No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Effects of stochastic and heterogeneous worker learning on the performance of a two-workstation production system
Abstract
Production systems in industries are undergoing transformative changes, with the rise of Industry 4.0 technologies amplifying the complexity of manual and semi-automated workstations, necessitating advanced training and adaptability from human workers. Human workers, due to their unique blend of cognitive and motor skills, thus flexibility, are indispensable and will continue to play a pivotal role. Because of their unique experiences and attributes, they inherently exhibit variability in their processing times and learning rates, which complicates frequent production ramp-ups. Recognizing the lack of comprehensive models that simultaneously account for stochastic processing times and heterogeneous learning during production ramp-ups, this study aims to bridge this gap. We developed an analytical model of a two-worker production system with an intermediate buffer by focusing on worker learning curves, stochastic processing times, and learning heterogeneity. Through an illustrative case, we derived insights into the performance of such systems, specifically in terms of measures including the mean throughput time of a batch, mean waiting time of a part in the buffer, mean idle time of workers, work-in-progress distribution, and buffer usage during the production run. We found that deterministic learning models can significantly underestimate the throughput times, and even consistent average learning rates can lead to variable throughput times based on the learning patterns. Our findings emphasize the need for production managers to consider these factors for realistic and effective production planning, underscoring the novelty of our approach in addressing these intricate dynamics to improve not only system performance, but also worker well-being.
... In recent years, cellular production has continuously penetrated and influenced various manufacturing modes (e.g., intelligent manufacturing, reconfigurable manufacturing, mass customization), showing an increasingly obvious trend of flexibility, collaboration and intelligence. To cope with the fluctuating production environment, the production cell also emphasizes the active role of humans, regarding high-quality workers and flexible cooperative production teams as important factors to improve the flexibility of the system (Bouaziz et al. 2022;Ranasinghe et al. 2024). Therefore, research on the cooperative behavior of production cell workers under a fluctuating environment is helpful to realize the best match between humans, machines and the environment to improve the manufacturing flexibility from the perspective of the organization. ...
... The manufacturing process of the production cell contains many discrete events, and the workers are active, autonomous and adaptable individuals; therefore, simulation technology based on complex system theory is considered to be a more effective way to study such complex production systems (Barbosa et al. 2015;Wang et al. 2016;Zhao et al. 2017). Current simulation studies on worker behavior show that organizational collaboration can significantly improve the performance of manufacturing systems (Liu et al. 2023;Mendez-Vazquez and Nembhard 2019;Ranasinghe et al. 2024;Zhang et al. 2015). However, in the actual production process, excessive collaboration also affects the system's smoothness, increasing labor training costs and communication costs, and reducing the overall system performance. ...
... Bouaziz et al. (2022) simulate the randomness of different workers' behaviors through Markov chains, testing the effects of different random behaviors on the performance of production systems. Ranasinghe et al. (2024) focus on workers' learning behavior, random behavior and learning heterogeneity, establishing a simulation model of a two-workstation production system with intermediate buffers to evaluate the effects of different learning models on the production system. Bokhorst et al. (2006) propose load-oriented collaboration and resource-oriented collaboration modes, analyzing the impact of these modes on system performance using the discrete event modeling method. ...
Flexible cooperative production workers are important for the performance and flexibility of the production cell. In this paper, a human–machine integrated hybrid modeling method is first proposed, which combines the discrete process model with the multi-agent model to simulate the collaborative behavior of workers in a discrete production process. Model structure, simulation process, and worker collaboration behaviors are illustrated in detail. An embedded framework of digital twins is established based on the simulation. Simulation experiments are then carried out on two changeover-based collaboration schemes in a variable-batch mixed-flow production cell under different production environment conditions, to obtain the dynamic and adaptive collaboration scheme. Both the performance indexes and the flexibility indexes of the production cell are evaluated, to achieve a quantitative evaluation of flexibility. The results show that the impact of workers’ behaviors on production performance differs for the various collaboration schemes and dynamic production environments. Autonomous collaboration scheme shows better performance only under high environmental fluctuation level, but has a higher flexibility considering all fluctuation levels. This study can help managers realize the best match between humans, machines and the environment by improving workers’ collaboration behaviors.
Industries are currently struggling with ageing workforce in modern production systems associated with Industry 4.0. The industrial socio-demographic problem is more and more present as the increasing of the ageing population results in the prolongation of the working life and the consecutively in the ageing of the workforce in industries. This paper aims to conduct a systematic literature review on the challenges and concerns of ageing operators, including the physical, cognitive, ergonomic, and well-being conditions of the ageing workforce in the Industry 4.0 environment. The ScienceDirect, Scopus, Web of Science and PubMed scientific databases were used to survey the studies and selected using PRISMA guidelines. This paper was structured and analysed by clusters: Ageing, Industry 4.0, Human Factors, and Ergonomics. These clusters were developed as research lines: Ageing as the socio-demographic challenge, Industry 4.0 as the technological development, Human Factors as the individual characteristics of the operator, and Ergonomics as the work environment. Thus, human-centric approaches and ideas are discussed with the insights and issues of Industry 4.0 technologies, Human Factors, and Ergonomics to achieve a sustainable system at the engineering and social level.
