ArticlePDF Available

The Role of Work-In-Process Inventory in Serial Production Lines

DOI:10.1287/opre.36.2.229
Authors:
Richard Conway
Richard Conway
Richard Conway
  • This person is not on ResearchGate, or hasn't claimed this research yet.
William Maxwell
William Maxwell
William Maxwell
  • This person is not on ResearchGate, or hasn't claimed this research yet.
John O. McClain
John O. McClain
John O. McClain
  • This person is not on ResearchGate, or hasn't claimed this research yet.
L. Joseph Thomas at Cornell University
L. Joseph Thomas
Download full-text PDF
Read full-text
Download citation

Abstract

We study simple, generic production systems to gain insight into the behavior of more complex systems. Results are presented for: identical workstations with and without buffers; balanced lines in which variability of processing times differs between stations; unbalanced lines; and lines with unreliable workstations. In general, buffers between workstations increase system capacity but with sharply diminishing returns. Position as well as capacity of the buffers are important. These results are preliminary, to be confirmed and extended by further study. 22 Refs.
Content uploaded by L. Joseph Thomas
Author content
All content in this area was uploaded by L. Joseph Thomas on Oct 25, 2016
Content may be subject to copyright.
This article was downloaded by: [128.253.95.89] On: 18 August 2015, At: 06:49
Publisher: Institute for Operations Research and the Management Sciences (INFORMS)
INFORMS is located in Maryland, USA
Operations Research
Publication details, including instructions for authors and subscription information:
http://pubsonline.informs.org
The Role of Work-in-Process Inventory in Serial Production
Lines
Richard Conway, William Maxwell, John O. McClain, L. Joseph Thomas,
To cite this article:
Richard Conway, William Maxwell, John O. McClain, L. Joseph Thomas, (1988) The Role of Work-in-Process Inventory in Serial
Production Lines. Operations Research 36(2):229-241. http://dx.doi.org/10.1287/opre.36.2.229
Full terms and conditions of use: http://pubsonline.informs.org/page/terms-and-conditions
This article may be used only for the purposes of research, teaching, and/or private study. Commercial use
or systematic downloading (by robots or other automatic processes) is prohibited without explicit Publisher
approval, unless otherwise noted. For more information, contact permissions@informs.org.
The Publisher does not warrant or guarantee the article’s accuracy, completeness, merchantability, fitness
for a particular purpose, or non-infringement. Descriptions of, or references to, products or publications, or
inclusion of an advertisement in this article, neither constitutes nor implies a guarantee, endorsement, or
support of claims made of that product, publication, or service.
© 1988 INFORMS
Please scroll down for article—it is on subsequent pages
INFORMS is the largest professional society in the world for professionals in the fields of operations research, management
science, and analytics.
For more information on INFORMS, its publications, membership, or meetings visit http://www.informs.org
Downloaded from informs.org by [128.253.95.89] on 18 August 2015, at 06:49 . For personal use only, all rights reserved.
Downloaded from informs.org by [128.253.95.89] on 18 August 2015, at 06:49 . For personal use only, all rights reserved.
Downloaded from informs.org by [128.253.95.89] on 18 August 2015, at 06:49 . For personal use only, all rights reserved.
Downloaded from informs.org by [128.253.95.89] on 18 August 2015, at 06:49 . For personal use only, all rights reserved.
Downloaded from informs.org by [128.253.95.89] on 18 August 2015, at 06:49 . For personal use only, all rights reserved.
Downloaded from informs.org by [128.253.95.89] on 18 August 2015, at 06:49 . For personal use only, all rights reserved.
Downloaded from informs.org by [128.253.95.89] on 18 August 2015, at 06:49 . For personal use only, all rights reserved.
Downloaded from informs.org by [128.253.95.89] on 18 August 2015, at 06:49 . For personal use only, all rights reserved.
Downloaded from informs.org by [128.253.95.89] on 18 August 2015, at 06:49 . For personal use only, all rights reserved.
Downloaded from informs.org by [128.253.95.89] on 18 August 2015, at 06:49 . For personal use only, all rights reserved.
