MDP408a

Production & Operations

Management (Facilities Planning)
Lecture #14
Facility Layout Planning (cont.)
 Today’s lecture
• Computerized layout algorithms
– Pairwise Exchange Method
– CRAFT
– Micro-CRAFT (MCRAFT)
– LOGIC
– MULTIPLE
Pairwise Exchange Method (i)

Assuming that department areas do not affect their assigned locations.

Pairwise Exchange Method (ii)
 Pairwise Exchange Method (iii)
• TCABCD = 5(15)+8(35)+10(50)+5(20)+3(35)+2(15)=1090

• TCBACD = 5(15)+8(20)+10(35)+5(35)+3(50)+2(15)=940

• TCCBAD = 5(20)+8(35)+10(15)+5(15)+3(35)+2(50)=810

• TCDBCA= 5(35)+8(15)+10(50)+5(20)+3(15)+2(35)=1010

• TCACBD = 5(35)+8(15)+10(50)+5(20)+3(15)+2(35)=1010

• TCADCB= 5(50)+8(35)+10(15)+5(15)+3(35)+2(20)=900

• TCABDC= 5(15)+8(50)+10(35)+5(35)+3(20)+2(15)=1090

C-B-A-D
 Pairwise Exchange Method (iv)

• Remarks:
– Not guaranteed to yield the optimal layout (as it
is a heuristic)
– It is possible to cycle back
– Symmetric layout arrangements are possible
– Only considers departments whose sizes or areas
do not affect and are not affected by their
locations.
 CRAFT-(Computerized Relative Allocation of
Facilities Technique)
• First computer-aided layout algorithm (1963)
• The input data is represented in the form of a
From-To chart, or qualitative data (activity
relationships).
• The main objective behind CRAFT is to minimize
total transportation cost:

• Department representation
o Discrete grids
o No shape restrictions
• Improvement-type layout algorithm
 Steps in CRAFT
1. Start with an initial layout with all departments made up of
individual square grids (Note: each grid represents the same
amount of space)
2. Calculate centroid of each department and rectilinear distance
between pairs of departments centroids (stored in a distance
matrix).
3. Find the cost of the initial layout by multiplying the
– From-To (flow) chart,
– unit cost matrix, and
– From-To (distance) matrix
4. Improve the layout by performing all-possible two-way exchanges
– At each iteration, CRAFT selects the interchange that results in the
maximum reduction in transportation costs
– These interchanges are continued until no further reduction is possible
– Only consider exchanging adjacent departments or departments with
equal areas
 Example
Flow Matrix
A
C A B C D E F
E
A - 0 8 0 4 0
B - 0 5 0 2
B
D C - 0 1 0
F D - 6 0
E - 4
Existing Layout F -

Assuming that the costs are all 1’s.

What are all possible pairwise departmental exchanges?

 A
C
E

B
D
F

Existing Layout
The distance and flow matrices are given as follows:

A B C D E F A B C D E F
A - 0 8 0 4 0 A - 3 3.5 6.5 7 10
B - 0 5 0 2 B - 5.5 3.5 8 7
F= C - 0 1 0 D= C - 3 3.5 6.5
D - 6 0 D - 5.5 3.5
E - 4 E - 3
F - F -

The current distance-based objective = 136

How did we calculate it?

Estimate the distance-based measure for possible pairwise exchanges…
 A
C
E

B
D
F

Existing Layout

Given the initial layout, which department pairs will not be considered for exchange?

A & D, E & A, F & B, C & F

All possible exchanges are:

A & B, A & C, A & F, B & D, B & C, B & E, D & C, D & E, D & F, E & F
So, we have here 10 alternative exchanges, we have to evaluate their
distance based objective value.
But we need to do it using a quick estimate. It does not have to be accurate for now.
 A A B C D E F
C A - 3 3.5 6.5 7 10
E
B - 5.5 3.5 8 7
D= C - 3 3.5 6.5
B
D D - 5.5 3.5
F E - 3
F -
Existing Layout Initial distance matrix

For instance, compute the estimated distance-based measure for the layout in which
departments E and D will be exchanged then evaluate its exact value.
When departments E and D are exchanged, the estimated distance-based measure is
evaluated as follows:
A B C D E F A B C E D F
A - 0 8 0 4 0 A - 3 3.5 6.5 7 10
B - 0 5 0 2 B - 5.5 3.5 8 7
F= C - 0 1 0 D= C - 3 3.5 6.5
D - 6 0 E - 5.5 3.5
E - 4 D - 3
F - F -

The estimated distance-based objective = 158  this is greater than the original value (136)
→ we have to do the same with all other possible exchanges then choose the best one.
 Suppose that minimum objective value is for departments E and D,
we now calculate their exact objective value by modifying the layout.
The modified layout will be:

A A
C C D
E

B B E
D
F F
o
Existing Layout

The corresponding exact distance matrix is:

A B C D E F
A - 3 3.5 6.5 7 10
B - 5.5 8.5 4 7
D= C - 3 3.5 6.5
D - 5 3.5
E - 3
F -

The exact distance-based objective = 158. This happened to be equal to the estimated
value. This is not going to happen all the time.
 Therefore, the new layout will be:

A
C D

B E
F
o

CRAFT will record the best solution found so far. Even though the current solution is
not better than the current best solution in this example (which is the initial given layout),
the algorithm will continue by investigating other pairwise departmental exchanges
starting from the current new layout.

