Academic journal article Journal of Management Information and Decision Sciences

# Heuristics for Scheduling Operations in MRP: Flowshop Case

Academic journal article Journal of Management Information and Decision Sciences

# Heuristics for Scheduling Operations in MRP: Flowshop Case

## Article excerpt

ABSTRACT

This paper investigates the issue of operations scheduling for Material Requirements Planning (MRP) Systems. Two general scheduling approaches, namely forward and backward procedures, are proposed to perform operations scheduling in MRP. Moreover, we develop various heuristic scheduling rules for use in the two general scheduling approaches. A simulation experiment is conducted to study the effectiveness of the proposed heuristics under the flowshop case--a manufacturing environment where all jobs processed in the shop must follow the same machine or workstation sequence. Lastly, we present and discuss the computational results obtained from our simulation experiment.

INTRODUCTION

This paper studies the issue of operations scheduling for Material Requirements Planning (MRP) systems under the general discrete manufacturing environment. The general discrete manufacturing is commonly referred to as multi-stage, production-inventory system (MS-PIS). The MS-PIS is important because it is considered as the most common type of production and manufacturing systems (Goyal & Gunasekaran, 1990). In the MS-PIS, each product or part requires one or more operations to be processed at some designated machines. The operations sequence of a product or part has to follow the specifications indicated on its routing sheet. It should be noted that different machines can be used to perform the same operation; however, they may possess different setup times and per unit processing times, as well as different per unit time costs.

MRP is a basic tool for performing the detailed material planning function in the manufacture of component parts and their assembly into finished items (Vollmann et al., 2005). The basic MRP inputs are: (1) master production schedule, (2) bill of materials, and (3) inventory records file; and the basic MRP output consists of a set of detailed schedules indicating, for each part, when and how much to order. In order to generate the output, the MRP logic takes an end-item requirements from the master production schedule and translates them into time-phased requirements for assemblies, parts, and raw materials using the bill of materials offset by lead time given by the inventory records file.

MRP has been a very powerful and pervasive tool adopted by numerous companies for over three decades; nonetheless, it possesses some major drawbacks. Two weaknesses are particularly noteworthy. The first one is the need to set planned lead time (Kanet, 1982; St. John, 1985), which is a very difficult, if not impossible, task. Plossl and Welch (1979) and Huge (1979) found that the waiting time in queue could amount to as much as 95% of the lead time. Hence, lead time is very much determined by how long it takes to obtain the required capacity, that is, the congestion level of the shop. As noted by Karmarkar (1993), from the perspective of MRP the lead time is an attribute of the part, rather than of the condition of the shop. The second noteworthy weakness is about determinating lot size quantity. According to Silver et al. (1998), most MRP systems do not provide support to any lot-sizing rules other than Economic Order Quantity (EOQ) and Lot For Lot (LFL). Even if the support is available, most users seem to rely on simple rules that may generate higher cost than is possible by using other rules.

A number of articles have appeared in the literature to discuss the possible integration of MRP and Just in Time (JIT). Nonetheless, most of them are focused on the conceptual level of the JIT philosophy. Flapper et al. (1991) propose one of the most rigorous frameworks for integrating MRP and JIT. Their three-step process makes use of MRP's backflushing and phantom features, and allows the JIT concepts to be utilized to the fullest extent. Other research in this area includes Benton and Shin (1998), Neely and Byrne (1992), Bose and Rao (1988), Belt (1987), Blackburn (1985), Cook and Muinch (1984), Discenza and McFadden (1988), and Heard (1984). …

Search by...
Show...

### Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.