Describing the influence of set-up optimised sequencing on output lateness of workstations

Productivity and schedule reliability are two key objectives when designing, planning and controlling an enterprise’s internal supply chain. Productivity can be increased by implementing set-up optimised order sequencing, which reduces the time and costs required for set-ups. However, since planned end-of-order-processing dates are ignored with set-up optimised sequencing, it tends to cause increased lateness and thus poorer schedule reliability. Accordingly, enterprises must weigh the highest possible productivity against a high schedule reliability. In this paper, we develop a logistic model that can be used by enterprises to facilitate describing the impact of set-up optimised sequencing on a workstation’s lateness behaviour. This logistic model depicts the extent of the output lateness’ standard deviation as a function of the mean WIP, mean workstation output and the input lateness’ standard deviation when applying set-up optimised sequencing. It is thus possible to quantitatively describe the effects of set-up optimised sequencing on the logistic objective ‘schedule reliability’ and compare possible productivity gains.     

Extending Little’s Law to single order throughput times

It has long been established that the mean throughput time is a result from the relationship between the work in process (WIP) and output rate. In reality, however, throughput times are frequently dispersed broadly around this mean. With the aid of a model originally developed for lateness, it is now possible to describe the throughput times of individual orders. Accordingly, the throughput time can be divided into two components: WIP-dependent throughput time and sequence-dependent throughput time.     

Decentralized WIP-oriented manufacturing control (DEWIP)

DEWIP is a manufacturing control system for job shop environments aiming at achieving short and reliable lead times by establishing WIP control loops between the manufacturing work centres. The paper describes the mode of function, the setting of parameters and simulation results of the new manufacturing control system. The setting of parameters is done with the aid of the funnel model and the theory of logistic operating curves, both developed at the Institute of Production Systems at the University of Hanover. The simulation is conducted using industrial data and makes it possible to assess DEWIP with regard to lead times, WIP level, performance and schedule reliability. DEWIP is compared both with an uncontrolled process and with the manufacturing control systems Load oriented order release (LOOR), Conwip and Polca. The results suggest that DEWIP and the models employed for the setting of parameters are suitable for job shop production and therefore offer a valuable alternative to prevailing centralized manufacturing control systems.     

Workload Control and Order Release: A Lean Solution for Make‐to‐Order Companies

Protecting throughput from variance is the key to achieving lean. Workload control (WLC) accomplishes this in complex make‐to‐order job shops by controlling lead times, capacity, and work‐in‐process (WIP). However, the concept has been dismissed by many authors who believe its order release mechanism reduces the effectiveness of shop floor dispatching and increases work center idleness, thereby also increasing job tardiness results. We show that these problems have been overcome. A WLC order release method known as “LUMS OR” (Lancaster University Management School order release) combines continuous with periodic release, allowing the release of work to be triggered between periodic releases if a work center is starving. This paper refines the method based on the literature (creating “LUMS COR” [Lancaster University Management School corrected order release]) before comparing its performance against the best‐performing purely periodic and continuous release rules across a range of flow directions, from the pure job shop to the general flow shop. Results demonstrate that LUMS COR and the continuous WLC release methods consistently outperform purely periodic release and Constant WIP. LUMS COR is considered the best solution in practice due to its excellent performance and ease of implementation. Findings have significant implications for research and practice: throughput times and job tardiness results can be improved simultaneously and order release and dispatching rules can complement each other. Thus, WLC represents an effective means of implementing lean principles in a make‐to‐order context.     

Shop Floor Control When Tacit Worker Knowledge is Important

We investigate a shop where the workers and supervisors have tacit knowledge of how to operate efficiently and where efficiency is important to providing capacity to meet demand. This tacit knowledge includes setup dependencies between products as well as which worker or machine is best suited for a particular product. We discuss a real‐world shop where this is the case. Management expects workers and supervisors to use their knowledge to schedule efficiently by monitoring their performance based on standards. The question that we explore is how management should control for due date performance in light of the discretion given to the workers and supervisors to sequence jobs on the basis of efficiency. We explore management control of due date performance through the use of order review and release (ORR) and management expediting. We find that although ORR is quite effective at reducing work‐in‐process (WIP) inventories, it may foster very late deliveries in a shop such as this. In fact, under such conditions, deftly executed expediting with no ORR at all can be far more effective at supporting all deliveries. Even improving ORR into a hybrid by actively updating path efficiencies (observed from supervisor/worker scheduling) did not support a change to this conclusion. Conversely, when conditions are created where tacit knowledge plays a reduced role or utilization is decreased, ORR delivers in a timely manner. The interaction between utilization, WIP levels, and worker knowledge all help dictate the appropriate control methodology.     

