LECTURE 5 - TRANSPORTATION AND ASSIGNMENT PROBLEMS

5.1 Transportation problem

The transportation model is a special class of linear programming that deals with shipping
a commodity from sources (e.g., factories) to destinations (e.g., warehouses). The objective is to
determine the shipping schedule that minimizes the total shipping cost while satisfying supply and
demand limits. In general, the transportation model can be extended to other areas of operation,
including, among others, inventory control, employment scheduling, and personnel assignment.

5.1.1 Mathematical Formulation of the problem

Let m : be the number of sources

n : be the number of destinations
: be the transportation cost per unit
: be the amount shipped
: be the amount of supply at source i
: be the amount of demand at source j

The objective of the model is to determine the unknowns that minimize the total
transportation cost while satisfying all the supply and demand restrictions:

Minimize

Subject to the constraints

for all i and j

Note: The transportation algorithm is based on the assumption that the model is balanced,
meaning that the total demand equals to total supply.

5.1.2 The Transportation Algorithm

The transportation algorithm follows the exact steps of the simplex method (Chapter 2).
However, instead of using the regular simplex table, we take advantage of the special structure of
the transportation model to organize the computations in a more convenient form. The
transportation table:

Destinations

AHA – Lecture 5 1
 ……. Supply,i
…….
…….
Sources
…… ……. ……. ……. ……. ……
…….
Demand,j …….

The steps of the transportation algorithm are as follows:

Step 1: Determine a initial basic feasible solution, and go to step 2.

Step 2: Use the optimality condition of the simplex method to determine the entering
variable from all the non-basic variables. If the optimality condition is satisfied,
stop. Otherwise, go to step 3.
Step 3: Use feasibility condition of the simplex method to determine the leaving
variable from the among all the current basic variables, and find the new basic
solution. Return to step 2.

5.1.3 Determination of the Initial Basic Feasible Solution

A general transportation model with m sources and n destinations has constraint

equations, one for each source and each destination. However, because the transportation model
is always balanced ( total supply = total demand), one of these equations is redundant. Thus, the
model has independent constraint equations, which means that the starting basic
solution consists of basic variables. Whenever the number of basic cells is less than
, the transportation is degenerate. To resolve degeneracy, the positive variables are
augmented by as many zero-valued variable as is necessary to complete basic
variables.
The special structure of the transportation problem allows securing a non-artificial starting
basic variable using one of the two methods:
1) Northwest-corner method
2) Least Cost method
3) Vogel approximation method
In general, the Vogel method yields the best initial basic solution, and the northwest-
corner method yields the worst. The trade-off is that the northwest-corner method involves the
least computation.

5.1.4 Northwest-Corner Method: The method starts at the northwest-corner cell of the table
(variable )

Step 1: Allocate as much as possible to the selected cell, adjust the associated amounts of
supply and demand by subtracting the allocated amount.

Step 2: Cross out the row or column with zero supply or demand to indicate that no further
assignments can be made in that row or column. If both a row and a column net to zero
simultaneously, cross out one only.

Step 3: If exactly one row or column is left uncrossed out, stop. Otherwise, move to the cell to
the right if a column has just been crossed out or below if a row has been crossed out.
Go to step 1.
Example 1:

AHA – Lecture 5 2
 Determine an initial basic feasible solution to the following transportation problem using
the Northwest-corner method.

Mill 1 Mill 2 Mill 3 Mill 4 Supply

Silo 1 10 2 20 11
15
Silo 2 12 7 9 20
25
Silo 3 4 14 16 18
10
Demand 5 15 15 15

Solution:

Mill 1 Mill 2 Mill 3 Mill 4 Supply

Silo 1 10 2 20 11
15
Silo 2 12 7 9 20
25
Silo 3 4 14 16 18
10
Demand 5 15 15 15

The total cost, z = 10 (5) + 2 (10) + 7 (5) + 9 (15) + 20 (5) + 18 (10) = 520

5.1.5 The Least Cost Rule: The only difference between the two rules is the criterion used for
selecting the successive basic variables. In the least cost
rule, the variable with the lowest shipping cost will be
chosen as the basic variable.

Example 2: Using Example 1, determine an initial basic feasible solution to the transportation
problem using the Least Cost Rule.

Solution:

Mill 1 Mill 2 Mill 3 Mill 4 Supply

Silo 1 10 2 20 11
15
Silo 2 12 7 9 20
25
Silo 3 4 14 16 18
10
Demand 5 15 15 15

The total cost, z = 2 (15) + 9 (15) + 20 (10) + 4 (5) + 18 (5) = 475

In general the least cost rule provides a better starting solution as compared to the northwest
corner rule. But this is not guaranteed in all problems. In fact, examples have been constructed
wherein the opposite is true.

AHA – Lecture 5 3
5.1.6 Vogel Approximation Method (VAM)

Step 1: For each row (column), determine a penalty measure by subtracting the smallest unit
cost element in the row (column) from the next smallest unit cost element in the same
row (column).

Step 2: Identify the row or column with the largest penalty. Ties are broken arbitrarily. Allocate as
much as possible to the variable with the least unit cost in the selected row or column.
Adjust the supply and demand, and cross out the satisfied row or column. If a row and a
column are satisfied simultaneously, only one of the two is crossed out, and the
remaining row (column) is assigned zero supply (demand).

Step 3:
(a) If exactly one row or column with zero supply or demand remains uncrossed out,
stop.
(b) If one row (column) with the positive supply (demand) remains uncrossed out,
determine the basic variables in the row (column) with the least unit cost. Stop.
(c) If all the uncrossed out rows and columns have zero supply and demand,
determine the zero basic variables with the least unit cost. Stop.
(d) Otherwise, go to step 1.

