Herbert Meyr

The general lotsizing and scheduling problem

The GLSP (GeneralLotsizing andSchedulingProblem) addresses the problem of integrating lotsizing and scheduling of several products on a single, capacitated machine. Continuous lotsizes, meeting deterministic, dynamic demands, are determined and scheduled with the objective of minimizing inventory holding costs and sequence-dependent setup costs. As the schedule is independent of predefined time periods, the GLSP generalizes known models using restricted time structures. Three variants of a local search algorithm, based onthreshold accepting, are presented. Computational tests show the effectiveness of these heuristic approaches and are encouraging for further extensions of the basic model.ZusammenfassungDas GLSP (GeneralLotsizing andSchedulingProblem) stellt ein neues Modell zur integrierten Losgrößen- und Reihenfolgeplanung dar. Betrachtet wird eine Maschine mit begrenzter Kapazität, für die Lose in kontinuierlicher Größe und die zugehörige Auflagereihenfolge bestimmt werden sollen. Zielkriterium sind minimale Lagerund (reihenfolgeabhängige) Rüstkosten für gegebene dynamische Bedarfe mehrerer Produkte. Die Lösungen sind unabhängig von einer vorab festgelegten Periodeneinteilung des Planungszeitraums. Das GLSP verallgemeinert daher bekannte Modelle, die eine bestimmte Zeitstruktur voraussetzen. Es werden drei unterschiedliche Local-Search-Heuristiken zur Lösung des Problems präsentiert, die auf Basis des „Threshold Accepting”-Prinzips arbeiten. Der Vergleich mit heuristischen und optimalen Lösungen für ein verwandtes Problem zeigt, daß die Ergebnisse durchaus ermutigend für Erweiterungen des Modells sind.

Simultaneous Lotsizing and Scheduling on Parallel Machines

This paper addresses the simultaneous lotsizing and scheduling of several products on non-identical parallel production lines (heterogeneous machines). The limited capacity of the production lines may be further reduced by sequence dependent setup times. Deterministic, dynamic demand of standard products has to be met without back-logging with the objective of minimizing sequence dependent setup, holding and production costs. The problem is heuristically solved by combining the local search metastrategies threshold accepting (TA) and simulated annealing (SA), respectively, with dual reoptimization. Such a solution approach has already proved to be successful for the single machine case. The solution quality and computational performance of the new heuristics are tested by means of real-world problems gathered from industry.

Simultaneous lotsizing and scheduling by combining local search with dual reoptimization

Abstract The contribution of this paper is twofold. On the one hand, the particular problem of integrating lotsizing and scheduling of several products on a single, capacitated production line is modelled and solved, taking into account sequence-dependent setup times. Thereby, continuous lotsizes, meeting deterministic dynamic demands, are to be determined and scheduled with the objective of minimizing inventory holding costs and sequence-dependent setup costs. On the other hand, a new general algorithmic approach is presented: A dual reoptimization algorithm is combined with a local search heuristic for solving a mixed integer programming problem. This idea is applied to the above lotsizing and scheduling problem by embedding a dual network flow algorithm into threshold accepting and simulated annealing, respectively. Computational tests show the effectiveness of the new solution method.

Planning Hierarchy, Modeling and Advanced Planning Systems

Publisher Summary It is important to identify different types of supply chains to derive fitting management strategies. Planning systems, which try to implement these strategies operatively, also have to be tailored to the particular requirements of the type of supply chain under consideration. The chapter introduces a typology of supply chains, which is illustrated by two contrasting examples: consumer goods manufacturing and computer assembly. A general framework for deriving the corresponding planning tasks of the respective supply chain type is provided. Planning concepts that fit the planning requirements can be designed by means of hierarchical planning (HP). The HP concepts, including recent developments, are reviewed in the chapter. It is shown that a common modular architecture, which is along the lines of HP, is underlying all the advanced planning systems (APS) and the functions of the typical modules are discussed.

