Rule-based vs. optimisation-based order release in workload control: A simulation study of a MTO manufacturer

https://doi.org/10.1016/j.ijpe.2011.09.012Get rights and content

Abstract

The paper presents a simulation study that compares multi-period models for order release optimisation (Input/Output Control with fixed lead times and a clearing function model) with a traditional workload control-based order release mechanism (control of aggregate workload with periodic release). For the optimisation models the study assumes periodic re-planning and thus can also assess the effects of demand predictability. The simulation model is based on a practical case of a manufacturer of optical storage media. The results indicate that optimisation models for order release planning largely outperform traditional workload control-based order release mechanisms even in the case of poor demand forecasts.

Section snippets

Problem description

Short, predictable flow times are an important goal in manufacturing planning and control (MPC), especially in environments where high flexibility in meeting customer demand is required. To attain this goal, both the manufacturing system and the manufacturing planning and control system must be designed appropriately. This paper concentrates on the contribution of manufacturing planning and control systems to keeping flow times at a pre-determined (low) level while at the same time maintaining

Short-term order release mechanisms versus multi-period order release planning: literature review

The order release approaches that are compared in this paper developed from two research traditions that have been pursued rather independently since the 1970s and early 1980s, starting from the same underlying idea.

The WLC concept and the importance of the order release function for the control of WIP and flow times was described in conceptual publications mainly in the 1980s (Bertrand and Wortmann, 1981, Bertrand et al., 1990, Zäpfel and Missbauer, 1993a). WLC aims at maintaining a

The order release methods

The hypotheses are tested using three order release methods: The rule-based approach is represented by a specified rule-based order released mechanism (abbreviated tradRM). The optimisation-based approach is represented by two models: an Input/Output Control model with fixed lead times (abbreviated IOC) and a clearing function model that can adjust the lead times to the load situation in the shop (abbreviated CF). We now describe these release methods.

Simulation model

In this section we describe the simulation model and structure it into its relevant sub-models: production system model, demand generation model and demand forecast model. For more details on the simulation setting, see Pürgstaller (2009).

Experimental setting

In the simulation study, 6 experimental factors were considered. These factors and their treatments are summarised in Table 5.

The parameter setting of the order release methods was determined using a procedure comparable to that described by Land (2004). Therefore initial runs with the traditional release mechanism at infinite and bounded WIP norms were performed for the investigated combinations of the total demand and product mix variability. Afterwards the WIP norms of the traditional

Results

The release algorithms tradRM (Section 3.1) and IOC (Section 3.2) are compared using settings for the load limit or for the planned lead times, respectively, that lead to (nearly) identical values for the total inventory (number of orders in the shop and in the final product inventory). The proportion of orders in time and the average tardiness of all orders (which is set to zero for orders in time; see Pinedo 1995, p. 12) are the relevant performance indicators. The CF model led to

Conclusions and research topics

We can conclude that order release planning using the Input/Output Control model largely outperforms the traditional release mechanism even if demand predictability is low, except in the case of largely constant demand and product mix. Eliminating the fixed lead time constraint as in the clearing function model potentially should further improve the results, although there are still some technical difficulties with this type of model. The results indicate that in a number of practical cases the

References (62)

  • I.P. Tatsiopoulos et al.

    Lead time management

    European Journal of Operational Research

    (1983)
  • G. Zäpfel et al.

    New concepts for production planning and control

    European Journal of Operational Research

    (1993)
  • G. Zäpfel et al.

    Production planning and control (PPC) systems including load-oriented order release—problems and research perspectives

    International Journal of Production Economics

    (1993)
  • S.C. Aggarwal

    MRP, JIT, OPT, FMS? Making sense of production operations systems

    Harvard Business Review

    (1985)
  • J.M. Asmundsson et al.

    Tractable nonlinear production planning models for semiconductor wafer fabrication facilities

    IEEE Transactions on Semiconductor Manufacturing

    (2006)
  • J.M. Asmundsson et al.

    Production planning with resources subject to congestion

    Naval Research Logistics

    (2009)
  • K.R. Baker et al.