In manual assembly systems, workers' differences in terms of skills, level of expertise and perceived physical effort largely affect the assembly line balancing and system performance. Traditional long-term strategic decisions may not respond to workforce changes and needs, resulting in frequent requests for line rebalancing. In this study, we propose a methodological framework and an easy-to-use Assembly Line Worker Assignment and Rebalancing Problem with different options: workers' assignment considering their performance variability, integration of worker dependent physical exertion constraints and possibility to use trainers to assist inexperienced workers. A bi-objective linear programming model is proposed aiming to minimise the cycle time and the number of reassigned tasks to respect the initial design while integrating new workers with different characteristics. The ε-constraint approach is used to build Pareto frontiers for this bi-objective problem. This approach is applied to three real cases. The obtained results show that the developed model can be successfully used in manufacturing companies to help the production managers to deal with workforce turnover and skills heterogeneity.
With the application of Advanced Manufacturing Technologies (AMTs), food manufacturers can operate outside of traditional volume-variety diagonal positions without the predicted performance penalties. Although food manufacturers produce high-variety and high-volume products with low cost and high quality, it is unavoidable that their high-variety production processes experience a high changeover frequency. Changeovers are not only time-consuming but are also accompanied by production losses. While the literature has emphasized the technical and technological aspects of changeovers, e.g. the Single Minute Exchange of Dies (SMED), most of these studies focussed on flexibility, time, and product quality. As food manufacturers’ production is capital-intensive with low-profit margins, it is essential for them to improve changeover efficiency and reduce changeover losses. This paper thus aims to understand changeover activities, identify the key factors influencing food losses during changeovers and develop a mechanism for improving overall production efficiency. To that end, an in-depth case study is conducted with a major European food manufacturer. This study identified nine key factors influencing changeover losses and developed an improvement framework with three key themes (working methods, knowledge management, and organizational design).
Industry 5.0 is a combination of organisational principles and technologies to design and manage operations and supply chains as resilient, sustainable, and human-centric systems. While the general notion of Industry 5.0 has been elaborated, its implications for future operations and supply chains remain underexplored. This paper contributes to the conceptualisation of Industry 5.0 from the perspective of viability. We contextualise a framework of Industry 5.0 through the lens of the viable supply chain model, the reconfigurable supply chain, and human-centric ecosystems. Our study uncovers the major dimensions that characterise Industry 5.0 as a technological-organisational framework. First, the major technological principles of Industry 5.0 are collaboration, coordination, communication, automation, data analytics processing, and identification. Second, Industry 5.0 covers four areas: organisation, management, technology, and performance assessment. Third, Industry 5.0 spans three levels: society level, network level, and plant level. Last but not least, Industry 5.0 frames a new triple bottom line: resilient value creation, human well-being, and sustainable society. We provide a definition of Industry 5.0 and discuss its implications by elaborating on the understanding of value in Industry 5.0, which spans the dimensions of profit, people, and society. We also discuss open research areas.
Many companies, despite there being opportunities for automation in production and logistics (P&L) systems, still rely on human workers due to their cognitive and motor skills. Taking Human Factor (HF) aspects into consideration when making P&L system design and management decisions is therefore important, an ignorance of HF potentially resulting in operator fatigue, discomfort, subsequent injuries and negative consequences for operator performance and the P&L system. A review of the literature shows that the majority of studies that take HF into consideration focus either on designing the workplace or on operation planning activities. There is also still a gap in the literature. Little has been published on P&L systems that incorporate HF and that combine different levels of short-term operational policy decisions (e.g. job allocation) and long-term system characteristic decisions (e.g. layout design). Current state-of-the-art frameworks that support the design and management of P&L systems and that take HF into consideration rarely consider different decision levels. This study proposes a new framework that incorporates HF into P&L systems by combining different levels of decisions to improve performance, quality, and well-being.
The new Industry 5.0 paradigm complements the well-known Industry 4.0 approach by specifically driving research and innovation to facilitate the transition to sustainable, human-centric and resilient industry. In the manufacturing context, workers' diversity in terms of experience, productivity and physical capacity represents a significant challenge for companies, especially those characterized by high staff turnover and manual processes with high workload and poor ergonomics. In seeking to address such challenges, this research adopts a human-centric perspective to define new flexible job arrangements by developing a new multi-objective job rotation scheduling model. The proposed model is unique in that it aims to achieve multiple job assignment objectives by simultaneously considering different socio-technical factors: workers' experience, physical capacity and limitations, postural ergonomic risks, noise and vibration exposure, and workers' boredom. The model's implementation in real environments can be supported by new sensor-based technologies that collect data on workers' efficiency, ergonomic scores and task performance and enable workers to participate in measuring perceived fatigue and boredom. The primary goal of our model is to find the most appropriate assignment of job and individual-flexible rest-break plan for each worker. The authors test the model application in an industrial setting. Useful managerial insights emerge and prescriptive recommendations are provided.
Sustainable production includes economic, environmental, and social aspects. However, social sustainability has received less attention, especially compared to the economic aspects. Next to technical and organizational measures, social improvements within supply chains can also be achieved through suitable production planning. Within production planning, production programs are determined, and the assignment of available resources (e.g., employees) is specified. Thus, the utilization and workload of employees are defined. This systematic literature review investigates to what extent such employee-related social aspects are reflected in production planning and discusses whether economic aspects dominate them. For this, a Scopus database search was carried out and 76 identified approaches were analyzed and categorized regarding the occurring employee-related social aspects and their implementation. Thus far, the approaches mainly consider single aspects on single planning levels. A consideration of a broad set of aspects along the entire production planning has rarely been studied. In particular, health and safety aspects are considered on the levels of assembly line balancing and job rotation. However, their impact is primarily determined by the specific settings of the decision-maker. To support decision-makers, only a few studies have investigated the effects based on real application scenarios. Further potential might be an extended modeling of social and economic interdependencies and a consideration of employee-related social aspects in medium- to long-term production planning.
