector machine and logistic regression. The smallest deviation between the measured and modeled values for these AI methods do not exceed 5%. Furthermore, using advancements in machine learning methods can enhance error mitigation in aerospace. The use of AutoML can play a key role in automatically finding an appropriate model which provides the best performance metrics and therefore the most reliable forecast for data prediction and error mitigation. [ABSTRACT FROM AUTHOR]

In the context of assembly lines, the term 'rest time' refers to those time of inactivity allowed to a worker after the completion of his/her tasks on a product. Rest times are often assigned to workers that perform ergonomically hazardous tasks, to prevent their occupational diseases. When rest times are assigned after balancing the line, the line becomes unbalanced and the ergonomic load of workers is reduced at the price of deteriorated line performance. In this paper, an approach able to assign rest times during the balancing procedure is presented. The approach consists of a genetic algorithm coupled with a discrete event simulator to evaluate the fitness function of individuals, which considers the performance, the ergonomic, and the cost aspects. The OCRA index has been utilised to assess the ergonomic risk of workers. Results show that the presented approach can significantly support firms in designing manual assembly lines: the possibility to assign rest times during the balancing procedure allows for increasing the quality of solutions in 103 out of 135 benchmark instances. The greatest improvements are obtained when the ergonomic load of tasks is relevant, by achieving good performances of the line with limited costs, thus defending companies' profitability. [ABSTRACT FROM AUTHOR]

In this paper, a new mixed integer programming (MIP) formulation is developed for balancing and scheduling of mixed model assembly lines with disjunctive precedence constraints among assembly tasks. To represent alternative precedence relations, AND/OR assembly graph was adopted. In case of alternative precedence relations, for each product multiple assembly plans exist, which can be represented by a set of alternative precedence subgraphs and only one of such subgraphs should be selected for each product. As the number of subgraphs exponentially increases with the number of disjunctive relations among the tasks, the computational complexity of simultaneous balancing and scheduling along with the assembly subgraph selection increases with the number of alternative precedence relations. Unlike the other MIP approaches known from the literature, the new model does not need the alternative assembly subgraphs to be to explicitly enumerated as input data and then used for indexing the variables. Instead, a new disjunctive precedence selection and task assignment variable and new constraints are introduced to optimally choose one relation for each subset of alternative precedence relations. The optimal solutions for computational examples of balancing and scheduling problems illustrate a superior performance of the new modelling approach. [ABSTRACT FROM AUTHOR]

Cyber-physical systems (CPS) play an important role in the implementation of new Industry 4.0 solutions, acting as the backbone infrastructure to host distributed intelligence capabilities and promote the collective intelligence that emerges from the interactions among individuals. This collective intelligence concept provides an alternative way to design complex systems with several benefits, such as modularity, flexibility, robustness, and reconfigurability to condition changes, but it also presents several challenges to be managed (e.g., non-linearity, self-organization, and myopia). With this in mind, this paper discusses the factors that characterize collective intelligence, particularly that associated with industrial CPS, analyzing the enabling concepts, technologies, and application sectors, and providing an illustrative example of its application in an automotive assembly line. The main contribution of the paper focuses on a comprehensive review and analysis of the main aspects, challenges, and research opportunities to be considered for implementing collective intelligence in industrial CPS. The identified challenges are clustered according to five different categories, namely decentralization, emergency, intelligent machines and products, infrastructures and methods, and human integration and ethics. Although the research indicates some potential benefits of using collective intelligence to achieve the desired levels of autonomy and dynamic adaptation of industrial CPS, such approaches are still in the early stages, with perspectives to increase in the coming years. Based on that, they need to be further developed considering some main aspects, for example, related to balancing the distribution of intelligence by the vertical and horizontal dimensions and controlling the nervousness in self-organized systems. [ABSTRACT FROM AUTHOR]

Embedded systems are the best solution to achieve high-performance edge terminal computing tasks. With the rapid increase in the amount of data generated by edge devices, it is imperative to implement intelligent algorithms with large amounts of data and computation on embedded terminal systems. In this paper, a novel multi-core ARM-based embedded hardware platform with a three-dimensional mesh structure was first established to support the decentralized algorithms. To deploy deep convolutional neural networks (CNNs) in this embedded parallel environment, a distributed mapping mechanism was proposed to efficiently decentralize computation tasks in the form of a multi-branch assembly line. In addition, a dimensionality reduction initialization method was also utilized to successfully resolve the conflict between the storage requirement of computation tasks and the limited physical memories. LeNet-5 networks with different sizes were optimized and implemented in the embedded platform to verify the performance of our proposed strategies. The results showed that memory usage can be controlled within the usable range through dimensionality reduction. The down-sampling layer as the base point of the mapping for the inter-layer segmentation achieved the optimal operation in lateral dispersion with a reduction of around 10% in the running time compared with the other layers. Further, the computing speed for a network with an input size of 105 × 105 in the multi-core parallel environment is nearly 20 times faster than that in a single-core system. This paper provided a feasible strategy for edge deployments of artificial intelligent algorithms on multi-core embedded devices. [ABSTRACT FROM AUTHOR]