Synchronous flow shop problems: How much can we gain by leaving machines idle?

In synchronous production lines it may be beneficial to leave machines idle instead of processing the next job immediately. In this paper, the effects of inserting voluntary idle times are discussed in more detail for different objective functions (minimization of makespan, total completion time, maximum lateness). Besides deriving theoretical bounds on how much can be gained by inserting idle times, an extensive computational study is conducted to empirically examine the actual improvements. For this, exact algorithms and heuristics capable of incorporating voluntary idle times are proposed to find (near-) optimal schedules. It can be seen that the potential gain is very large in theory, while the empirical results indicate that in general only small improvements are achievable in practice.     

资源受限下平行工序顺序对优化的0-1规划模型

资源受限是工程项目时刻都可能面对的挑战。由于资源限制，需要将原项目计划中相互之间无优先关系的平行工序调整为顺序工序。平行工序顺序化可导致项目工期延迟，因此需考虑如何使项目工期延迟最小。该平行工序顺序优化问题是项目调度问题，也是排列组合问题，通常难度很大，包括一些NP-hard问题。本文主要研究该问题的一类典型子问题——平行工序顺序对优化，即如何将项目中某2n个平行工序调整为n个顺序工序对，并且对项目工期的影响最小。该问题的总方案数可达到（2n）!/n!。本文借助工序网络（如CPM网络），运用简单的时间参数量化了平行工序顺序化对项目工期的影响，进而降低问题的求解难度，建立了纯0-1规划模型。实验验证了该模型的求解效率，求解100个平行工序规模的问题平均耗时0.2605秒，而求解500个平行工序规模的问题平均耗时10.66秒。     

An intelligent workstation controller for integrated planning and scheduling of FMS cell

This paper presents an experimental design developed to determine a combination of robust planning and scheduling rules for an intelligent workstation controller (IWC). The IWC is used as part of the control system for an automated flexible manufacturing system. A three-level hierarchical control structure (shop, workstation and equipment)is adopted in order effectively to control a shop-floor. At the top level is a shop controller which receives orders and their associated manufacturing information, and manages interactions among workstations. The IWC defines and resolves the production control activities necessary to coordinate a group of equipment controllers so as to ensure the completion of orders. Specifically, the IWC is responsible for selecting a specific process routeing for each part, allocating resources, scheduling and coordinating the activities across the equipment, monitoring the progress of activities, detecting and recovering from errors, and preparing reports. These activities are accomplished using planning, scheduling, and execution functions. In order to demonstrate the effectiveness and robustness of the IWC, all the controllable and uncontrollable factors need to be identified and detailed. Controllable factors are those which are set by the controller and cannot be directly affected by the production environment. Uncontrollable factors are those which vary with the production environment and cannot be directly changed by the controller developer. The objective of the paper is to illustrate how a good set of controllable factors that dictate the IWC's effectiveness and robustness over various uncontrollable factors can be determined. Due to the number of possible combinations of all the factors, a complete validation of the effectiveness and robustness of the IWC is extremely time consuming and far beyond the scope of this paper. Therefore, a few planning and scheduling strategies are selected and a formal experiment is conducted. The experiment illustrates how significant performance dependencies for various planning and scheduling strategies can be identified.     

An Application of Master Schedule Smoothing and Planned Lead Time Control

Make‐to‐order (MTO) manufacturers must ensure concurrent availability of all parts required for production, as any unavailability may cause a delay in completion time. A major challenge for MTO manufacturers operating under high demand variability is to produce customized parts in time to meet internal production schedules. We present a case study of a producer of MTO offshore oil rigs that highlights the key aspects of the problem. The producer was faced with an increase in both demand and demand variability. Consequently, it had to rely heavily on subcontracting to handle production requirements that were in excess of its capacity. We focused on the manufacture of customized steel panels, which represent the main sub‐assemblies for building an oil rig. We considered two key tactical parameters: the planning window of the master production schedule and the planned lead time of each workstation. Under the constraint of a fixed internal delivery lead time, we determined the optimal planning parameters. This improvement effort reduced the subcontracting cost by implementing several actions: the creation of a master schedule for each sub‐assembly family of the steel panels, the smoothing of the master schedule over its planning window, and the controlling of production at each workstation by its planned lead time. We report our experience in applying the analytical model, the managerial insights gained, and how the application benefits the oil‐rig producer.     