The recent shift to wellbeing, sustainability, and resilience under Industry 5.0 has prompted formal discussions that manufacturing should be human-centric-placing the wellbeing of industry workers at the center of manufacturing processes, instead of system-centric-only driven by efficiency and quality improvement and cost reduction. However, there is a lack of shared understanding of the essence of human-centric manufacturing, though significant research efforts exist in enhancing the physical and cognitive wellbeing of operators. Therefore, this position paper presents our arguments on the concept, needs, reference model, enabling technologies and system frameworks of human-centric manufacturing, providing a relatable vision and research agenda for future work in human-centric manufacturing systems. We believe human-centric manufacturing should ultimately address human needs defined in an Industrial Human Needs Pyramid-from basic needs of safety and health to the highest level of esteem and self-actualization. In parallel, human-machine relationships will change following a 5C evolution map-from current Coexistence, Cooperation and Collaboration to future Compassion and Coevolution. As such, human-centric manufacturing systems need to have bi-directional empathy, proactive communication and collaborative intelligence for establishing trustworthy human-machine coevolution relationships, thereby leading to high-performance human-machine teams. It is suggested that future research focus should be on developing transparent, trustworthy and quantifiable technologies that provide a rewarding working environment driven by real-world needs.
This study is based on calculation of standard minute value (SMV) of a T-shirt by determining the manpower and machine allocation for particular target. An experimental investigation for the distribution of SMV for each and every operation requires for making a T-shirt and provides a clear and detail concepts for determining line balancing, machine requirements, man power allocation, operational sequence of the manufacturing process for setting a definite target within a reasonable efficiency. This study is a detailed discussion and distribution of SMV of a T-shirt by work study which will assist to minimize the SMV for having a better synchronization with man, machine, materials and methods to achieve higher efficiency. A general overview over efficient productive method to make a T-shirt is given in this paper.
The effects of workforce differences on manufacturing systems have attracted the attention of a wide range of researchers in recent years. The differences between workers in terms of skills, age, gender and anthropometric measures have a large impact on production system performance. In this study, the workforce differences factors in production system design and modelling were investigated, with the aim of understanding how the differences between workers could influence a production system and how they had been considered in previous studies. The papers selected from the Scopus database were categorised based on whether how human factors are incorporated into manufacturing system optimisation and design approaches is discussed therein or is not. To find relevant papers, two sets of keywords were defined: (1) keywords relating to the differences between workers and (2) keywords relating to the kind of problem under study. Furthermore, the investigated papers helped highlight the strengths and weaknesses of the existing literature and derive a discussion on the possible future research steps. ARTICLE HISTORY
The fourth industrial revolution we currently witness changes the role of humans in operations systems. Although automation and assistance technologies are becoming more prevalent in production and logistics, there is consensus that humans will remain an essential part of operations systems. Nevertheless, human factors are still underrepresented in this research stream resulting in an important research and application gap. This article first exposes this gap by presenting the results of a focused content analysis of earlier research on Industry 4.0. To contribute to closing this gap, it then develops a conceptual framework that integrates several key concepts from the human factors engineering discipline that are important in the context of Industry 4.0 and that should thus be considered in future research in this area. The framework can be used in research and development to systematically consider human factors in Industry 4.0 designs and implementations. This enables the analysis of changing demands for humans in Industry 4.0 environments and contributes towards a successful digital transformation that avoid the pitfalls of innovation performed without attention to human factors. The paper concludes with highlighting future research directions on human factors in Industry 4.0 as well as managerial implications for successful applications in practice.
The 4th industrial revolution (Industry 4.0, I4.0) is based upon the penetration of many new technologies to the industrial world. These technologies are posed to fundamentally change assembly lines around the world. Assembly systems transformed by I4.0 technology integration are referred to here as Assembly 4.0 (A4.0). While most I4.0 new technologies are known, and their integration into shop floors is ongoing or imminent, there is a gap between this knowledge and understanding the form and the impact of their full implementation in assembly systems. The path from the new technological abilities to improved productivity and profitability has not been well understood and has some missing parts. This paper strives to close a significant part of this gap by creating a road map to understand and explore the impact of typical I4.0 new technologies on A4.0 systems. In particular, the paper explores three impact levels: strategic, tactical, and operational. On the strategic level, we explore aspects related to the design of the product, process, and the assembly system. Additionally, the paper elaborates on likely changes in assembly design aspects, due to the flexibility and capabilities that these new technologies will bring. Strategic design also deals with planning and realizing the potential of interactions between sub-assembly lines, kitting lines, and the main assembly lines. On the tactical level, we explore the impact of policies and methodologies in planning assembly lines. Finally, on the operational level, we explore how these new capabilities may affect part routing and scheduling including cases of disruptions and machine failures. We qualitatively assess the impact on performance in terms of overall flow time and ability to handle a wide variety of end products. We point out the cases where clear performance improvement is expected due to the integration of the new technologies. We conclude by identifying research opportunities and challenges for advanced assembly systems.
The occurrence and characteristics of human learning and forgetting are extensively researched in many fields. Thus, the literature on learning curves is abundant. Within production and operations management, learning curves can describe the performance improvement of workers due to repetitions or experience, which makes them a useful tool for managerial decision making. Human learning is very relevant to manufacturing firms that are labor-intensive, especially where labor is costly. Also, technology and organization learning frequently occur in many organizations, even when manufacturing is not labor-intensive. It is not surprising, therefore, that, despite the plethora of prior work in learning curves, research in this important area continues to appear in the production and operations management literature. A comprehensive, systematic review of the literature, however, has never been conducted.
