A heuristic algorithm for capacity sensitive requirements planning

Available lot sizing rules for use in MRP (Material Requirements Planning) systems ignore capacity limitations at various work centers when sizing future orders. Planned order releases are instead determined by the tradeoff only between the item's set up and inventory holding costs. This limitation can cause unanticipated overloads and underloads at the various work centers, along with higher inventories, poorer customer service, and excessive overtime.This research explores one way to make MRP systems more sensitive to capacity limitations at the time of each regeneration run. A relatively simple heuristic algorithm is designed for this purpose. The procedure is applied to those planned order releases that standard MRP logic identifies as mature for release. The lot sizes for a small percentage of these items are increased or decreased so as to have the greatest impact in smoothing capacity requirements at the various work centers in the system. This algorithm for better integrating material requirements plans and capacity requirements plans is tested with a large scale simulator in a variety of manufacturing environments. This simulator has subsequently undergone extensive tests, including its successful validation with actual data at a large plant of major corporations.Simulation results show that the algorithm's modest extension to MRP logic significantly helps overall performance, particularly with customer service. For a wide range of test environments, past due orders were reduced by more than 30% when the algorithm was used. Inventory levels and capacity problems also improved. Not surprisingly, the algorithm helps the most (compared to not using it at all as an MRP enhancement) in environments in which short-term bottlenecks are most severe. Large lot sizes and tight shop capacities are characteristic of these environments. The algorithm works the best when forecast errors are not excessive and the master schedule is not too “nervous.”This proposed procedure is but one step toward making MRP more capacity sensitive. The widely heralded concept of “closed-loop” MRP means that inventory analysts must change or “fix up” parts of the computer generated material requirements plan. What has been missing is a tool for identifying the unrealistic parts of the plan. Our algorithm helps formalize this identification process and singles out a few planned order releases each week. This information comes to the analyst's attention as part of the usual action notices. These pointers to capacity problems go well beyond capacity requirements planning (CRP) and would be impossible without computer assistance.Our study produced two other findings. First, short-term bottlenecks occur even when the master production schedule is leveled. The culprits are the lot sizing choices for items at lower levels in the bills of material. “Rough-cut” capacity planning, such as resource requirements planning, therefore is not a sufficient tool for leveling capacity requirements. It must be supplemented by a way to smooth bottlenecks otherwise caused by shop orders for intermediate items. Second, the disruptive effect of large lot sizes is apparent, both in terms of higher inventories and worse customer service. Large lot sizes not only inflate inventories, but paradoxically hurt customer service because they create more capacity bottlenecks. The only reason why management should prefer large lot sizes is if set-up times are substantial and cannot be efficiently reduced. This finding is very much in step with the current interest in just-in-time (JIT) systems.     

A comparison of MRP lot sizing methods considering capacity change costs

Capacity management is the planning and leveling of resources required (load) against the resources available (capacity). In this study, the lot size models used by Material Requirements Planning (MRP) had a major effect on the work center load profiles generated by Capacity Requirements Planning (CRP). Therefore, the selection of lot size models for MRP systems is an important decision for capacity management as well as materials management.The results of this study highlight the operating characteristics of specific lot size models considering setup, inventory carrying, and capacity associated costs. For example, the Economic Order Quantity model and the Lot-For-Lot model in certain situations can help level load. The Periodic Order Quantity and Least Total Cost models especially for high cost structure ratios tend to result in erratic and lumpy work center load profiles. The reasons for such operating behavior are examined. Other concepts and relationships important to capacity management are discussed.     