The business world is becoming more competitive. Therefore, it is crucial to increase the flexibility of production by decreasing the time used in the processes of preparing the production lines for new items' production, reducing changeover and setup times. This paper presents a case study where the main goal is to reduce the setup time of welding robots. Single Minute Exchange of Die (SMED) was implemented, using other tools such as the Spaghetti Diagram, ERCS Analysis (Eliminate, Rearrange, Combine, Simplify), Gemba Walk, Standardized Work, Flowcharts, and Pareto Diagram. The setup time decreased by 36% in the welding robots studied, decreasing the motions by 43% during the changeover process and reducing the time from the categories: "transportation", "main", "other", and "waiting". In addition to SMED implementation, this study offers an integrated study of several Lean tools and Quality tools to achieve the maximum reduction of changeover and setup times. [ABSTRACT FROM AUTHOR]

Bodywork assembly in bus manufacturing is a task that involves several operations. In Mexico, many bus manufacturing companies have non-flexible production lines, which makes model change very problematic. In response to this situation, the SMED methodology was applied to make more flexible production lines in terms of setup reduction for model changing. In this paper, the urban bus assembly line was studied, which is composed of sub-assembly stations that subsequently feed the station that assembles the unit, and that joints all the frames of the required model, the aim of the paper is to propose an approach to reduce the production time through the use of folding references. With the application of the SMED, it was observed that the production line improved in the transition from one body structure model to another, obtaining results of 56.2% in the time reduction. [ABSTRACT FROM AUTHOR]

This paper presents a solid methodology for improving the efficiency and productivity of assembly lines using Lean Manufacturing tools, in particular the Define, Measure, Analyze, Improve, and Control approach (DMAIC) and line balancing techniques, followed by a concrete application in a case study of a wiring industry assembly line. The first phase of the approach ensured a clear definition of the problem using the who, what, where, when, why, and how tool (5W1H) and a description of the manufacturing process. The measurement phase allowed the calculation of the Takt time (TT) and the timing of the cycle times of the 17 stations of the line with the use of data collected on the standardized work combination table (SWCT) documents. This facilitated the analysis phase by first establishing a Yamazumi chart showing the distribution of the load between the line's stations and allowing the identification of bottleneck stations, and then analyzing the situation through the 5-Why tools and the Ishikawa diagram. Thanks to the innovation phase and the ideal balancing conditions developed in this paper, it was possible to balance the line's stations using an action plan whose effectiveness was monitored during the control phase, improving efficiency from 78% to 95% with a saving in manpower by reducing the number of operators from 17 to 14. [ABSTRACT FROM AUTHOR]

In manual order picking systems, a paper list specifying the name, location, and amount of each item is often given, and this would exert mental pressure on pickers while finding target shelf bins in high-density picking environments. Augmented reality (AR) can provide a friendly alternative to improve the manual order picking performance by conveying picking information into visual instruction. Nevertheless, there is still no systematic consensus to deploy the pick-by-AR associated with actual warehouse workplaces. To establish a spatial correspondence between the visual guidance with the actual workplace, the multi-marker-based global map about the warehouse floor is established in advance. Instead of traditional single marker-based pick-by-AR methods, the warehouse floor-related map can provide an accurate and continuous navigation performance for intuitive AR guidance, allowing the picker to move freely on the warehouse floor without limiting to some certain locations. Besides, a systematic pick-by-AR solution is available by integrating the proposed method to a lightweight wearable AR device, and this easy-to-deploy pick-by-AR solution can alleviate the worker's mental effort while doing picking action. Finally, the pick-by-AR system described in the paper is deployed in the automobile assembly line, and experimental results illustrate that the proposed method can improve the picking efficiency while reducing the picking errors compared with the current paper-based order picking. [ABSTRACT FROM AUTHOR]

Two-sided assembly lines balancing problem has been extensively studied in recent years. However, only limited attention has been paid to balancing mixed-model two-sided assembly lines problem (MTALBP). Moreover, the majority of balancing research assumes the setup as negligible, although it is ubiquitous in the assembly process. As the non-increment activities, setups occur in two ways: forward and backward setups. According to our best knowledge, no published work in literature on MTALBP has simultaneously considered forward and backward setups. In this paper, the problem of balancing mixed-model two-sided assembly lines with setups (MTALBPS) is considered. The purpose of this paper is twofold. The primary objective is to develop a mixed-integer programming (MIP) mathematical model to formulate the type-I problem of MTALBPS. The secondary objective is to propose an effective variable neighbourhood search (VNS) algorithm to solve it, especially for the large-sized problems. In addition, to test the effectiveness of the proposed approaches, a number of test problems from the literature with up to 148 tasks are solved and compared with the lower bound. The results demonstrate that the proposed algorithm is effective and produces very close results to the lower bound in a reasonable time. [ABSTRACT FROM AUTHOR]