This article aims at giving a general overview of learning curves in production and operations management. It maps this specific research area, synthesizes its research findings, highlights its key application areas and explores future research directions. This work achieves this goal by presenting the results of a systematic literature review on the applications of learning curves in production and operations management. First, a framework that includes typical learning curve models is developed. The fundamental characteristics of learning curves and their applications in production and operations management are then identified. This framework is used to categorize the literature. This study also presents a discussion of the most important and informative articles in each of the major categories, as well as highlighting future research opportunities.
ract
This study is based on calculation of standard minute value (SMV) of
a T-shirt by determining the manpower and machine allocation for particular target. An experimental investigation for the distribution of SMV for each and every operation requires for making a T-shirt and provides a clear and detail concepts for determining line balancing, machine requirements, manpower allocation, operational sequence of the manufacturing process for setting a definite target within a reasonable efficiency. This study is a detailed discussion and distribution of SMV of a T-shirt by work study which will assist to minimize the SMV for having a better synchronization with man, machine, materials and methods to achieve higher efficiency. A general overview over efficient productive method to make a T-shirt is given in this paper.
Apparel manufacturing is a labour intensive industry which is characterized by rapid product changes. Each time such changeover of product type occurs, it demands operator learning for the process to reach a steady-state. Higher the number of product type changes, higher the adverse impact of learning on the production performance. Purpose of this paper is to analyse and model this learning process in batch assembly (sewing) lines to improve the accuracy of forecasting, production planning and inventory control. For this purpose, empirical data of daily production efficiency is collected from a large scale apparel manufacturer in Sri Lanka over a period of 8 months. Product types or styles are categorized as " Repeat Styles " or " New Styles " based on whether the operators have prior experience in assembling the product type or not. Modelling of learning curves for above two categories of styles was carried out separately, using nonlinear regression analysis. This analysis indicates that the hyperbolic learning curve models are the best fitted learning curve for both style categories with different parameter estimates. Models were validated for both style categories using a separate set of daily efficiency data from same sewing lines.
Production ramp-up is a critical step in the life cycle of a new product, and efficiently managing ramp-ups is a key to business success and market leadership. To support the planning of ramp-ups in practice, researchers have developed decision support models in the past that help to solve problems that arise during the ramp-up phase, such as lot sizing, the assignment of workers to workplaces or the determination of the capacity of the production equipment. Decision support models for production ramp-up typically consider the specific characteristics of this phase, such as uncertainty, growth in demand, worker learning or imperfect production processes. The aim of this paper is to provide a comprehensive overview of decision support models for production ramp-up and to identify areas where more research is needed. First, the paper develops a conceptual framework of production ramp-up by categorising typical planning problems and process characteristics of the ramp-up phase. Secondly, a systematic literature review with a focus on mathematical planning models for the ramp-up phase is conducted. The analysis shows that various decision support models that help to realise an efficient production ramp-up exist, but that there are still many opportunities for future research in this area.
Assembly lines are useful for mass production of standard as well as customized products. Line balancing is an important issue, in this regard an optimal or near optimal balance can provide a fruitful savings in the initial cost and also in the running cost of such production systems. A survey of different problems in different types of assembly lines and some of the critical and on going research areas are highlighted here. The provided research information is momentous for the research community in assembly line area to proceed further in the presented issues of assembly lines. © 2014 Higher Education Press and Springer-Verlag Berlin Heidelberg.
A learning curve is a well known tool which describes the relation between the performance of a task and the number of repetitions of that task. A learning curve model may be used as a prediction tool in various applications of operations planning and control. Previous models have treated learning as a deterministic phenomenon, whereas this paper addresses its stochastic nature and treats learning as a random process. Exact expressions of the variance, of the coefficient of variation and of the probability density function of the performance, expressed as a function of the number of repetitions of a task, are derived for the power model. Mathematicalanalysis reveals that the coefficient of variation is independent of the number of repetitions. The probability density function type was also found to be independent of the number of repetitions. Analysis of actual learning data validated the theoretical findings. These findings enable improved forecasting of future performance by using a stochastic rather than a deterministic estimation.
Purpose
The purpose of this paper is to study the prevalence of human learning in the order picking process in an experimental study. Further, it aims to compare alternative learning curves from the literature and to assess which learning curves are most suitable to describe learning in order picking.
Design/methodology/approach
An experimental study was conducted at a manufacturer of household products. Empirical data was collected in the order picking process, and six learning curves were fitted to the data in a regression analysis.
Findings
It is shown that learning occurs in order picking, and that the learning curves of Wright, De Jong and Dar‐El et al. and the three‐parameter hyperbolic model are suitable to approximate the learning effect. The Stanford B model and the time constant model led to unrealistic results.
Practical implications
The results imply that human learning should be considered in planning the order picking process, for example in designing the layout of the warehouse or in setting up work schedules.
Originality/value
The paper is the first to study learning effects in order picking systems, and one of the few papers that use empirical data from an industrial application to study learning effects.