Resolving uncertainty in manufacturing systems

This paper examines the effectiveness of three commonly practiced methods used to resolve uncertainty in multi-stage manufacturing systems: safety stock under regenerative material requirements planning (MRP) updates, safety capacity under regenerative MRP updates, and net change MRP updates, i.e., continuous rather than regenerative (periodic) updates. The use of safety stock reflects a decision to permanently store materials and labor capacity in the form of inventory. When unexpected shortages arise between regenerative MRP updates, safety stock may be depleted but it will be replenished in subsequent periods. The second method, safety capacity, overstates the MRP capacity requirements at the individual work centers by a prescribed amount of direct labor. Safety capacity either will be allocated to unanticipated requirements which arise between MRP regenerations or will be spent as idle time. The third method, net change, offers a means of dealing with uncertainty by rescheduling instead of buffering, provided there is sufficient lead time to execute the changes in the material and capacity plans.Much of the inventory management research has addressed the use of safety stock as a buffer against uncertainty for a single product and manufacturing stage. However, there has been no work which evaluates the performance of safety stock relative to other resolution methods such as safety capacity or more frequent planning revisions. In this paper, a simulation model of a multi-stage (fabrication and assembly) process is used to characterize the behavior of the three resolution methods when errors are present in the demand and time standard estimates. Four end products are completed at an assembly center and altogether, the end products require the fabrication of twelve component parts in a job shop which contains eight work centers. In addition to the examination of the three methods under different sources and levels of uncertainty, different levels of bill of material commonality, MRP planned lead times, MRP lot sizes, equipment set-up times and priority dispatching rules are considered in the experimental design.The simulation results indicate that the choice among methods depends upon the source of uncertainty, and costs related to regular time employment, employment changes, equipment set ups and materials investment. For example, the choice between safety stock and safety capacity represents a compromise between materials investment and regular time employment costs. The net change method is not designed to deal effectively with time standard errors, although its use may be preferred over the two buffering alternatives when errors are present in the demand forecasts and when the costs of employment changes and equipment set ups are low. The simulation results also indicate that regardless of the method used, efforts to improve forecasts of demands or processing times may be justified by corresponding improvements in manufacturing performance.     

The effect of lot-sizing on workload variability

Lot-sizing models which group demand requirements for one or more consecutive time periods into a single production run have received considerable attention in recent years. Material Requirements Planning (MRP) systems must, for instance, make a lot-size decision for each planned order release. Existing decision models attempt to minimize the sum of setup plus inventory holding costs. However, lot-sizing tends to increase the work center load variability, and, consequently, the costs associated with changing production levels from period to period should be incorporated into the economic analysis. This study is concerned, first of all, with analytically describing the relationship between dynamic lot-sizing models and workload variability. Secondly, in order to account for production level change costs we propose a simple modification to existing heuristic models. Lastly, we employ a simulation model to empirically extend these results to a typical MRP multiechelon production environment. An example is included to show clearly that with cost premiums for overtime and severance or guaranteed minimum costs for undertime the traditional lot-sizing techniques significantly underestimate actual costs and can lead to very costly policies.Mean, variance and coefficient of variation of period work time requirements are derived as a function of several algorithm characteristics. Average cycle time (number of periods covered by a single batch) is found to be the most influential factor in determining workload variability. Variance grows approximately in proportion to this cycle time with the proportionality constant being the square of average period workload. Cycle time and demand variability also contribute to workload variability. Results indicate that for a given average cycle time, the EOQ method will minimize workload variability. When N products utilize the same work center, the coefficient of load variation will be reduced by a factor of N^{?$\text{1}\text{2}$} unless requirements are positively correlated. Positive correlation would result when products have similar seasons or parent items. In this case grouping such products cannot help reduce variability.In order to incorporate production level change costs into existing heuristics we may simply introduce a term consisting of a penalty factor times average cycle time. The penalty factor represents the costs of period by period production level changes. Several popular heuristics are extended in this fashion, and it is found that solutions are still readily obtainable, requiring only modifications to setup or holding cost parameters.The effects of level change costs are examined via simulation for a specific yet typical environment. It is found that when setup costs are significant, traditional lot-sizing heuristics can provide cost savings and service level improvements as compared to lot-for-lot production. However, whereas for our model the obtainable profit improvement from lot-sizing was 25% in the case of freely variable capacity, actual improvements were only one half as large when reasonable hiring and firing practices and overtime and undertime costs were considered. Consequently, management needs to consider carefully labor costs and work center product relationships when determining a production scheduling method.     

An empirical investigation of different strategies for material requirements planning

Master production scheduling, component lot sizing, and capacity requirements planning represent three important modules of material requirements planning (MRP) systems. Coordinating these three modules has been largely dependent on managerial judgment and experience. In this research, five different strategies for integrating these modules are empirically investigated. These strategies differ mainly in the extent to which the modules are coordinated. The impact of the shop, product, component and cost characteristics is measured by varying ten experimental factors. The research findings so obtained provide several guidelines on the effectiveness of each strategy in different environmental settings.     

An evaluation of overtime policies for a repair shop

This article presents the first attempt to develop and examine overtime policies for a repair shop environment. Overtime is used to augment repair capacity as needed to offset short-term demand fluctuations. If overtime provides sufficient additional repair capacity, it may be possible to reduce investment in spare parts inventory.The use of overtime in a repair shop requires managerial attention to several issues. In this article, the following five issues are examined: The relationship between overtime policies and spares stocking levels The timing of overtime—reactive or proactive The amount of overtime to use The level in the product structure at which overtime is most beneficial The priority scheduling and labor assignment policies usedSix overtime policies are developed that explore the above issues. These are examined using a simulation model of a hypothetical repair shop. Since the focus of the article is on overtime policies, a single labor assignment policy is used in conjunction with two priority scheduling rules.The results indicate that reactive overtime policies work well in this environment and overtime is most effective at the lowest level of the product structure, where repair times are relatively shorter. In addition, lowest level parts provide more usage flexibility to handle anticipated future failures.     