The iterations will continue until a specific stopping criteria is satisfied.

Another Example
Another Example – Contd.
Another Example – Contd.

Exchange departments B and C

 CRAFT Facts
• CRAFT only exchanges departments that are
– Adjacent (share at least one common edge)
– Have equal areas
• Adjacency is a necessary but not sufficient criteria for
swapping departments
• Quality of final solution depends on the initial layout
• Final solution may be locally optimal, not globally
optimal
 Pitfalls of CRAFT
 Exchange departments 2 and 4 in the layout shown below. All
departments are fixed except 2 & 4.
5 5 5 6 6
5 5 5 6 6
5 5 5 6 4
5 5 5 4 4
2 2 2 2 2
1 1 2 3 3
1 1 2 3 3

 If the flow from A to B is 4, A to C is 3, and B to C is 9, and all move

costs are 1, what is the layout cost?
C C C C C C
C B B B B C
C B A A B C
C B B B B C
C C C C C C
 MICRO-CRAFT (MCRAFT)
• Can exchange any two departments whether they are
adjacent or not
• Shifts other departments “automatically” if the exchanged
departments are not adjacent and equal sized
• Uses “bands” and assigns grid within band to a
department
• User input:
– Length and width of building
– # of bands
– Layout vector or fill sequence
• Computes grid size, number of rows and columns
• Procedure
– Starts from top-left corner of building
– Sweeps bands in a serpentine fashion
MCRAFT - Example
 MCRAFT - Facts
• Band widths are equal
– A department may occupy two bands
• Uses a From-To chart for input flow data
• Limitation: fixed department
– Fixed department may move
– MCRAFT’s primary strength is also its primary
weakness
 LOGIC
• Layout Optimization with Guillotine Induced Cuts
• Inputs are a from-to chart, total area, dept areas
• Distance-based objective function
• Construction algorithm
• Divides building into smaller and smaller sections by successive
horizontal or vertical cuts
• Stops when each section is a single department
• Constructs a tree to keep track of cuts and subsections
• Choice of cuts made using search strategy
Cut-Tree for Example
A,B,C,D,E,F,G,H
Example
 MULTIPLE
• MULTI-floor Plant Layout Evaluation
• Can also be used for single floors
• Similar to CRAFT
• Uses From-To chart
• Objective function identical to CRAFT
• Not restricted to rectangular shapes
• Discrete representation
• Improvement-type algorithm
• Uses two-way exchanges
• Like MCRAFT, MULTIPLE can exchange non-adjacent
departments
• Uses space-filling curves to place departments
 MULTIPLE
• Layout vector denotes order of department
placement
• Fill sequence is determined by Space Filing Curve
(SFC)
• Can exchange departments by changing their
position in layout vector
• SFC can bypass fixed departments or areas
• Can also be used as a construction procedure
• SFC can move around a fixed department
Example
 Using filling curves
 How does MULTIPLE use these curves to
generate layouts?
▪ With a curve established for a facility
1. Define a department layout sequence
2. Place the departments in the facility in department sequence
along the curve, grid by grid
 Using filling curves
 Example
Alternative filling curves
 Multi-Floor Facility Layout
 More common in older cities and buildings
and in countries where land is very
expensive
 Less flexibility of use of total facility space
▪ A multi-floor building may have enough total
space but may not be able to accommodate
departments on different levels without
splitting
 Multi-Floor Facility Layout
(cont.)

 Additional decisions/constraints
▪ The number and location of vertical material
handling devices
▪ Restriction of certain departments to specific
floors
• First floor restrictions due to floor loading capacity
• Ceiling heights
• Heat/chemical use and/or generation
▪ Departments may have to be kept contiguous
 Multi-Floor Facility Layout
(cont.)

 Modified distance-based objective

m m
z =  (cijH dijH + cijV dijV ) fij
i =1 j =1

where,

cijH = Cost of moving one unit one horizontal distance unit

cijV = Cost of moving one unit one vertical distance unit
 Multi-Floor Facility Layout
(cont.)

 Example

Floor r Floor s
Dept. j

Dept. i
Vertical Travel