Learning curves (LCs) are deemed effective tools for monitoring the performance of workers exposed to a new task. LCs provide a mathematical representation of the learning process that takes place as task repetition occurs. These curves were originally proposed by Wright in 1936 upon observing cost reduction due to repetitive procedures in production plants. Since then, LCs have been used to estimate the time required to complete production runs and the reduction in production costs as learning takes place, as well as to assign workers to tasks based on their performance profile. Further, effects of task interruption on workers’ performance have also being modeled by modifications on the LCs. This wide variety of applications justifies the relevance of LCs in industrial applications. This paper presents the state of the art in the literature on learning and forgetting curves, describing the existing models, their limitations, and reported applications. Directions for future research on the subject are eventually proposed.
Production ramp-up is an important phase in the lifecycle of a manufacturing system which still has significant potential for improvement and thereby reducing the time-to-market of new and updated products. Production systems today are mostly one-of-a-kind complex, engineered-to-order systems. Their ramp-up is a complex order of physical and logical adjustments which are characterised by try and error decision making resulting in frequent reiterations and unnecessary repetitions. Studies have shown that clear goal setting and feedback can significantly improve the effectiveness of decision-making in predominantly human decision processes such as ramp-up. However, few measurement-driven decision aides have been reported which focus on ramp-up improvement and no systematic approach for ramp-up time reduction has yet been defined. In this paper, a framework for measuring the performance during ramp-up is proposed in order to support decision making by providing clear metrics based on the measurable and observable status of the technical system. This work proposes a systematic framework for data preparation, ramp-up formalisation, and performance measurement. A model for defining the ramp-up state of a system has been developed in order to formalise and capture its condition. Functionality, quality and performance based metrics have been identified to formalise a clear ramp-up index as a measurement to guide and support the human decision making. For the validation of the proposed framework, two ramp-up processes of an assembly station were emulated and their comparison was used to evaluate this work.
Industry 4.0 was presented more than a decade ago as the fourth industrial revolution, aiming to significantly raise the level of sophistication of interconnected technologies and thus increase manufacturing industries’ profits. However, because the technology-driven narrow focus of Industry 4.0 on performance and profit fails to explain how to increase prosperity for all the stakeholders involved, the European Commission has introduced the concept of Industry 5.0. This vision overcomes the weaknesses of Industry 4.0 by paying explicit attention to outcomes for humans in the system and establishing an environment to create human-centric, resilient, and sustainable systems. Considering these developments, this position paper and editorial introducing the special issue of the International Journal of Production Research elaborates on the transition from Industry 4.0 to 5.0 through 10 papers focusing on the human-centric pillar of Industry 5.0 and its impacts on production and logistics system design and management. This work presents guidance for a more systemic approach needed in future research: to include empirically grounded works and data-driven multimethod approaches that consider diversity in system operators and human factors demands holistically in order to incorporate ethical implications missing from Industry 4.0 – in the pursuit of Industry 5.0 systems.
Prior studies have investigated the relationship between Industry 4.0 technologies and work. This paper acknowledges the contributions of such studies and builds on their perspective to broaden the understanding of the contribution of Industry 4.0 technologies to specific worker capabilities and manufacturing activities. The aim is to build a conceptual framework to consolidate a common view on this growing yet fragmented issue by integrating a wide range of findings from the literature. The study adopts a systematic literature review approach to systematize such knowledge in a singular and consolidated perspective on Industry 4.0 technologies and work. The study analyzes 80 papers in this field and investigates how different Industry 4.0 technologies are related to workers’ manufacturing activities. Eight main manufacturing activities were considered to frame the analysis: assembly, maintenance, training, quality control, movement, machine operation, product and process design, and production planning and control. Eight worker capabilities that Industry 4.0 technologies can enhance were also considered: super-strength capability, augmented capability, virtual capability, healthy capability, smart capability, collaborative, social capability, analytical capability. Based on these 80 papers, this paper conceptualizes Smart Working-related technologies for Operators 4.0 and shows their benefits and limitations as described in the literature. The study shows how these manufacturing activities and worker capabilities can be supported by Industry 4.0 technologies, which is useful for future research and the design of operational processes in the Industry 4.0 context.
To generate continuing profits as well as to retain market share, companies need to regularly introduce new products into the market, replacing outdated products and enticing buyers to purchase new products. Thus, planning the product transition from the phase-out of a product to the phase-in of a product is a challenging task for both production planners and marketing managers. One significant concern in particular is how to account for the learning effect associated with the production of two successive products that share a single type of production resource. To address this problem of designing production plans for two successive products, we develop an MINLP (Mixed-Integer Non-Linear Programming) model that minimizes the total operating cost for managing the product transition. In doing so, we consider the number of production resources, the lengths of the product production periods, and the number of setups for both products during the ramp-down and ramp-up periods. We study the behaviors of the proposed model with the help of numerical experiments by varying the values of key parameters, including the learning effect and its impact on performance under different operational scenarios. Leveraging the results from the sensitivity analyses, the operational characteristics for managing the product transition are discussed.
Currently, the ageing population is getting higher attention in production systems due to the increasing percentage of the ageing workforce, which persists longer in the working environment. Older workers are more vulnerable because of a possible decrease in physical capacities but can have a higher experience level. It is important to evaluate these factors during production process decisions and adapt the working environment to such workers. In such a way, production systems performance and workers’ well-being are jointly achieved. In companies aiming to include these aspects, digitalisation and automation can provide tools that can be used at tactical, strategic or operational levels. For example, at the operational level, wearable sensors, smart devices or trackers can be used in analysing the working time, effort and physical fatigue of operators via human-energy expenditure. These measures can be used to assess the required recovery time. However, the most challenging step is to include these measures in the decision-making process to organise the working process for operators according to their profiles and health restrictions. In this paper, we show how fatigue can be incorporated into operational processes such as scheduling in the context of a Dual-Resource Constrained (DRC) job shop. We consider several models to integrate rest allowance into the scheduling process and present a heuristic approach to assign operations to workers and machines. The impact of rest allowance on system performance is measured and managerial insights are provided.