Integer goal programming for multistage lot sizing: Experimentation and implementation

This article considers the issues involved in implementing a large-scale multistage lot sizing model in a pharmaceutical manufacturing environment, and reports on a series of sensitivity experiments that subsequently examined the critical impacts of system capacity and inventory policy on the specification of a multistage schedule. The model was initially developed as an aggregate scheduling aid for a class of tablet pharmaceuticals. The manufacture of tablet products is a serial-type process characterized by batch flow. The model placed multiple resource capacity restrictions on various stages of the multistage system to ensure the feasibility of the resultant schedules. Difficulties in estimating penalty costs for shortages were circumvented by employing a multi-objective formulation of the lot-sizing problem.Various obstacles encountered during the implementation process are discussed. Model implementation encompassed the development of an approximation algorithm for efficiently solving the large-scale problem. The performance of the algorithm was evaluated by examining the closeness to optimality of the solutions obtained using the procedure. Performance statistics are presented for the sensitivity experiments discussed herein. Another essential aspect of implementation involved the timely revision of model input parameters. This facet of implementation proved to be at least as important to management as the efficient provision of a “good” solution.Model experimentation centered on variations in the capacities of bottleneck resources and changes in target inventory parameters whose values are predetermined by company policy. The first set of experiments was designed to demonstrate the importance of system capacity to the lot-sizing process and to illustrate that the location of a system bottleneck can vary over time as a function of both internal and external factors. The results indicated that lot-sizing procedures that fail to incorporate capacity information or that focus on a single bottleneck production stage in order to schedule production are severely limited with respect to practical and/or long-term applicability.The final set of inventory-related experiments indicated that the specification of target ending inventory levels was a crucial factor in the lot-sizing process. The determination of appropriate target levels must reflect the inherent trade-off between the objective of minimizing shortages and the desire to avoid excessive inventory accumulations. The generation of usable model information was found to be contingent upon the realistic definition of target level parameters.     

The effects of input control in a simple scheduling model

Input control is a generic procedure for smoothing production workload by delaying work during intervals of heavy load. While input control techniques have several practical benefits, they also have an inherent disadvantage. By restricting the set of jobs available for scheduling, an input control procedure removes some of the scheduling options that would otherwise be available. This paper examines the impact of such a procedure in a simple simulation model.The simulation model represents a production shop in a simplified way, as a single machine, but the production control system has three distinct parts. The first part assigns due-dates to customer orders, taking into consideration the size of each job and the workload in the shop. The second part is a job releasing rule that implements input control. The third part is a priority dispatching procedure that is aimed at meeting due-dates. By representing this three-part control system the model provides an opportunity to explore the interdependence between input control and other control procedures.Reinforcing previous research, the simulation experiments confirm that modified due-date priorities perform more effectively than other basic priority rules when performance is measured by average tardiness. Moreover, the experiments indicate that performance under the modified due-date regime is never improved by the use of input control. On the other hand, with dispatching rules that rely on shortest-first or critical ratio priorities, the experiments indicate that input control is sometimes advantageous. The effects of input control on scheduling performance thus appear to be somewhat complicated, and further experiments were designed to explore some of the relationships involved. The principal finding, however, provides a warning that input control can be counterproductive.     

Application of work center control rules in a job shop environment

Work center control rules, defined as a combination of job dispatch rules and short-term work center capacity adjustments, are analyzed using queueing theory. Promising rules are evaluated with a job shop simulation model. Simulations comparing work center control rules to the critical ratio rule for job dispatching indicate that work center control can increase performance to customer due date while simultaneously reducing average work in process inventory. The work center control rules are easily implemented by shops currently using input/output control and daily dispatch lists.     

