An Evaluation of Group Scheduling Heuristics in a Flow-Line Manufacturing Cell

Abstract. When a company adopts cellular manufacturing and creates a ce ll,

one operational problem that must be addressed is how to schedule parts within

the cell. Many studies have investigated scheduling rules in a cellular

manufacturing environment. However, there has been little consensus on the

best sc heduling rule to use. To address this lack of consensus, this study

evaluated the best scheduling rules from most of these studies in a flow-line

cell. The impact of two environmental factors, setup to runtime ratio and

number of part families, was also investigated. Out of the five best

scheduling rules found, three of these had not been investigated in previous

group scheduling studies. The scheduling rule that most often performed best

was selecting the part family with the most waiting jobs and sequencing these

jobs in shortest processing time order, a relatively simple rule. The more

complex rules generally showed poorer performance.

Abstract. Why have results of Cellular Manufacturing (CM) implementations

been so diverse? Why after three decades do the conditions for the economic

viability of Group Technology remain unresolved? The authors interpret the above as evidence that present Cellular Manufacturing concepts and methodology are faulty due to the following:

1. The vast majority of cell formation techniques are suboptimizing CM

designs by ignoring the primary objective of CM, setup reduction, or confounding this objective with other objectives.

2. No present methods of cell formation and few scheduling techniques

consider sequential dependencies of setups and instead make general assumptions

that are in many situations invalid. The purposes of this article are to

focus attention on the primary objective of CM, offer reasons why present CM

design methods fail to consistently attain the acclaimed benefits of CM, and

offer tenets for successful CM designs.