Minimising throughput loss in assembly lines due to absenteeism and turnover via work-sharing

In this paper we analyse the loss of throughput rate of assembly line caused by slow pace of substitute workers (replacing absentees) having no prior experience in the required tasks. We proposed work-sharing mechanisms that improve the balance of the workload during the learning period. The proposed mechanisms add to the experienced neighbouring workers some of the workload of the inexperienced worker substituting an absentee. We call this workload ‘shared work’. After the performance of the substitute workers improves due to learning, the shared work is re-assigned to them (relieving their experienced neighbours). We provide analytic expressions for the line throughput rate, which is determined by sets of bottleneck workstations. These sets of consecutive workstations consist of the inexperienced workers replacing the absentees and the experienced workers assisting them during the learning periods. The decision variables of this model are: (1) the amount of shared work, and (2) the time in which the shared work is re-assigned to the substitute worker. Unique optimal values of these two variables are found via numerical study, for buffered and non-buffered lines. Experiments show that the proposed work-sharing mechanisms can significantly improve the line’s throughput, compared to the original system without work-sharing.     

A novel metric for determining the constraining effect of resources in manufacturing via simulation

This paper provides a novel method for determining the constraining effect of resources in a manufacturing system using discrete event simulation. Traditionally manufacturing systems are constrained by one or more bottlenecks. Eliminating or mitigating the bottleneck will speed up the system throughput. However, bottlenecking resources generally only refer to machines, and primarily focus on flow-shops not job-shops. One important resource we believe that is often overlooked is workers and their associated skills, and we propose that a particular skill could be flagged as a bottleneck resource. We define new metrics known as resource constraint metrics (RCM) for measuring the constraining effect of a resource on the entire manufacturing system. These metrics are flexible and differentiate between the constraining effects of machines and their requested skills. The metrics can also deal with complex workflows with alternative routing, alternative resources, calendars (a necessary consideration when dealing with workers), worker performance, and multiple modes of operation of machines (e.g. run, setup, and maintenance). The use of RCMs in simulation aids in real-world decision-making, by determining which resource should be focussed on and improved to reduce the overall system feeling constrained. This will have the effect of increasing throughput or at least providing the capacity for increased throughput.     

Cross-training policies and team performance

Labour flexibility is an important issue in the design and development of teams. By means of an analytical study this paper investigates the impact of cross-training on team performance. Four cross-training policies are distinguished and compared according to their effects on important performance measures, such as the load of the bottleneck worker and the number of newly used qualifications seen in various situations. These measures indicate a team's effectiveness and efficiency. Furthermore, they are related to psychological, social and organizational aspects of team functioning. A task assignment heuristic serves as a tool for comparing the cross-training policies. This is applied in an experimental study in which the team situation varies with respect to the cross-training policy, the level of absenteeism, and the variation in the mix of orders. It is shown that a worker-oriented cross-training policy, which attempts to spread multi-functionality evenly among employees, performs well. This policy offers the opportunity to shift work among employees to keep the load of the bottleneck worker low. The required coordination effort for allocating workers to tasks is also relatively low. In line with previous research, this study shows a diminishing positive effect of expanding the level of labour flexibility. The coordination effort required to apply additional flexibility appears to increase linearly with the amount of additional cross-training. The diminishing positive effect and the linearly increasing coordination effort suggest that managers should critically consider the level of labour flexibility in worker teams.     

G公司生产流程的瓶颈分析与改进

采用瓶颈模型对G公司生产流程的瓶颈进行了分析,并提出改进措施,数据分析表明,所提出的改进方案对提高企业有效产出率有效.     

