|
|
Group Technology for Workcell Design
The New Conventional Wisdom Lean manufacturing has become so popular it is now the "conventional wisdom." The worrisome thing is that rather than using fundamental knowledge and sound analysis, many companies (and consultants) work by prescription or formula. The formula may not fit the problem.
The Toyota view of cellular manufacturing is a good example. Toyota workcells typically have high volumes and a very narrow range of products. The result is a highly efficient, U-shape with inherently balanced workstations.
So, what happens when the Toyota workcell formula does not fit, as in a job shop? Occasionally the company forges ahead, heedless of contradictions, and the results can be disastrous. More commonly, the organization simply dismisses workcells as not applicable.
The popular model limits your view of workcells. It is difficult to see how they apply in a low-volume shop. Nevertheless, the concept of cells does apply, it does work, and the relative advantages are even greater than in a high-volume shop. Manufacturers with up to 10,000 or more products can use workcells. The paradox is that the larger the number of parts, the more likely they will form viable families. Cellular manufacturing, in fact, originated in job shops, not at Toyota.
Group Technology to the Rescue
Such high-variety, low-volume situations require sophisticated analysis. The resulting workcells are quite different than Toyota's simple U-shape cells. Group technology (GT) is the key that unravels these complex product-process mixes. A subsequent article will explore the differences between Toyota (dedicated) cells and GT cells. A third article will describe a general design procedure that applies to all workcells.
GT focuses on similarities in products. It forms families of parts and machine groups to manufacture them. The machine group forms the nucleus of a workcell. Figure 1 shows how a GT brings order to a seemingly random set of machined parts.


 Figure 1 -- Before and after grouping
Intuitive Grouping
"Eyeball" and experience are the simplest, easiest and most common ways to group products. The analysts simply figure it out in their heads. However, this only works for a simple product-process mix. Figure 2 illustrates.

 Figure 2 -- Intuitive grouping
Production Flow Analysis (PFA)
A second grouping method uses current processes and routings. It employs a matrix of parts and machines. Figure 3 is an oversimplified example. In theory, PFA can deal with any size matrix, but this assumes that all similar parts have similar routings. For a variety of reasons, this rarely happens.
The PFA matrix usually has many anomalies. With smaller matrices of a few hundred parts and a dozen or so processes, the analyst recognizes and fixes the anomalies. With larger matrices, this task becomes overwhelming and the methodology breaks down.

 Figure 3 -- Production flow analysis (click on image to enlarge)
Coding and Classification
A third method for grouping is coding and classification (C&C). Here, a "code number" (separate from the part number) contains information about each part. This number identifies material, size, shape, function, process and other characteristics.
Parts with similar characteristics should have similar processes. Coding enables an analyst to standardize process routings and then group the parts into families. The example of Figure 4 shows a small part of a database for castings. Each digit position represents a particular part characteristic. For example, the third position indicates the casting's finished weight. Notice the similarity of code numbers within each group.
  Figure 4 -- Coding and classification example (click on image to enlarge)
Coding and classification is the most sophisticated and versatile of the GT analysis methods. It can cope with huge numbers of parts and processes. However, coding and classification present some major challenges that are not usually evident to an inexperienced practitioner. For example, it takes a lot of time and effort and the design of the code structure is critical. In addition, coding inconsistencies and errors must be avoided. A successful project requires experience and judgment in coding system design, initial coding and family development.
Summary
Group technology enables cellular manufacturing in job shops where it often seems impractical. The results are large decreases in inventory with increased productivity and quality. It is a cornerstone of lean manufacturing strategy for many manufacturers.
Other articles in this series examine:
How to Design Workcells
A Cellular Manufacturing Case Study
By Quarterman Lee, president of Strategos, Inc. Chapters of his book, "Facilities and Workcell Design," can be downloaded from http://www.strategosinc.com/facdes.htm.
Copyright © 2010 Society of Manufacturing Engineers
SME Customer Service: 1-800-733-4763
|
|
|
|
 |
| February 9, 2004 Issue: |
|
|