Part Family and Manufacturing

Cell Formation
Cellular Manufacturing • "Given the machine-part matrix, form
groups of machines and part families in
such a way y that p
part families can be
Chapter 18, Groover. fully processed in a group of machines"
(Ham et al., 1985).
• An entry in a machine-part matrix is 1 to
indicate that a machine is used to
process a part, and 0 otherwise.
19

Part Family & Manufacturing Cell Example: Part-Machine Incidence matrix

Formation: General Procedure PART PART
1. Define the scope of the study including system 5 7 2 4 6 1 8 3 5 4 6 7 3 8 2 1
boundaries exceptions
2. Identify the similarity attributes of interest C 1 1 1 1 D 1 1 1 1
3. Simplify E 1 1 C 1 1 1 1

MAC
MAC

• Group obviously similar parts into a representative part-type

A 1 1 1 E 1 1

CHINE
CHINE

• Group pieces of equipment that must stay together into a

representative machine-type
F 1 1 F 1 1
4. Find process plans using part-types and machine-
types found in the previous step D 1 1 1 1 A 1 1 1
clusters
5. Determine the Part-Machine Incidence Matrix based B 1 1 1 B 1 1 1
on the process plans found in the previous step
6. Find the best Product Families and Machine Cells The aim is to form ‘natural’ machine-part clusters; i.e., form groups of
using clustering methods machines and part families in such a way that part families can be fully
processed in a group of machines
20 21

Rank Ordering Clustering Method

Analytical Methods
• Rank Ordering Clustering Technique
– facilitates the identification of part-machine
1. Clustering
clusters on the part-machine incidence
• Rank Order Method matrix
• Covered in textbook
• Th
The rows andd columns
l off 0’
0’s and
d 1’
1’s
2. Manufacturing Cell Layout are considered binary numbers.
• Hollier Methods 1 and 2 • The method repeatedly sorts rows and
• Method 2 covered in textbook columns until no change occurs.

22 23

1
Example: Apply the Rank Ordering Clustering Method to the previous example
C
5
1
7
0
2
0
4
1
6
1
1
0
8
0
3
1
binary
10011001
Handling Exceptions Form
E
A
0
0
1
0
0
1
0
0
0
0
0
1
1
0
0
1
01000010
00100101 • Exceptions are parts that cannot be groups of
F 0 1 0 0 0 0 1 0 01000010
machines and
D
B
1
0
1
0
0
1
1
0
1
0
0
1
0
0
0
1
11011000
00100101
completed in one cell. part families
in such a way that
After sorting ROWS
5 7 2 4 6 1 8 3 binary
• In order to handle exceptions part families can
D
C
1
1
1
0
0
0
1
1
1
1
0
0
0
0
0
1
11011000
10011001 – Incur inter-cellular material handling be fully processed
E 0 1 0 0 0 0 1 0 01000010 Answer: in a group of
F
A
0
0
1
0
0
1
0
0
0
0
0
1
1
0
0
1
01000010
00100101
Machine clusters: – Modify process plan machines
B 0 0 1 0 0 1 0 1 00100101 {A,B}, {C,D}, and
Bin 110000 101100 000011 110000 110000 000011 001100 010011
{E F}
{E,F} – Duplicate machines
After Sorting COLUMNS
5 4 6 7 3 8 2 1 binary Part families: Example: Eliminate the exceptions in the example by duplicating machines.
D 1 1 1 1 0 0 0 0 11110000 {1,2,3}, {4,5,6}, {7,8}
C
E
1
0
1
0
1
0
0
1
1
0
0
1
0
0
0
0
11101000
00010100 Exceptions: Add a machine D2 to Cell {EF}, and add a machine C2 to Cell {A,B}.
F
A
0
0
0
0
0
0
1
0
0
1
1
0
0
1
0
1
00010100
00001011
{3,7} The resulting cells are: {A,B,C2}/{1,2,3}, {D2,E,F}/{7,8}, and {D,C}/{4,5,6}.
B 0 0 0 0 1 0 1 1 00001011 5 4 6 7 3 8 2 1
Bin 110000 110000 110000 101100 010011 001100 000011 000011
D 1 1 1 0 0 0 0 0
After sorting ROWS exceptions C 1 1 1 0 0 0 0 0
5 4 6 7 3 8 2 1 binary
D 1 1 1 1 0 0 0 0 11110000 E 0 0 0 1 0 1 0 0
C 1 1 1 0 1 0 0 0 11101000 F 0 0 0 1 0 1 0 0
E 0 0 0 1 0 1 0 0 00010100 Stop – no more D2 0 0 0 1 0 0 0 0
F 0 0 0 1 0 1 0 0 00010100
A 0 0 0 0 1 0 1 1 00001011 changes. A 0 0 0 0 1 0 1 1
B 0 0 0 0 1 0 1 1 00001011
Bin 110000 110000 110000 101100 010011 001100 000011 000011 24 B 0 0 0 0 1 0 1 1 25
C2 0 0 0 0 1 0 0 0
clusters

(Adapted from 15.8 & 15.11) The following table lists the weekly quantities and routings of ten parts that
are being considered for cellular manufacturing in a machine shop. Parts are identified by letters and
machines are identified numerically. After applying the Rank-ordering clustering method, two machine