Data driven bottleneck detection of manufacturing systems

Bottlenecks within a production line significantly reduce the productivity. Quick and correct identification of the bottleneck locations can lead to an improvement in the operation management of utilising finite manufacturing resources, increasing the system throughput, and minimising the total cost of production. Most of the current bottleneck detection schemes focus on the long-term bottleneck detection problem and an analytical or simulation model is usually needed. Due to recent developments, short-term process control and quick decision making on the plant floor have emerged as important qualities for operation management. This research proposes a new data driven method for throughput bottleneck detection in both the short and long term. The method utilises the production line blockage and starvation probabilities and buffer content records to identify the production constraints without building an analytical or simulation model. The method has been verified both analytically and by simulation. An industrial case study has also been used in order to demonstrate the implementation and validate the efficiency of the proposed bottleneck detection method.     

A data-driven approach to multi-product production network planning

The clearing function models the non-linear relationship between work-in-process and throughput and has been proposed for production planning in environments with queuing (congestion) effects. One approach in multi-product, multi-stage environments has been to model the clearing function at the bottleneck machine only. However, since the bottleneck shifts as the product release mix changes, this approach has its limitations. The other approach is the Alternative Clearing Function formulation, where the clearing function is first estimated at the resource level using piecewise linear regression from simulation experiments, and then embedded into a linear programme. This paper develops an alternative to the Allocated Clearing Function formulation, wherein system throughput is estimated at discrete work-in-process points. A mixed integer programming formulation is then presented to use these throughput estimates for discrete release choices. The strength of the formulation is illustrated with a numerical example and the new approach is compared with the ACF.     

Simulation study of a bottleneck-based dispatching policy for a maintenance workforce

Maintenance is important for production operations and for continuous improvement. Appropriate dispatching of the maintenance workforce to quickly respond to equipment failures and carry out preventive services can improve system productivity. The first-come-first-served policy is typically used in many manufacturing industries. In this paper, we present a priority-based dispatching policy, a dynamic bottleneck policy, based on the analysis of real-time data. In such a policy, priority is assigned to the bottleneck machine after a fixed time period, and the maintenance worker will service the high-priority machine (i.e. bottleneck machine) first when multiple service requests are received. It is shown by extensive simulation experiments that this policy can lead to a greater improvement in system throughput compared with the first-come-first-served policy. To implement such a policy, the appropriate time period for data collection and the frequency for carrying out bottleneck analysis are investigated. In addition, a sensitivity study suggests that the results obtained are insensitive to machine downtime, efficiency, and reliability models.     

一个改进的单抓钩周期性排序模型及其在自动化学处理线中的应用

计算机控制的抓钩广泛用于自动化学处理线的工件的运送。抓钩的排序直接影响系统的生产率，抓钩排序的目标是对运送进行排序以极大化生产率。当某工序处理时间非常长时，该工序成为瓶颈。为了去除该瓶颈，系统可以为该工序设计多个处理槽，这称为“多重处理槽”问题。本文提出一个改进的混合整数规划模型以求解有“多重处理槽”的单抓钩周期性排序问题的最优解。实例表明所提出的方法是有效的。     

The effect of variability on the unbalanced production line

Several algorithms have been presented for manpower scheduling of the unbalanced production line. This paper uses the best performing algorithm. One Worker At A Time (OWAAT), as a vehicle to assess the effect of variability in selected factors on the performance of the unbalanced production line. Several different performance measures were considered; results indicate that worker efficiency, size of buffer stocks and method for penalizing shortages significantly affect performance, whereas worker absenteeism does not. Additional observations were also made on the effect of these factors on the number and size of shortages.     

Bottleneck detection of complex manufacturing systems using a data-driven method

Bottlenecks within a production line significantly affect system productivity. Most current bottleneck detection schemes focus on the long-term bottleneck detection problem using an analytical or simulation model. Furthermore, these studies are restricted to serial tandem lines only. This research focuses on extending the newly developed data-driven method for throughput bottleneck detection from a serial line to a manufacturing system with a complex layout. Within these complex systems, two specific layouts are considered: the concurrent process and the closed-loop feedback process. The method is verified using simulation case studies. An industrial case study is examined to demonstrate the practicality of this approach and to validate the efficiency of the proposed bottleneck detection method.     