In the time of industry 4.0 and big data, methods which are based on the collection and the processing of a large amount of data in order to support managerial decisions have outstanding significance. The learning curve theory pertains to these methods. The purpose of this paper is to explore some application possibilities of the classical learning curve in manufacturing and service operations. The learning effect assumes that as the quantity of units manufactured increases, the time needed to produce an individual unit decreases. The function describing this phenomenon is the learning curve. Various learning curves have been developed and applied in the area of production economics and much research studies the significance of the learning effect in management decisions. This study summarizes the main learning curve models and demonstrates how learning can be considered in three classical areas of operations management. First, the calculation of economic manufacturing quantity in the presence of learning is studied. Next, the effect of learning in break-even analysis and assembly line balancing is explored. The results show that with the consideration of the learning effect, calculations become more complex and require greater efforts, but the application of the learning curve concept can provide valuable insight both at operational and strategic levels.
Production ramp-up is a key phase during the introduction or changeover of a production system. Process calibration and tuning are inevitably required to make such a system fully operational and let it reach its maximum production yield. A complex decision-making process takes place in order to optimally tune the system and requires a long time for testing and experimenting that will determine the system behaviour. This work considers the sequential nature of ramp-up and proposes a Cyber-Physical Systems approach based on data capturing, learning mechanisms and knowledge extraction, leading to an Industry 4.0 compliant Decision Support System (DSS) for human operators. The proposed system is implemented as an online DSS and also supports offline learning using previously gathered knowledge. A number of experiments have been carried out on a micro scale assembly station, validating the expected benefits of the proposed DSS. Results show a reduction of over 40% in the number of ramp-up steps required when using the DSS.
Many new model launches in Europe do not meet the quality, cost and production targets set by automotive companies during production ramp-up. Although increasing car complexity could be a key factor in explaining this issue, it is not sufficient to explain why production sites are not able to meet these important targets. This article analyses whether the production system used in European car manufacturing plants are suitable for achieving performance goals during the production ramp-up phase. First, the differences between serial production and the production ramp-up phase are analyzed through the conceptual framework of the European Production System in the automotive industry. Second, a case study using real data from a European automotive plant is presented, and the evidence calls into question the appropriateness of maintaining the same production system during the ramp-up phase. The study concludes that there is a need to make the production system in Europe more flexible during the ramp-up phase. In particular, changes should be more oriented towards fostering organizational improvement capability in order to reduce the stops that occur in the production line, avoid using pull systems as a constraint or limitation, and help to build a new culture of stopping to fix problems.
In this paper, we consider a product manufacturer that seeks to leverage the potential of human learning
to develop the capacity of its workforce and to reduce its costs. Unlike much of the literature in this area,
we do not assume that the rate at which individuals learn is known with certainty. We present a two-stage
stochastic programming model of the related production planning problem that quantifies the impact of worker
assignment decisions to produce through an exponential learning curve which we linearize to yield a mixed
integer linear program that can be solved efficiently. With this stochastic program, we perform a rigorously
designed computational study and statistical analysis to derive tactics and managerial insights for how an
organization should plan its production operations about assignment, cross-training and practicing. Results
suggest that explicitly recognizing uncertainty in learning rates would reduce costs and that when dealing
with assignment decisions, the leading factor to consider is the mean learning rate. On the other hand, when
dealing with cross-training and practicing decisions, the learning rate variance is more predictive. We also
assess the impact of explicitly considering stochastic forgetting rates in the productivity curve, finding that in
the optimal assignment schedule, workers practice more and always specialize.
Mass Customization (MC) implies in large product variety and reduced production batch sizes. Tasks that rely on human ability are especially affected in this context as workers must quickly adapt to the requirements of new models. That adaptation process takes place differently in each worker, justifying the development of strategies to group workers with similar learning profiles. This paper proposes a framework to form homogeneous groups of workers based on their learning profiles by integrating learning curve (LC) modeling and cluster analysis. For that, performance data are collected and modeled through LCs, such that LC parameters quantify workers’ adaptation to tasks. Principal Components Analysis (PCA) is applied to the dataset consisting of the LC parameters; PCA outputs give rise to an importance index that guides a backward parameter selection process. After each LC parameter is removed from the dataset, a new grouping using the Fuzzy C-Means (FCM) clustering technique is carried out using the remaining LC parameters, and the quality of the formed groups is assessed by means of three metrics. When applied to a shoe manufacturing process, 8 out of the original 29 LC parameters were retained, and used to insert workers into two clusters. The reduced subset of parameters elevated the quality of the clustering procedure by 29.4% (from 0.476 to 0.616) according to the Silhouette Index; similar improvements were indicated by the other two metrics. The retained parameters are related to workers’ performance at the first repetition and previous experience, highlighting the importance of well-designed training programs.