Use of the net present value criterion in a random job shop where early shipments are forbidden

Both practitioners and researchers in the field of Operations Management have suggested that shop scheduling should be an integral component in both the strategic and tactical plans for an organization's assets. This paper examines the use of an accepted measure of return on assets, net present value (NPV), in a simulated shop scheduling environment where early shipment of jobs before their due dates is forbidden. In addition, early shipment of raw materials to the shop is also forbidden. This shop environment is consistent with the prevalent practice in industry of accepting orders only on a just-in-time basis to reduce purchased parts inventories. The NPV measure provides a means of balancing a variety of performance criteria that have been treated as separate objectives previously, including work-in-process inventory, finished goods inventory, mean flow time and mean tardiness, while also providing a means of measuring monetarily the value of various shop scheduling approaches.The NPV performance of priority scheduling rules and order release policies is measured in this research through the simulation of a random job shop under a variety of environmental conditions. It is found in a comparison of priority rules that use time-based information with those that use job value information that the Critical Ratio rule provides higher average performance than the three other rules used in the study. However, in some situations that are consistent with JIT practice, value-based priority rules also perform well. The use of a mechanism for delaying the release of jobs to each work center in the shop provided higher average NPV when shop utilization was set at a low level of 80%, while immediate release of work upon its arrival to the shop provided superior performance at a higher shop utilization level of 94%. While JIT materials delivery and costing yields higher NPV, it did not alter the relative ranking of priority rule/release policy combinations. In addition, it was found that environmental factors, including average job length, average number of tasks per job and level of tardiness penalty, resulted in greater variations in NPV performance than the institution of a JIT raw materials delivery policy.     

“牛鞭效应”下的供应链库存管理研究

为发挥库存在企业的运作过程中的作用。改善客户服务水平,在库存位置接近客户时,满足较高的客户服务要求。文章在讨论供应链库存管理中牛鞭效应的基础上,主要研究供应链库存管理办法,提出了供应商的库存管理、联合库存管理和利用第三方物流供应商来管理库存三种方法。使得适当的库存存在不仅保证销售活动的顺利进行,而且提高了实际销售量,有助于降低采购和运输成本,产生一定的经济效益。     

Ways to reduce inventory

Almost every business today is looking for ways to reduce inventories. However, the unsaid goal is to reduce that inventory while at the same time maintain or improve customer service. Today, every business operation is concerned with methods to improve customer service. The real trick is to accomplish that task without increasing inventory. This article discusses how some of the leading companies accomplish the task of reducing inventory and increasing customer service at the same time.     

供应商管理库存对顾客服务水平影响的研究

在知识经济时代,传统的库存管理思想已经不能满足市场发展的需要,强调供应链整体合作是库存理论的进一步发展趋势。而供应商管理库存是未来供应链库存管理的一种重要模式,受到越来越多理论界的重视和研究。而库存管理思想对顾客服务水平有什么影响,却很少有人对此做出模型研究。本文基于提前期对此展开讨论,提出供应商管理库存能够提高补给周期供给水平(CSL)和产品供给率(Fr),降低供应链中总成本,保证整条链快速响应顾客需求,提高顾客服务水平。     

Integrating MRP (materiel requirements planning) into modern business

Time is the commodity of the '90s. Therefore, we all must learn how to use our manufacturing systems to shorten lead time and increase customer satisfaction. The objective of this article is to discuss practical ways people integrate the techniques of materiel requirements planning (MRP) systems with just-in-time (JIT) execution systems to increase customer satisfaction. Included are examples of new ways people use MRP systems to exemplify the process of continuous improvement--multiple items on work orders, consolidated routings, flexing capacity, and other new developments. Ways that successful companies use MRP II for planning and JIT for execution are discussed. There are many examples of how to apply theory to real life situations and a discussion of techniques that work to keep companies in the mode of continuous improvement. Also included is a look at hands-on, practical methods people use to achieve lead time reduction and simplify bills of material. Total quality management concepts can be applied to the MRP process itself. This in turn helps people improve schedule adherence, which leads to customer satisfaction.     

The use of workload information to control job lateness in controlled and uncontrolled release production systems

The performance of a workload dependent scheduling and due date assignment rule is investigated. The rule uses time-phased workload information and time-phased capacity information. Performance comparisons are made with a rule that uses only time-aggregated workload information. The performance for two methods of releasing jobs to the shop is investigated: an uncontrolled method where the jobs are released at random, and a controlled method where jobs are released to maintain a specific workload norm. Computer simulation is used as the research tool. The mean and standard deviation of lateness are used as performance measures.The results indicate that the use of time-phased workload information may decrease the variance of the lateness, as compared with the use of time-aggregated workload information only. However, the magnitude of the effect depends on the type of sequencing rule used. Furthermore, if time-phased information is used, control of the mean lateness requires that the amount of capacity available for loading used in the loading procedure differs from the amount of capacity actually available for production. It is shown that by selecting adequate parameter values, both a constant mean lateness and a small variance of lateness can be obtained with this type of assignment rule.These results are valid for both types of release situations. So even in the case where the workload of the shop is under strict control, and the mean operation flow time does not, therefore, vary, the use of time-phased workload information can decrease the variance of the lateness.     