Chained cross-training of assembly line workers

To function properly, assembly lines require the presence of every worker. When a worker is absent, management must scramble quickly to find a replacement. Cross-training workers to perform multiple tasks mitigates this difficulty. However, since cross-training is costly and limited by learning capacity and can confound the search for quality problems, it should be used judiciously. The present paper proposes a training strategy called chaining in which workers are trained to perform a second task, and the assignments of task types to workers are linked in a chain. It is shown that chaining is a practical and effective strategy for prioritizing cross-training to compensate for absenteeism on assembly lines.     

Flexible capacity design for the Focus Factory – a case study

The objective of this research study is to predict throughput and material flow requirements under a number of different scenarios with a focus on designing flexible capacity. The use of discrete-event simulation integrated with Excel provides proactive decision support for strategic, long-term capacity planning. This study presents a comprehensive capacity analysis of a Focus Factory at 3M, a Fortune 500 manufacturing company. Traditional capacity analysis typically provides throughput per given time over a set time horizon. The simulation model developed predicts monthly throughput, identifies bottleneck areas, and determines resource requirements for the Focus Factory while properly sizing storage and staging areas of raw materials, work-in-process, and finished goods. The decision support model provides a tool to the business team that links the sales plan with the “best” operational scenario. It is also used in cost-benefit and sensitivity analyses to predict capacity shortfalls and plan for capital investment as the business grows.     

Determination of batch size at a bottleneck machine in manufacturing systems

In manufacturing systems, there often exists a bottleneck machine whose capacity is equal to or less than the market demand. Any idle or waste time at the bottleneck machine directly impacts the output of the entire plant because it results in a loss of throughput. In order to maximize the capacity utilization by less setup losses at the bottleneck machine, the parts are often produced in batches. Traditionally, most batch sizing decisions are made based on the economic order quantity model where setup and inventory holding costs are considered. This paper presents an alternative method to determine batch size at a bottleneck machine. We present a new objective function and cost factors for batch sizing and investigate queuing and throughput models. A linear search algorithm is introduced to find the optimal throughput rate and batch size at the same time. Numerical examples are examined to see how the batching algorithm works.     

A statistical framework of data-driven bottleneck identification in manufacturing systems

Data-driven bottleneck identification has received an increasing interest during the recent years. This approach locates the throughput bottleneck of manufacturing systems based on indicators derived from measured machine performance metrics. However, the variability in manufacturing systems may affect the quality of bottleneck indicators, leading to possible inaccurate detection results. This paper presents a statistical framework (SF) to decrease the data-driven detection inaccuracy caused by system variability. Using several statistical tools as building blocks, the proposed SF is able to analyse the logical conditions under which a machine is detected as the bottleneck, and rejects the proposal of bottleneck when no sufficient statistical evidence is collected. A full factorial design experiment is used to study the parameter effects of the SF, and to calibrate the SF. The proposed SF was numerically verified to be effective in decreasing the wrong bottleneck detection rate in serial production lines.     

The effects of workforce heterogeneity on productivity in an experiential learning environment

This paper investigates the effects of heterogeneity in various workforce learning/forgetting characteristics on productivity. An important aspect of this study is that worker heterogeneity is specifically modeled and examined by including the variance of key parameters as controlled experimental factors in addition to these parameters’ means. Initially, we analytically investigate the impact of heterogeneity in each of the four parameters in the learning/forgetting model used in this study. The analytical results are subsequently verified and further investigated using computer simulation. Specifically, seven macro-level factors are investigated including the rate of turnover, the mean and variance of the latent initial expertise existing in the workforce for the tasks to be performed, the mean and variance of worker steady-state productivity levels, the variance of the rate of worker learning, and the variance of working forgetting. The simulation models were developed and informed by empirical data from industry. The results indicate significant effects for mean initial expertise, mean steady-state productivity, the variance of initial expertise, the variance of the rate of worker learning, and the variance of worker forgetting. Results have managerial implications in the comparison and selection of various workforce populations for production systems.     