Demand Fulfilment and ATP

The planning process that determines how the actual customer demand is fulfilled is called demand fulfillment. The demand fulfillment process calculates the first promise date for customer orders and — thus — strongly influences the order lead-time and the on time delivery.1 In today’s competitive markets it is important to generate fast and reliable order promises in order to retain customers and increase market share. This holds particularly true in an e-business environment: Orders are entered on-line in the e-business front end, and the customer expects to receive a reliable due date within a short time period.

Customer segmentation, allocation planning and order promising in make-to-stock production

Modern advanced planning systems offer the technical prerequisites for an allocation of “available-to-promise” (ATP) quantities—i.e. not yet reserved stock and planned production quantities—to different customer segments and for a real time promising of incoming customer orders (ATP consumption) respecting allocated quota. The basic idea of ATP allocation is to increase revenues by means of customer segmentation, as it has successfully been practiced in the airline industry. However, as far as manufacturing industries and make-to-stock production are concerned, it is unclear, whether, when, why and how much benefits actually arise. Using practical data of the lighting industry as an example, this paper reveals such potential benefits. Furthermore, it shows how the current practice of rule-based allocation and consumption can be improved by means of up-to-date demand information and changed customer segmentation. Deterministic linear programming models for ATP allocation and ATP consumption are proposed. Their application is tested in simulation runs using the lighting data. The results are compared with conventional real time order promising with(out) customer segmentation and with batch assignment of customer orders. This research shows that—also in make-to-stock manufacturing industries—customer segmentation can indeed improve profits substantially if customer heterogeneity is high enough and reliable information about ATP supply and customer demand is available. Surprisingly, the choice of an appropriate number of priority classes appears more important than the selection of the ATP consumption policy or the clustering method to be applied.

Customer Orientation in Advanced Planning Systems

Customer satisfaction is a major objective of Supply Chain Management. This also applies to “Advanced Planning Systems” (APS), computerized planning systems which try to support the various planning processes to be tackled in supply chains. “Order Promising” and “Demand Fulfillment” software modules of these systems are helpful in promising short and reliable customer order delivery dates and in tracking customer orders along the whole process between order entry and order delivery.

Revenue management and demand fulfillment: matching applications, models, and software

Recent years have seen great revenue management successes, notably in the airline, hotel, and car rental businesses. Currently, an increasing number of industries, including manufacturers and retailers, are exploring ways to adopt similar concepts. Software companies are taking an active role in promoting the broadening range of applications. Additionally technological advances, including smart shelves and radio frequency identification (RFID), are removing many of the barriers to extended revenue management. The rapid developments in supply chain planning and revenue management software solutions, scientific models, and industry applications have created a complex picture, which is not yet well understood. It is not evident which scientific models fit which industry applications and which aspects are still missing. The relation between available software solutions and applications as well as scientific models appears equally unclear. The goal of this paper is to help overcome this confusion. To this end, we structure and review three dimensions, namely applications, models, and software. Subsequently, we relate these dimensions to each other and highlight commonalities and discrepancies. This comparison also provides a basis for identifying future research needs.

Simultaneous lotsizing and scheduling problems: a classification and review of models

The current paper presents a structured overview over the literature on dynamic simultaneous lotsizing and scheduling problems. We introduce a classification scheme, review the historical development of research in this area and identify recent developments. The main contribution of the present review is the discussion of the historical development of the body of knowledge in the field of simultaneous lotsizing and scheduling and the identification of recent trends. This helps to reveal research opportunities, but it can also be helpful in the selection of appropriate models for industrial applications.

Combining the principles of variable neighborhood decomposition search and the fix&optimize heuristic to solve multi-level lot-sizing and scheduling problems

In this paper a new heuristic is proposed to solve general multi-level lot-sizing and scheduling problems. The idea is to cross-fertilize the principles of the meta-heuristic Variable Neighborhood Decomposition Search (VNDS) with those of the MIP-based Fix&Optimize heuristic. This combination will make it possible to solve the kind of problems that typically arise in the consumer goods industry due to sequence-dependent setups and shifting bottlenecks. In order to demonstrate the strength of this procedure, a GLSP variant for multiple production stages is chosen as a representative. With the help of artificial and real-world instances, the quality of the solution as well as the computational performance of the new procedure is tested and compared to a standard MIP-solver.