Downloaded from informs.org by [128.253.95.89] on 18 August 2015, at 06:49 . For personal use only, all rights reserved.
Downloaded from informs.org by [128.253.95.89] on 18 August 2015, at 06:49 . For personal use only, all rights reserved.
Downloaded from informs.org by [128.253.95.89] on 18 August 2015, at 06:49 . For personal use only, all rights reserved.
... Reference [33] also connects buffer allocation with downtime. It shows that onedowntime buffering is sufficient to regain about 50% of production losses if the downtime is constant (deterministic). ...
... Thus, for in-process buffers, this research follows references [31] and [33] and provides additional results on rules-of-thumb for buffer capacity allocation necessary to accommodate downtime and achieve the desired production rate of serial production lines. One of the main purposes of this study is to generalize the results using several different machine models, such as those based on the exponential, gamma, log-normal, and ...
Lean Buffering in Production Systems: A Quantitative Approach
Article
Full-text available
  • Jan 2003
Dissertation (Ph.D.)--University of Michigan.
View
... By adding an additional die at four stations, we find that, without any initial inventory, the throughput of options X1234, X1245, and X2345 is essentially equivalent. When combined with initial inventory, however, option X2345 is far superior, indicating that position as well as buffer size are important (Conway et al. 1988). By placing the bottleneck at the beginning of the line, the extra inventory at the other stations allows them to pass on rolls larger than six for a number of rounds (Gupta and Boyd 2011). ...
... The graphs for the other capacity options follow a similar pattern. This coincides with the findings in Conway et al. (1988) and Martin (2007) that a small amount of WIP can produce significant benefits, but the returns eventually diminish sharply. ...
Playing Goldratt’s Dice Game in Large Classes
Article
Full-text available
  • May 2022
We provide and describe an Excel-based simulation of a classic dice game. Instructors can use the simulation in large classes to allow student teams to try to select the best combination of capacity and inventory-improvement options to maximize profit. Instructors generate the results during class in real time. The game also provides excellent debriefing opportunities for further insight. In this paper, we not only present the game itself, but we perform extensive numerical analysis to develop deeper insight regarding the effect of the different options. This analysis provides instructors with tools to modify our original costs to produce different winning combinations and to potentially play the game multiple times to strengthen the managerial insights derived in this illustration of addressing the problem of dependent events and statistical fluctuations in assembly lines. Instructors of analytics or modeling courses could even ask their students to design the basic Monte Carlo simulation model themselves. Survey evidence has shown enthusiastic student endorsement of the game, and pretest/posttest analysis suggests strong learning effects.
View
... Work in process (WIP) is a company's product that is in the production process, and still has not yet been produced and has not been sent to the storage area for the finished product. According to Conway, Maxwell, McClain, & Thomas (1988) [11] WIP is a company's product that has gone thoroughly the initial process and has not been processed. If there are more WIPs, the impacts that can be occurred are as follows: 1. ...
View
... These results highlight the influence of the buffer size, as confirmed by multiple authors such as [28]. In the numerical experiments presented, a simple line composed by only three workstations have been considered; for sure production efficiency differences between "G" and "B" scenarios would increase with the length of the line. ...
A Stochastic-Based Model to Assess the Variability of Task Completion Times of Differently Aged and Experienced Workers Subject to Fatigue
Conference Paper
Full-text available
  • Sep 2023
In current industrial scenarios, the new paradigm of Industry 5.0 (I5.0) is gaining interest: considering the Industry 4.0 paradigm as a base, I5.0 is aimed at reaching a more sustainable, human-centric, and resilient industry. Under the I5.0 human-centric perspective, behavioural issues assume high criticality, thus requiring a more reliable prediction of operators' performances. In manual assembly lines, operators' performances are characterized by a stochastic behaviour over time. System's and operator's features cause the variability of task completion time: the former is related to properties of the work environment (e.g., ergonomics, cycle time), the latter is related to the intrinsic stochastic behaviour of operators. Furthermore, workers' features and their different attitudes to becoming fatigued, influence performance variability. In this context, the authors propose a new stochastic model that expresses the variability of execution times of operators involved in manual assembly lines by considering their differences in age, experience, and fatigue state. The novelty of the proposed model relies on considering the stochastic behaviour of workers influenced by age, experience as well as fatigue. The effectiveness of the proposed model is tested through numerical experiments of a job rotation scheduling problem to maximize productivity with proper worker-workstation assignments.