Setting corporate goals and measuring organizational effectiveness--a practical approach.

Shareholders, customers, suppliers, lenders, the community, government and regulatory agencies and employees--the stakeholders in an organization--are the ones entitled to anticipate the satisfaction of societal demands placed on the organization. Strategic planning has to take these into account in order to be able to plan how the demands can be met. Therefore, the stakeholders as such must be identified and placed in an order of priority. There are those without whom the organization could not exist and those who have a judicial position in relationships to the organization. It then becomes important to establish the criteria by which each will judge the organization's effectiveness. One way is to approach the stakeholder directly. Another is to use a deductive approach based on a literature survey or polling representatives. In practice a combination can be used. If this is done then the crucial effectiveness criteria for each stakeholder group can be identified. It is important, of course, to realize that not only can these criteria change, but also the determinants of stakeholder power can vary. They depend on what resources each group possesses, whether there are alternative stakeholders, what authority and influence is held. In turn the effectiveness will be influenced by the pressures brought to bear on the power of the stakeholder. These can progress from societal pressures to political issues to legislated requirement and, finally, to punitive action. Another way of regarding them is as moving from strategic to emerging to current. An appreciation of this will help environmental scanning.     

Multi-objectives exception management model for semiconductor back-end environment under turnkey service

On-time delivery is the need for customer satisfaction which is a critical factor in the survival of the semiconductor industry. The on-time-delivery performance of the whole semiconductor turnkey service depends on the performance of the back-end factories. Unfortunately, undesirable and inevitable production variations make it difficult to maintain and improve a factory's performance and more objectives such as cycle time, throughput rate and the due-date accuracy need to be simultaneously considered. This paper presents an exception management model in order to compromise the contradicting needs of delivery accuracy, throughput rate and cycle time. The exception model can be divided into three parts: (1) an integrated performance index is proposed to compromise multiple performance measures; (2) an AWDL (available WIP deviation level) determination model is designed to gather proper AWDLs for triggering exceptions and (3) a WIP correction action is proposed to make abnormal WIP levels back to normal levels as soon as possible. To evaluate the proposed WIP exception management model, a simulation model is constructed and experiments are then conducted. The simulation results show that the proposed model helps back-end factories to set proper exception triggering conditions, reduce uncertainty occurrences and achieve better performances on due dates.     

Manufacturing cycle time reduction using balance control in the semiconductor fabrication line

In semiconductor manufacturing, wafer fabrication is the most complicated and important process, and is composed of several hundred process steps and involves several hundred machines. The productivity of the manufacturing process depends mainly on controlling the balance of WIP (work-inprogress) flow to achieve maximum throughput under short manufacturing cycle times. This paper discusses how to determine the proper WIP level for operations, against which balance status can be measured. Balance measurement can be applied in mathematical modelling for bottleneck scheduling and operations management of the fabrication line. Performances are evaluated through computational experiments to show that balance driven management leads to 15-33% more production in 21% shorter manufacturing cycle time than production driven management.     

基于瓶颈负荷的BTO供应链响应时间优化

供应商生产的低效率是BTO供应链实际运作过程中的瓶颈,主要表现为供应链整体响应时间长和在制品库存高.针对该类供应链的瓶颈特点,利用排队模型建立了一种基于瓶颈负荷的BTO供应链响应时间优化策略.然后,通过算例分析了订单数量、等待订单数量和产能利用率,以及系统内的概率分布等参数对在制品库存水平和供应链整体响应时间的影响.最后,得出的结论表明,基于瓶颈负荷的控制对缩短供应链整体响应时间和在制品库存水平的控制均是有效的.     