Manufacturing Cell Layout clusters are formed: {2, 3, 4, 7}/{A, E, G, I, H} and {1, 5, 6}/{B, C, D, F, J}. (a) Determine the part
incidence matrix for the first cluster. (b) Determine the flow diagram, identifying the flows ‘in’ and ‘out’
the cell. (c) Recommend a cell machine layout using Hollier Method 1.
Part Weekly quantity Machine routing Part Weekly quantity Machine routing
• Once machine clusters are identified, one A 50 3→2→7 F 60 5→1
B 20 6→1 G 5 3 →2 → 4
needs to decide the best machine layout to C 75 6→5 H 100 3 →2 → 4 → 7
implement. D 10 6→5→1 I 40 2→4→7
E 12 3→2→7→4 J 15 5→6→1
• For a good cell layout one must consider
– Technological considerations (a) The From-To diagram
to
(b) The flow diagram

– Streamlined material flow 2

2
-
3 7
62
4
145
62

• Hollier methods specify the machine layout 3 167 -

from

7 - 12 167 167 145 140 190

3 2 4 7
that maximizes the proportion of in-sequence 4 140 -
12

moves within the cell. 40 17

26 28

Solve the previous problem using Hollier Method 2:

Hollier Method 2 2 3
Step 1
7 4 From From
Step 2
To From/To Order
sums sums ratio
1. Develop the From-To Chart 2 - 62 145 207 2 207 167 1.24 2
3 167 - 167 3 167 0 ∞ 1
2. Determine the From/To ratio for each 7 - 12 12 7 12 202 0.06 4
4 140 - 140 4 140 140 0.89 3
machine by dividing the ‘From-sum’ by the To 167 0 202 157
‘To-sum’ for each machine
Sequence: 3 → 2 → 4 → 7
3 Arrange
3. A the
h machines
hi i order
in d off d
decreasing
i
From/To ratios 1. Develop the From-To Chart
• Machines with high ratios are placed at the 2. Determine the From/To ratio for each machine by dividing the ‘From-
beginning of the flow. sum’ by the ‘To-sum’ for each machine
• In the case of ties, place machines with higher 3. Arrange the machines in order of decreasing From/To ratios
• Machines with high ratios are placed at the beginning of the flow.
‘From’ values first.
• In the case of ties, place machines with higher ‘From’ values first.

31 32

2
 Solve the previous problem using Hollier Method 2:
Step 1 Step 2 Performance Measures for
2 3 7 4 From From To From/To Order

2 - 62 145 207 2
sums
207
sums
167
ratio
1.24 2
Machine Sequences in GT Cells
3 167 - 167 3 167 0 ∞ 1
7 - 12 12 7 12 202 0.06 4 • Percent of in-sequence moves: considers moves
4 140 - 140 4 140 140 0.89 3 between adjacent machines in the flow direction
To 167 0 202 157 (high values are good)
Sequence:
q 3 → 2 → 4 → 7 • Percent of backtracking moves: considers moves
between adjacent machines opposite the flow
1. Develop the From-To Chart
direction (low values are good)
2. Determine the From/To ratio for each machine by dividing the ‘From- • Percent of bypassing (forward) moves: forward
sum’ by the ‘To-sum’ for each machine moves where the machines are not adjacent to each
3. Arrange the machines in order of decreasing From/To ratios other
• Machines with high ratios are placed at the beginning of the flow. • Percent of backward moves: backward moves where
• In the case of ties, place machines with higher ‘From’ values first. the machines are not adjacent to each other

33 34

Example: Determine the goodness of the machine sequence in the previous example in terms of %
in sequence, % backtracking moves, % forward moves, and % backward moves. Recall that the The Flow Diagram showing the machine sequence may
machine sequence is 3→2→4→7, and that the process plans, and the From-to chart of moves are:
suggest what machine layout to use. For instance
Part Weekly quantity Machine routing Part Weekly quantity Machine routing
A 50 3→2→7 F 60 5→1
(b) The flow diagram
B 20 6→1 G 5 3 →2 → 4
62
C 75 6→5 H 100 3 →2 → 4 → 7
D 10 6→5→1 I 40 2→4→7
167 167 145 140 190
E 12 3→2→7→4 J 15 5→6→1 3 2 4 7
12
to
2 3 7 4 (b) The flow diagram 40
62
17
2 - 62 145
3 167 -
from
m

190
7 - 12 167
3
167
2
145
4
140
7
4 140 - 12

In-sequence moves Backtracking moves 190

7
40 17

(# in-sequence moves) = 167+145+140 = 452 moves 62

(# backtracking moves) = 12 moves 167 167
(# bypassing {forward} moves) = 62 moves
3 2 12 140
(# total moves) = 526 moves 145
(% in-sequence moves) = 100(452/526) = 85.93% 4 17
(% backtracking moves) = 100(12/526) = 2.3% 40
(% forward moves) = 100(62/526) = 11.8% The machine sequence
35 36

Machine-Part Group Formation

Solution Procedures
Solution Procedures
1. Production flow analysis (Burbidge, 1971) • From Chapter 18 in your text, study
2. Rank order clustering (King 1980) – Rank Order clustering algorithm (18.6)
3. Clustering
g Analysis
y – Similarityy Coefficients – Performance measures (18.6)
(Anderberg, 1973; McAuley, 1972;
• From lecture notes
Seifoddini & Wolfe, 1986)
– Hollier Method 2
4. Graph Partitioning
(See Kusiak and Chow (1988) for a literature
survey)

37 38