The average time in the system (or average throughput time) is recognized as a crucial performance measure of manufacturing systems and, in recent years, has become even more relevant as markets have shifted towards mass-customization scenarios, since it directly affects the average inventory in the system. Modern manufacturing systems design paradigms, Industry 4.0 on top, focus on the need to achieve high responsiveness by shortening throughput time as the main lever to maintain cost efficiency and competitiveness in such market scenarios. In the contexts in which the human impact on production operations is not negligible, the learning phenomenon may reasonably affect the average time in the system. This paper analyses the learning effect on this performance measure through batching policies. Here, batching is considered as the possibility to group customized orders presenting similarities in the production processes, thus making it possible the exploitation of learning effect. Both single- and multi-item systems under Markovian arrivals are studied, also deterministic and stochastic processing times are considered. The problem that we pose consists in determining the batch size for each product that minimizes the average time in the system taking into consideration the learning effect, which is included by means of the Plateau model. A solution procedure for each case is discussed and, through numerical experiments, the sensitivity of the model to variations in parameter values is studied.
Many problems in manufacturing systems can be formulated via Markov stochastic modeling. This paper gives a review and classification of timed models of manufacturing systems with particular emphasis on Markov models. As the associated Markov chains of even small systems are characterized by the well-known state explosion or largeness problem, the review continues on with the models and methods for the numerical solutions of large Markov chains. In addition, the software tools are summarized. Finally, the paper provides some challenges and directions for further research on the modeling of manufacturing systems.
The learning curve is a powerful tool for estimating either the production cost or time in production analysis. Our work focuses on the characterization of the uncertain nature of the learning effect. Traditionally, the learning curve of a new product can be derived by processing the past data obtained for similar products. However, due to the complicated environment, in some cases, the data for similar products may be lacking, preventing the use of the traditional methods to find the learning curve. To address this issue, we develop an uncertain learning curve by utilizing uncertainty theory. Moreover, some useful theorems are developed to characterize the uncertain learning curve. An experiment is conducted to compare the proposed model and the standard model. Finally, we apply the uncertain learning curve to the single-machine optimization problem.
Decision support models for production and logistics have neglected human factors to a large extent in the past. For this reason, researchers have called for the development of more realistic planning models for production and logistics problems incorporating human factors. This call has recently been addressed by an increasing number of innovative approaches, which makes the study of human factors in decision support models for production and logistics a more and more maturing research field.
The aim of this paper is to present the evolution of works that incorporate human factors into decision support models for production and logistics. For this reason, a systematic literature review is performed. The results of the literature review reveal the current state of research and highlight areas and approaches where additional research is promising.
Assembly lines with non-constant task time attribute are widely studied in the literature. For the SALBP-II assembly line balancing problem, we take account of stochastic task time changes, which is more practical than the deterministic times often assumed in industrial application. An algorithm – ENCORE, which leverages the traditional algorithm SALOME2, is proposed to address the assembly line balancing problem with stochastic task time attribute. Computational and statistical experiments are conducted to show the efficiency of proposed algorithms over traditional methods with regards to the improvement of total production times.
The aim of this research is to demonstrate how human learning models can be integrated into discrete event simulation to examine ramp-up time differences between serial and parallel flow production strategies. The experimental model examined three levels of learning rate and minimum cycle times. Results show that while the parallel flow system had longer ramp-up times than serial flow systems, they also had higher maximum throughput capacity. As a result, the parallel flow system frequently outperformed lines within the first weeks of operation. There is a critical lack of empirical evidence or methods that would allow designers to accurately determine what the critical learning paramters might be in their specific operations, and further research is needed to create predictive tools in this important area.
The learning curve is a fundamental model used by engineers in cost estimating. In industry, it is typical to use the deterministic model for projecting cost, which is also suggested in standard textbooks. However, the parameters for the model are obtained from actual data, which usually come from a stochastic process. In this technical note, we investigate a particular phenomenon of the stochastic learning model that indicates that a bias may exist in the parameter estimates simply due to random behaviour in learning. The findings suggest that, on average, projections of cost from a model whose parameters are estimated from early data points are, on average, optimistic about the future cost reduction.
The time that a part may spend in a buffer between successive operations is limited in some manufacturing processes. Parts that wait too long must be reworked or discarded due to the risk of quality degradation. In this paper, we present an analytic formulation for the steady-state probability distribution of the time a part spends in a two-machine, one-buffer transfer line (the part sojourn time). To do so, we develop a set of recurrence equations for the conditional probability of a part's sojourn time, given the number of parts already in the buffer when it arrives and the state of the downstream machine. Then we compute the unconditional probabilities of the part sojourn time using the total probability theorem. Numerical results are provided to demonstrate how the shape of the distribution depends on machine reliability and the buffer size. The analytic formulation is also applied to approximately compute the part sojourn time distribution in a given buffer of a long line. Comparison with simulation shows good agreement. © 2015 Elsevier B.V. and Association of European Operational Research Societies (EURO) within the International Federation of Operational Research Societies (IFORS). All rights reserved.
The effect of workers’ learning curves on the production rate in manual assembly lines is significant when producing relatively small batches of different products. This research studies this effect and suggests applying a work-sharing mechanism among the workers to improve the makespan (time to complete the batch). The proposed mechanism suggests that adjacent cross-trained workers will help each other in order to reduce idle times caused by blockage and starvation. The effect of work sharing and buffers on the makespan is studied and compared with a baseline situation, where the line does not contain any buffers and work sharing is not applied. Several linear programming and mixed-integer linear programming formulations for makespan minimization are presented. These formulations provide optimal work allocations to stations and optimal parameters of the work-sharing mechanism. A numerical study is conducted to examine the effect of buffers and work sharing on the makespan reduction in different environment settings. Numerical results are given along with some recommendations regarding the system design and operation.