某大型装置现场可靠性的贝叶斯分析

可靠性、维修性和可用性是描述复杂装置质量属性的三个重要指标。在某大型装置的现场故障和维修数据的经验分析基础上,运用分段线性模型确定了装置可靠性改进趋势。针对装置现场故障和维修数据的小样本问题,给出了装置现场可靠性和维修性的贝叶斯模型,并用WinBUGS软件对其进行了求解计算。继而分析给出了装置可用性最恰当描述——任务可用度的闭合表达形式。最后,通过与极大似然法进行对比,发现应用贝叶斯方法分析该大型装置现场可靠性更有效。     

Coordinated scheduling of production and delivery stages with stage-dependent inventory holding costs

This paper considers a problem of integrated decision-making for job scheduling and delivery batching wherein different inventory holding costs between production and delivery stages are allowed. In the problem, jobs are processed on a facility at a production stage and then delivered at the subsequent delivery stage by a capacitated vehicle. The objective is to find the coordinated schedule of production and delivery that minimizes the total cost of the associated WIP inventory, finished product inventory and delivery, where both the inventory costs are characterized in terms of the weighted flow-time and the delivery cost is proportional to the required number of delivery batches. It is proved that the problem is NP-hard in the strong sense. Thereupon, three heuristic algorithms are derived. Some restricted cases are also characterized as being solvable in polynomial time. Numerical experiments are conducted to evaluate the performance of the derived heuristic algorithms.     

The effective use of computer resources

The distribution over time of jobs arriving for service at a computer installation will affect the turnaround time experienced by them, as does the sequence of service for jobs having arrived. However, if the pattern of delays due to congestion can be determined in advance, the submission times of jobs may be adjusted to increase the effectiveness of the system.A theoretical model is developed in which users relate the net value of their jobs to submission time and turnaround time and the organization imposes submission and sequencing rules to increase the net gain of the system. A simulation is used to illustrate specific system environments and evaluation methods.     

THROUGHPUT TIME REDUCTION: TAKING ONE'S MEDICINE

Managing throughput time and its relation to work-in-process (wip) inventory and customer service is the focus of this paper. This research combines theory, simulation results, and the analysis of corporate data in an effort to address the issues associated with how one company (Eli Lilly) managed a reduction in their throughput times and an improvement in their delivery reliability. The results for this company suggest that production control decisions—expediting and de-expediting—can lead to a vicious circle of decisions, which in turn can lead to increased levels of WIP inventory and higher and more unpredictable throughput times.     

The effect of setup time reduction on its variance

Earlier research has found that the variability in setup time can play a central role in which it produces paradoxes and can harm effective waiting time, WIP, and capacity as one reduces setup time. Specifically, if setup time is reduced without a reduction in its variance, the waiting time in the system can increase and so can WIP. Of course, the matter worsens if setup variance increases as one reduces its duration. Given such findings, it becomes critical to find out what happens to setup time variance when we cut setup time. It is unlikely that setup variability can increase as one reduces setup duration.     

An investigation of output flow control,bottleneck flow control and dynamic flow control mechanisms in various simple lines scenarios

Flow control mechanisms have been a topic of academic research for several years. With the growth of business-wide information systems such as enterprise resource planning and supply chain, better planning, scheduling and control of the business transformation process is required in order to achieve increased throughput, reduced inventories, shorter lead times and reduced tardiness. This research compares two new approaches to flow control, output flow control and bottleneck flow control to a real-time flow control system, dynamic flow control. Both output and bottleneck flow control mechanisms are much simpler to implement and manage than dynamic flow control in that they do not require continual feedback and rescheduling. Line characteristics, such as location of breakdowns with respect to the bottleneck, the location of the bottleneck when breakdowns occur, and the impact of variability of processing times on the performance measures (output, WIP level, lateness, and number of tardy jobs) for these three flow control mechanisms are compared. Both output and bottleneck flow control mechanisms perform favourably (particularly bottleneck) under different scenarios and warrant further study across a wider range of scenarios (mixed models, job shops, etc.).     

Lot sizing in workload control systems

An important basis for workload control (WLC) is the existence of functional relationships between the mean level of work-in-process (WIP) and the values of important goal variables, like average flow time, capacity utilization, etc. These functional relationships are largely influenced by the lot sizes. This means that the usual objective of lot sizing must be supplemented by considering the impact of lot sizes on the relationships between WIP and the other goal variables. Here it is shown that this insight leads to flow time oriented lot sizing models. This type of lot sizing models is analysed. It is argued that the derivation of simple rules for lot sizing is an important research topic, and a model for deriving such rules is presented. Some rules are derived from the model, assuming that the batches are processed by an M/G/1 server, and it is shown that these rules support insights based on simulation in the 1980s. Topics for future research are outlined as well.