    An analytical framework for evaluating rolling schedules

    Management Science

    (1979)
  • Bechte, W. 1980. Steuerung der Durchlaufzeit durch belastungsorientierte Auftragsfreigabe bei Werkstattfertigung....
  • B. Belt

    Integrating capacity planning and capacity control

    Production and Inventory Management

    (1976)
  • D. Bergamaschi et al.

    Order review and release strategies in a job shop environment: a review and a classification

    International Journal of Production Research

    (1997)
  • J.W.M. Bertrand et al.

    Production Control and Information Systems for Component-Manufacturing Shops

    (1981)
  • J.W.M. Bertrand et al.

    Production Control: A Structural and Design Oriented Approach

    (1990)
  • S. Bhaskaran

    Simulation analysis of a manufacturing supply chain

    Decision Sciences

    (1998)
  • M. Dawande et al.

    Forecast horizons for a class of dynamic lot size problems under discrete future demand

    Operations Research

    (2007)
  • A.G. de Kok et al.

    Planning supply chain operations: definition and comparison of planning concepts

  • A. De Toni et al.

    Production management techniques: push–pull classification and application conditions

    International Journal of Operations and Production Management

    (1988)
  • Flexsim Software Products, Inc., 〈http://www.flexsim.com/〉 (retrieved...
  • H. Grünwald et al.

    A framework for quantitative comparison of production control concepts

    International Journal of Production Research

    (1989)
  • S.T. Hackman et al.

    A general framework for modeling production

    Management Science

    (1989)
  • L.C. Hendry et al.

    Job release: part of a hierarchical system to manage manufacturing lead times in make-to-order companies

    The Journal of the Operational Research Society

    (1991)
  • C.C. Holt et al.

    Planning Production, Inventories and Work Force

    (1960)

    • Cited by (32)

    • Integrated and hierarchical systems for coordinating order acceptance and release planning

      2022, European Journal of Operational Research
      Citation Excerpt :

      The most mature modeling approach for workload-dependent lead times is the use of clearing functions (CFs), which represent a functional relationship, or metamodel, relating the amount of work available to a production resource in a planning period and its expected output in that period (Graves, 1986; Haeussler & Missbauer, 2014; Karmarkar, 1989; Missbauer, 2011; Missbauer & Uzsoy, 2011). The term refers to the fraction of the current workload that can be processed to completion (“cleared”) by a resource in a planning period (Asmundsson, Rardin, & Uzsoy, 2006), allowing planned WIP and output to be adjusted to the load situation in the shop (Pürgstaller & Missbauer, 2012). Rather than being a relationship that is valid for each and every observation, a CF describes an average relation over a wide range of operating states (Asmundsson et al., 2006; Kacar & Uzsoy, 2010).

    • Average reward adjusted deep reinforcement learning for order release planning in manufacturing

      2022, Knowledge-Based Systems

    • Towards an Automated Application for Order Release

      2022, Procedia CIRP

    • Comparison of two optimization based order release models with fixed and variable lead times.

      2020, International Journal of Production Economics
      Citation Excerpt :

      In contrast, optimization models in category (2) optimize the timing of production within the lead time and thus fully utilize the buffering capabilities of WIP to achieve production smoothing. These models use the logic of input output control (Wight, 1970; Plossl, Wight, 1973; Belt, 1976) that can be integrated into order release planning models (see de Kok, Fransoo, 2003; Spitter et al., 2005; Puergstaller, Missbauer, 2012). A major disadvantage of models with fixed, exogenous lead times is their inability to consider the nonlinear relationship between resource utilization and lead times, which is well-known from practice and queueing theory.

    • Order release planning by iterative simulation and linear programming: Theoretical foundation and analysis of its shortcomings

      2020, European Journal of Operational Research

    • Order release in ETO systems - A basic optimization-based approach

      2019, IFAC-PapersOnLine
    View all citing articles on Scopus
    View full text
    Copyright © 2011 Elsevier B.V. All rights reserved.