SRC Task 1 Meeting 1227

1
SRC Task 1 Meeting12/27

• Vic Chen

2
Efficiency Measurement of Alternative Machine
Component Grouping Solutions via Data Envelopment
Analysis

• Scott M. Shafer and John W. Bradford
• Department of Management, Auburn University
• Department of Management, University of Miami
• IEEE Transactions on Engineering Management
  1995.

3
Guideline

• Objective In cellular manufacturing (CM)
  system, find the efficient machine grouping
  solutions.
• Problem 1 What is an appropriate number of
  clusters?
• Problem 2 How to dispatch the operators?

4
Problem Descriptions

• The company has 19 machines (most of CNC type)
  and routes for the 150 highest usage parts.
• Use the cell algorithm developed by Vakharia and
  Wemmerlov with 310 clusters.

5
Result of Cell Algorithm

• More clusters get more machines.
• Contained more than 50 machines is treated as a
  remainder cell.

6
Operators Dispatching Rule

• One worker is available for each machine in the
  remainder cell for the long distance or complex
  process.
• At one extreme, one worker is available for each
  machine.
• The other extreme, one worker is available for
  each machine cell.
• Others.

7
Result of Simulations

• A total of 47 simulation runs were run for two
  years beyond an initial two-week startup period.

8
Applied DEA Model

• A DMU (decision-making unit) is a combination of
  machine groups and workers.
• Performance index
• A scale transformation is employed by subtracting
  the original value from the maximum value .

9
Results of DEA

• DMU 1, 9, 10, 13 are efficient.

10
Discussions of Results

• Compute the costs of four efficient combinations,
  then choose the cheapest one.
• We can provide flexibility to inputs and outputs
  that are relevant to a particular plants
  objectives. Like the following constraint in
  primal model.

11
Excepted Efficient Results
The number must be an integer.

• The adjusted data suggested may not be possible.