View
... In general, as CV increases, the process performance for production gets worse. Conway et al. (1988) investigated the effect of WIP inventories in flowshop production [14]. Through simulation, they concluded that the larger the CV, the lower the capacity of a production line, and WIP inventories were important to recover the capacity. ...
Stability of trade-off balancing in one-stage production scheduling
Article
  • Sep 2022
Production scheduling faces three challenges, which are inconsistent key performance indicators (KPIs), processing time uncertainties, and production schemes. Applying modern portfolio theory (MPT), Li et al. (2021) proposed a ToB(α) heuristic to balance trade-offs in one-stage production. However, production schemes for optimizing average performance of individual KPIs, trade-off values, or worst-case scenarios affect the stability of a process differently, especially with processing time uncertainties. We propose an innovative approach using transfer functions for stability (TF4S) in balancing trade-offs in production scheduling. Our TF4S approach provides a systematic way to analyze the stability of one-stage production and can be extended to production scheduling for classic m-machine flow lines.
View
... Previous research on stochastic flow lines suggested that buffers should be installed at the center or near the center of balanced flow lines (see, e.g., Hillier & Boling, 1967;Hillier & Boling, 1979or Conway, Maxwell, McClain, & Thomas, 1988. Table 11 lists different cases for the allocation of spare parts in a balanced flow line. ...
Evaluation of Unreliable Flow Lines with Limited Buffer Capacities and Spare Part Provisioning
Article
We consider a model of an N-stage flow line with stochastic processing times and interstage buffers that decouple adjacent production stages. Machine downtimes are induced by failures of critical machine components. Each machine is assumed to have exactly one of these failure-prone components. To achieve high machine availability, it is assumed that spare parts for those failure-prone critical components are kept in stock. Failed components are immediately replaced with new, functioning components, and a one-for-one replenishment policy is applied for the restocking of those spare parts. We present a novel decomposition approach to approximate the average throughput and inventory for a system with an arbitrary number of machines, buffers, and spare parts. With a detailed numerical study, we analyze the impact of different parameter constellations on the approximation quality. We demonstrate the remarkable accuracy of our method by comparing our results with both exact and simulated values. Using our method, we further study the complex interaction and partial substitution effects between buffer sizes and spare part base-stock levels on the logistical performance of the flow line.
View
... Because each workstation in the production line has a different cycle time, the scheduler must consider whether a push production environment can facilitate a smooth flow between workstations and a similar production capacity across all the workstations. When either or both of these conditions cannot be ensured, unbalanced production lines and work-in-process (WIP) accumulation are likely to result [2,3]. The kanban production system adopted by lean production advocates is based on pull demand such that it is also referred to as a pull production system [4]. ...
A Comparative Study of Unbalanced Production Lines Using Simulation Modeling: A Case Study for Solar Silicon Manufacturing
Article
Full-text available
  • Jan 2022
In the solar silicon manufacturing industry, the production time for crystal growth is ten times longer than at other workstations. The pre-processing time at the ingot-cutting station causes work-in-process (WIP) accumulation and an excessively long cycle time. This study aimed to find the most effective production system for reducing WIP accumulation and shortening the cycle time. The proposed approach considered pull production systems, and the response surface methodology was adopted for performance optimization. A simulation-based optimization technique was used for determining the optimal pull production system. The comparison between the results of various simulated pull production systems and those of the existing solar silicon manufacturing system showed that a hybrid production system in which a kanban station was installed before the bottleneck station with a CONWIP system incorporated for the rest of the production line could reduce the WIP volume by 26% and shorten the cycle time by 16% under the same throughput conditions.