感知服务质量与顾客满意和顾客忠诚关系的实证研究

在实践中,中国服务业发展迅速,但营销水平有待提高,很多管理人员对营销以及服务管理的理解有限。在理论中,对感知服务质量各个维度、顾客满意和顾客忠诚之间的关系讨论很多,尚无定论。实证研究发现,是顾客满意正向影响感知服务质量各个因子,而不是相反,说明基于交易的顾客满意会影响对服务进行长期综合评价的感知服务质量。只有忠诚度一个因子影响顾客忠诚,因此企业在培养忠诚顾客上应该有所侧重。顾客满意和顾客忠诚的关系被验证并非线性关系,在线性方程模型中未能验证成功。     

公司库存优化方法浅析

库存是企业生产管理的重点内容之一,库存管理所要解决的核心问题是如何保证在不断提升客户服务满意度的前提下将库存成本控制在最低的范围之内。每个企业应当根据自身的现实条件和市场环境来确定最适合本企业的库存管理方案。N公司想要保持一个较低的库存水平,首先需要对库存材料进行分类．得出真正需要重点关注的物料清单,运用合适的采购订货策略．力求达到在按时满足客户需求的前提下使库存成本最低．．     

Manufacturing lead time accuracy

This article addresses the question of accuracy of planned lead times (PLTs) that are used with a material requirements planning system. Lead time error is defined as the difference between an item's PLT and the actual lead time (flow time) of an order to replenish the item. Three related topics are discussed: the relationship between system performance and average lead time error, the transient effect on work-in-process (WIP) inventory of increasing PLTs, and the relative accuracy of three methods of determining PLTs. A distinction is made between available and WIP inventory. The former includes any purchased item, fabricated part, assembly, or finished good that is in storage and available for use or delivery. WIP denotes materials associated with open orders on the shop floor.It was concluded that average lead time error has a considerable affect on system performance. PLTs that are on average too long or too short increase available inventory; and the further the average error is from zero, the more pronounced the increase. Contrary to conventional wisdom, increasing PLTs will increase the service level (decrease backorders), unless PLTs are already severely inflated and MPS uncertainty (forecast error) is small. If PLTs are inflated, decreasing them will decrease the number of setups per unit time in the case of considerable demand uncertainty. Contrary to conventional wisdom, increasing PLTs causes only a transient rise WIP inventory.The fact that the average lead time error has a significant effect on the three areas of system effectiveness mentioned above does not imply that a given order's lead time should be managed in a way that forces its actual lead time to match the PLT. Stated another way, the material planner may use the latest information to manage a given order's lead time; however, if the average discrepancy between the actual and planned lead times is large, system performance can be improved by changing the PLTs to approximate the average flow times.Three methods that have been proposed for determining PLTs are compared. They are historical averages of the actual flow times, calculated lead times based on standard times and historical averages of the queuing time at the appropriate work centers, and the QUOAT lead time proposed by Hoyt. The third was found to perform poorly unless the work content of all operations is identical. With one exception, no differences were found between the first two methods. The simpler historical average method was superior to the calculated lead time in the case where the work content of each operation varies and when considerable demand uncertainty exists.The results are based on simulation experiments employing a generalized MRP/Job-Shop stochastic simulation model. The program launches orders based on standard MRP logic, reschedules open orders by moving the due date in or out to coincide with revised need dates, moves manufacturing orders through a job shop, schedules the delivery of purchase orders, and updates inventory levels. The product structure tree contained eight distinct items, with four levels and one end item. There is no reason to believe that the conclusions would be any different had a larger system been studied.     

服务作坊及其营销策略探讨

美国学者罗杰·施米诺依据定制及其交互程度和劳动力密集程度两个维度,将服务分为四类,即服务工厂、服务作坊、大众服务和专业服务。本文阐述了其中的服务作坊及其特征,分析了其面临的营销机遇和挑战,进而提出了具体的营销策略,这对服务作坊开展营销活动具有现实指导意义。     

如何在降低库存成本的同时取得满意客户服务质量

库存水平与顾客服务质量是两个效益悖反的指标,如何能以较合理的库存水平实现较高的顾客服务质量呢? 文中通过客户的 ABC 分析、产品的 ABC 分析以及产品-客户九象限图分析,探讨了解决这个问题的方法.