For almost a century, researchers and practitioners have studied learning curves in production economics. Learning, in this context, refers to performance improvements of individuals, groups or organizations over time as a result of accumulated experience. Various learning curves, which model this phenomenon, have been developed and applied in the area of production economics in the past. When developing planning models in production economics, the question arises which learning curve should be used to best describe the learning process. In the past, the focus of the literature has been on empirical studies that investigated learning processes in laboratory settings or in practice, but no effort has been undertaken so far to compare existing learning curves on a large set of empirical data to assess which learning curve should be used for which application. This study systematically collected empirical data on learning curves, which resulted in a large database of empirical data on learning. First, the data contained in the database is categorized with the help of meta-tags along different characteristics of the studies the data was taken from. Second, a selection of well-known learning curves is fitted to the empirical datasets and analyzed with regard to goodness of fit and data characteristics. We identify a set of data/task characteristics that are important for selecting an appropriate learning curve. The results of the paper may be used in production economics to assist researchers to select the right learning curve for their modeling efforts.
In a two-sided assembly line, tasks can be executed simultaneously on both sides of the line. One task cannot be started until both of its direct predecessors on the left and right sides are completed. Therefore, the start time of the task is the maximum of the two predecessors’ finish times. In many realistic situations, it is assumed that the task times are independent and normally distributed with known means and variances. However, the maximum of two normal variables is not normally distributed, but can be well approximated by results from extreme value theory. In this paper, we utilise these results to develop a solution methodology to balance two-sided assembly lines with stochastic task times, minimising the line length and the number of stations while guaranteeing all tasks are completed within the cycle time with a given confidence level.
This paper studies the effects of learning and forgetting on a two-stage production system and the position of a potential bottleneck in the system. We start with developing a model for a two-stage serial production system where semi-finished items are fed by the first stage to the second stage, which, in turn, processes the items to their final state. The finished items are transferred either to a subsequent stage or to customers. The paper assumes that both stages of the production system considered are subject to learning and forgetting effects. Learning quickens the production rate as more experience is gained (i.e., when the number of repetitions increases), while forgetting has the opposite effect when production is intermittent (i.e., experience is lost over production breaks). The paper studies how different values of the learning and forgetting parameters influence the ratio of the production rates of both stages and the flow of material in the system. The results of the paper indicate that learning may cause a bottleneck to shift its position in a production system. This happens when an initially slower stage overtakes a previously faster stage over time due to a higher learning rate. The paper thus contributes to the literature on moving bottlenecks and provides practitioners with a model that helps predicting where bottlenecks may arise in the production system, and which enables the system to flexibly react to moving bottlenecks.
In the ramp-up phase, or time to volume of new products, pronounced learning effects are observed. They are present especially on assembly lines producing mass-goods because of a high number of repetitions of the tasks. Shortening the ramp-up phase and reaching the steady-state production as soon as possible generates main advantages for firms that introduce new products. Moreover, a careful planning of the ramp-up stage is getting even more important in view of shorter product life cycles and a growing importance of the 'time to payback' financial indicators. Former studies on incorporation of learning effects into assembly line balancing have limited applicability, because they rely on unrealistic assumptions. We model learning effects, based on general and realistic assumptions, as an extension of the Simple Assembly Line Balancing Problem. We propose exact and heuristic solution procedures and perform extensive computational tests. We found that for instances similar to the problems, which arise in firms, the duration of the learning stage can be reduced by up to 10% if our specialised methods are applied.
Order picking is a time-intensive and costly logistics process as it involves a high amount of manual human work. Since order picking operations are repetitive by nature, it can be observed that human workers gain familiarity with the job over time, which implies that learning takes place. Even though learning may be an important source of efficiency improvements in companies, it has largely been neglected in planning order picking operations. Mathematical planning models of order picking that have been published earlier thus provide an incomplete picture of real-world order picking, which affects the quality of the planning outcome. To contribute to closing this research gap, this paper presents an approach to model worker learning in order picking. First, the results of a case study are presented that emphasize the importance of learning in manual order picking. Subsequently, an analytical model is developed to describe learning in order picking, which is then evaluated with the help of numerical examples. The results show that learning impacts order picking efficiency. In particular, the results imply that worker learning should be considered when planning order picking operations as it leads to a better predictability of order throughput times. In addition, the effects of learning are relevant for the allocation of available resources, such as the allocation of workers to different zones of the warehouse. The results of the numerical analysis indicate that it is beneficial to assign workers with the lowest learning rate in the workforce to the fast moving zone to gain experience.
The learning curve or start-up model is extended to three examples of labour-intensive manufacture-automobile assembly, apparel manufacture and the production of large musical instruments. Prediction of the parameters of the model and the occurrence of steady-state "plateaux" are examined in relation to the findings. The results of the investigation suggest the wider applicability of the learning curve model in labour-intensive industries, provide additional data to evaluate an empirical approach to parameter prediction and demonstrate the difficulty of generalizing about the occurrence and predictability of steady-state plateaux. In addition, the application of different cost indices in learning curve research is demonstrated among new product start-ups in both single and multiple facilities.
Mainly deal with queueing models, but give the properties of many useful statistical distributions and algorithms for generating them.
Models are developed which 'explain' why new automobile assembly systems are abandoning traditional moving assembly line concepts if human operators perform most of the tasks. The new systems incorporate parallel stations, asynchronous work flow and small inventories. However, the paralleling creates problems in synchronizing parts delivery, and models are developed to illustrate these problems and the effect of alternative part delivery systems and the impact of job resequencing requirements.