View
Flexible assembly layouts in smart manufacturing: An impact assessment for the automotive industry
Article
  • Sep 2022
Currently, automotive manufacturers take the concept of flexible manufacturing to an unprecedented level, considering the deployment of flexible assembly layouts (also known as matrix production systems) in which automated guided vehicles transport bodyworks on individual routes between assembly stations. To this end, a methodological framework that allows to assess the impact of technology choice decisions between traditional assembly lines and flexible assembly layouts as well as the impact of different flexibility levers and configurations is necessary for optimal decision support. We provide such a framework based on analytical insights and a chance-constrained problem formulation. We further show how this problem formulation can be solved optimally using a tailored branch-and-price algorithm. Our results quantify the impact of different flexibility configurations in flexible assembly layouts. We show that flexibility enables a simultaneous improvement in worker utilization and work in progress, resolving a classical trade-off in manufacturing systems. Moreover, we find that worker utilization and output are up to 30% higher in flexible assembly layouts compared to line assembly layouts. Further, flexible assembly layouts prove to be especially beneficial during the ramp-up of vehicles with alternative drivetrain technologies, such as the current transition to electric vehicles.
View
Some Lean Manufacturing Tools
Chapter
  • Jan 2022
Lean manufacturing is a methodology that focuses on tools and practices implementation in production lines, which provides improvements and better performance in all operative processes. These tools optimize production processes to achieve cost reduction by reducing waste generated throughout the production process. This chapter gives an overview of 10 lean manufacturing tools used to optimize the manufacturing process, indicating the main benefits, enablers, and barriers.KeywordsLM descriptionLM tools
View
Dynamic line balancing in unpaced mixed-model assembly lines: A problem classification
Article
  • Feb 2022
Mixed-model assembly lines are almost never perfectly balanced in industrial practice, nor can sequencing fully compensate imbalances. A promising approach to achieve both equal loads for operators and high throughput for manufacturers is dynamic line balancing. Besides the tasks fixed allocated to stations, some tasks can be shared between adjacent stations. Decisions are made on the tactical design regarding the identification of shared tasks and during the operation assigning shared tasks to stations for each job. This work provides a classification of the respective tactical and operational decision problems, classifies existing literature accordingly and derives an agenda for future research.
View
The Best Order for Queues in Series
Article
Full-text available
  • Apr 1985
An important design problem for queueing systems is to determine the best order for two or more service stations in series. For given external arrival process and given service-time distributions, the object is to determine the order of the stations (to be used by all customers) that minimizes the expected equilibrium sojourn time per customer. Unfortunately, very little is known about this problem because exact analysis is extremely difficult. This paper applies approximation methods recently developed for networks of queues to obtain approximation formulas and useful heuristic design principles.
View
Finite Queues in Series with Exponential or Erlang Service Times—A Numerical Approach
Article
  • Apr 1967
This paper considers a queuing system consisting of N service channels in series where each channel has an exponential or Erlang holding time and except for the first channel a finite queue, and where the input process is such that the first queue is never empty. The measures considered are the steady-state mean output rate and mean number of customers in the system excluding the first queue. First, a procedure is described, for obtaining these measures, that is relatively efficient computationally. Second, an exceptionally efficient procedure is developed for approximating the mean output rate for the case of exponential holding times. It is demonstrated that this procedure provides an excellent approximation for most cases and that it is computationally feasible for large problems. Third, extensive new numerical results are obtained.
View
A viable alternative to the ‘method of stages’ solution of series production systems with Erlang service times
Article
  • Oct 1976
The mean production rate of a series production system with identical Erlang stations is obtained by solving an integral equation formulated by Muth. In this method, one needs to solvE only k simultaneous equations rathor than the k(k + 1) (k + 3) equations generated by the ‘ method of stages ’ approach. As such, it is much more feasible computationally for realistic systems with large Erlang parameters (k). Results are reported for values of k up to 25.
View
Queueing Networks: A Survey of Their Random Processes
Article
  • Sep 1985
We review three topics in queueing network theory: queue length processes, sojourn times, and flow processes. In the discussion of the queue length processes we present results for the continuous-time process and several embedded processes. Then we compare continuous-time processes with embedded processes. In considerable generality we present results for mean sojourn times and discuss the distributions of sojourn times. In the discussion of flow processes we present results for various queueing systems.