A closed-loop supply chain with stochastic product returns and worker experience under learning and forgetting

This paper addresses a single-manufacturer single-retailer closed-loop supply chain with stochastic product returns considering worker experience under learning and forgetting in production and inspection of returned items at the manufacturer. Customer demand is assumed to be dependent linearly on the retail price, and it is fulfilled by using both manufactured and remanufactured products. The manufacturer delivers the buyer’s order quantity in a number of equal-sized batches. The optimal number of shipments, the shipment size and the retail price are determined by maximising the average expected profit of the closed-loop supply chain. It is observed from the numerical study that high learning effects in production and inspection lead to high recovery rates of used products, which, besides an economic advantage, may have a positive effect on the environment. Even though forgetting has an adverse effect, the average expected profit of the closed-loop supply chain is much higher than that of the basic model which ignores worker learning.     

RAPID for high-performance computing systems: architecture and performance evaluation

The limited bandwidth and the increase in power dissipation at longer communication distances and higher bit rates will create a major communication bottleneck in high-performance computing systems (HPCS), affecting not only their performance, but also their scalability. As a solution, we propose an optical-interconnect-based architecture for HPCS called reconfigurable all-photonic interconnect for parallel and distributed systems (RAPID) that alleviates the bandwidth density, optimizes power consumption, and enhances scalability. We also present two cost-effective design alternatives of the architecture, a modified version called M-RAPID and an extended version called E-RAPID that minimizes the cost of the interconnect based on the number of transmitters required. We perform a detailed simulation of the proposed RAPID architecture and compare it to several electrical HPCS interconnects. Based on the performance study, RAPID architecture shows 30%-50% increased throughput and 50%-75% reduced network latency as compared to HPCS electrical networks.     

Job order releasing and throughput planning for multi-priority orders in wafer fabs

To meet the production target of multi-level (multiple priority rank) orders in wafer fabs, this paper uses a hierarchical framework based on a mathematical model, and without the assistance of any simulation tool, to build a production scheduling system to plan wafer lot releasing sequence and time. This system first applies capacity loading analysis to set up the batch policy for each level (rank) of orders. Next, the production cycle time of each product level is estimated with considerations of batching and loading factor. The cycle time is then used to derive system control parameters such as the most appropriate level of work in process (WIP) and the number of daily operations on the bottleneck workstation. Lastly, a Constant WIP mechanism is applied to establish a wafer release sequence table and a throughput timetable. The due date designation for each specific order can hence be confirmed. With the comparison with the result of simulation, it shows that under the designed system the performance and planning measures in the master production schedule can be drawn up quickly and accurately, and the system throughput target and due date satisfaction can be achieved. Overall, the proposed production scheduling system is both effective and practicable, and the planning results are supportive for good target planning and production activity control.     

Crane scheduling with time windows in circuit board production lines

This paper studies a crane scheduling problem with time windows in a flow-shop type production system. Feasibility of a state of the system is first discussed. Then, based on the insight derived from the mathematical programming formulation of feasibility, an optimization-based heuristic algorithm for real-time scheduling of the system is developed. Computer simulation on randomly generated problems shows that this algorithm has an excellent performance in maximizing throughput without defective jobs.     

Workload balancing capability of pull systems in MTO production

Pull systems focusing on throughput time control and applicable in situations with high variety and customisation are scarce. This paper compares three unit-based pull systems that can cope with such situations: POLCA, CONWIP and m-CONWIP. These systems control the shop floor throughput time of orders by limiting the number of orders on the shop floor. However, their effectiveness in terms of reducing total throughput time is questioned. Theory states that an improvement in the average total throughput time will be due to the workload balancing capability of a pull system, but that many pull systems lack this capability. This paper shows that this workload balancing capability exists for POLCA and m-CONWIP, but not for CONWIP. The magnitude of the effect differs strongly, depending on the configuration of the system, the order arrival pattern and the variability of the processing time of the orders.