View
The location of variable work stations and the performance of production flow lines
Article
  • Oct 1976
The effect of imbalance in the variability of station work-time distributions is considered. The effect of buffer capacity, service-time variability and the order of constant and variable work stations on output efficiency and average idle time is examined.
View
Sequential Arrays of Waiting Lines
Article
  • Dec 1956
The problem considered in this paper is that in which a sequence of service operations must be performed on the units to be serviced. Poisson arrivals and exponential service times are assumed. Four particular cases of service facilities in series are treated, involving infinite storage space between stages, no storage space between stages, finite storage space between stages, and the case of the unpaced belt-production line. A comparison is given for two stages, and the maximum possible utilization is obtained for all cases discussed.
View
The interstage buffer storage capacity of non-powered assembly lines A simple mathematical approach
Article
This paper considers the properties of the interstage buffer storage capacity as a ‘ decoupling agent ’ in multi-station non-powered assembly lines, where the economics of production are achieved by balancing delay costs against stock-carrying costs and storage space costs. A simplified mathematical model for exponential service time distribution based on the ‘ cyclic queue ’ concept has been developed, taking into account the three cost parameters, to optimize the buffer storage capacity. A numerical example and sensitivity analysis of the model are presented.
View
Analysis of in-process buffers for multi-stage transfer line systems
Article
The paper deals with the role of buffer stock in multi-stage transfer lines by presenting computer simulation results. It is shown that an n-stage line should be designed such that the lowest stage production rate occurs in the nth stage, the second lowest in the first stage, the third lowest in the (n - 1 )st stage, the fourth lowest in the second stage and so on, to maximize line production rate. The paper also demonstrates that the total buffer capacity should be allocated such that (a) the difference between the production rates of the stages on either side of a storage point is minimized, and (b) the production rate of the stages before the storage point is slightly greater than that of the stages following the storage point
View
Properties of duality in tandem queueing systems
Article
  • Dec 1975
Duality in the queueing theory has been discussed by Finch [2], Foster [3], Ghosal [4] and others in connection with storage problems since the late 1950's. For example, Foster considered in his descriptive paper queues with maximum of N waiting-spaces and derived duality principles to relate the number of customers to the number of waiting spaces, since the number of waiting spaces decreased by one as the number of customers increased by one. In relationship between the number of customers and the number of waiting spaces, the arrival and service processes are simply reversed. This is the reason why the dual system in a single server queueing system has commonly been defined by interchanging the arrival process and the service process in the primary system. On the other hand, duality in tandem queueing systems has been applied only to the following special cases. Gordon and Newell [5] considered cyclic queueing systems where N customers advanced sequentially in clockwise. They studied the relationship between a customer and a waiting space as follows: when a customer completed his service at the ith stage and advanced to the (i+l)th stage, waiting spaces at the (i+l)th stage were decreased by one and those at the ith stage were increased by one just like one of waiting spaces at the (i+l)th stage completed her 'service', that is, waiting spaces advanced sequentially in counterclockwise. Therefore they defined the dual system as reversing the order of service in the primary system. Makino [8] analyzed the reversibility for some two and three stage tandem queueing systems where he compared the ordinary system with the reversed order-of-service system. The above authors treated negative exponential service processes and have applied the standard method of the queueing analysis to these problems. When all service time distributions are negative exponential, the process of queue lengths becomes Markovian so the balance equation method can be applied though there still remains insurmountable computational problems. However, when service processes are more complicated and realistic, an alterna
View
An Efficient Decomposition Method for the Approximate Evaluation of Tandem Queues with Finite Storage Space and Blocking
Article
  • Apr 1987
This paper presents an efficient method for evaluating performance measures for a class of tandem queueing systems with finite buffers in which blocking and starvation are important. These systems are difficult to evaluate because they have large state spaces and because they cannot be decomposed exactly. The approximate decomposition approach we describe is based on system characteristics such as conservation of flow. Comparisons with exact and simulation results indicate that the approach is very accurate.
View