Example 3:
Using previous example, determine an initial basic feasible solution by VAM
Solution:

The total cost, z = 15 (2) + 15 (9) +10 (20) + 5 (4) + 5 (18) = 475

5.1.7 Iterative Computations of the Transportation Algorithm

After determining the initial solution, we use the following algorithm to determine the
optimum solution:

AHA – Lecture 5 4
Step 1: Use the simplex optimality condition to determine the entering variable as the current
non-basic variable that can improve the solution. If the optimality condition is satisfied,
stop. Otherwise, go to step 2.

[Optimality Condition: The entering variable in a maximization (minimization) problem is

the nonbasic variable having the most negative (positive) net evaluation. Ties are
broken arbitrarily. The optimum is reached at the iteration where all the net evaluation of
the nonbasic variables are nonnegative (nonpositive)]

Step 2: Determine the leaving variable using simplex feasibility condition. Change the basis, and
return to step 1.

[Feasibility Condition: For both the maximization and the minimization problems, the
leaving variable is the basic variable associated with the smallest nonnegative ratio (with
strictly positive denominator). Ties are broken arbitrarily]

Example 4:
Solve the transportation model of example 1, starting with the northwest-corner solution.

Solution:

Mill 1 Mill 2 Mill 3 Mill 4 Supply

Silo 1 10 2 20 11
5 10 15
Silo 2 12 7 9 20
5 15 5 25
Silo 3 4 14 16 18
10 10
Demand 5 15 15 15

The determination of the entering variable from among the current non-basic variable
(those that are not part of the initial basic solution) is done by computing the non-basic
coefficients in the z-row, using the method of multipliers.

In the method of multipliers, we associate the multipliers and with row i and column j
of the transportation table. For each current basic variable , these multipliers have to satisfy
the following equations:

for each basic

In example 3.1, we have 7 variable and 6 equations correspond to the six basic variables.
To solve these equations, the method of multipliers calls for arbitrarily setting .

Next, we use and to evaluate the non-basic variables by computing

for each non-basic

The preceding information, together with the fact that for each basic , is
actually equivalent to computing the of the simplex table.

AHA – Lecture 5 5
 Because the transportation model seeks to minimize cost, the entering variable is the one
having the most positive coefficients in the . (NB: If it is a maximization problem, the
entering variable is the one having the most negative coefficients).

Mill 1 Mill 2 Mill 3 Mill 4 Supply

Thus,
Silos 1 10 2 20 11
choose
5 10 15
as
the Silos 2 12 7 9 20
entering 25
5 15 5
variable.
Silos 3 4 14 16 18
10
10
Having
Demand 5 15 15 15
determined as the entering variable, we need to determine the leaving variable. The selection
of as the entering variable means that we want to ship through this route because it reduces
the total shipping cost. What is the most that we can ship through this route? If route (3, 1) ships
, then the maximum value of is determined based on two conditions:

(a) Supply limits and demand requirements remain satisfied.

(b) Shipments through all routes must be non-negative.

These two conditions determine the maximum value of and the leaving variable in the
following manner. First, construct a closed loop that starts and ends at the entering variable cell
(3, 1). The loop consists of connected horizontal and vertical segments only (no diagonals are
allowed). Except for the entering cell, each corner of the closed loop must coincide with the basic
variable.

Next, we assign the amount to the entering variable cell (3, 1). For the supply and
demand limits to remain satisfied, we must alternate subtracting and adding the amount at the
successive corners of the loop.

Supply
10 2 20 11
5 10 (-16) (4) 15
12 7 9 20
(3)
25
5 15 5
4 14 16 18
(9) (-9) (-9)
10
10
Demand 5 15 15 15
The maximum value of is 5, which occurs when both and reach zero level.
Because only one current basic variable must leave the basic solution, we can choose either
or as the leaving variable. We arbitrarily choose to leave the solution.

The selection of as the entering variable and as the leaving variable requires
adjusting the values of the basic variables at the corners of the closed loop. Given the new basic
solution, we repeat the computation of the multipliers u and v.

AHA – Lecture 5 6
 Mill 1 Mill 2 Mill 3 Mill 4 Supply
Choose
Silos 1 10 2 20 11
as 15 15
the
entering Silos 2 12 7 9 20
variable. 25
0 15 10
Silos 3 4 14 16 18 Supply
10
5 10 2 20 5 11
Demand 5
(-9) 15 15 15
(-16)
15
(4) 15
12 7 9 20
(-6) (0)
25
0 15 10
4 14 16 18
(-9) (-9)
10
5 5
Demand 5 15 15 15
The closed loop shows that the entering variable and that the leaving variable is
. Given the new basic solution, we repeat the computation of the multipliers u and v.

Supply
10 2 20 11 15
(-13) 5 (-16) 10
12 7 9 20 25
(-10) 10 15 (-4)

4 14 16 18 10
5 (-5) (-5) 5
Demand 5 15 15 15
Since the new are now negative for all non-basic , the last table is optimal.
The following table summarizes the optimal solution

From Silo To Mill Number of truckload

1 2 5
1 4 10
5.1.8
Unbalanced 2 2 10
2 3 15
3 1 5
3 4 5
Optimal solution = $435
Transportation Problem

AHA – Lecture 5 7
Case 1 : If total supply > total demand [ ]

Introduce a dummy destination in the transportation table. The cost of the transporting to
this destination is all set to zero. The demand at this destination is assumed to be

Case 2 : If total supply < total demand [ ]

Introduce a dummy source in the transportation table. The cost associated with are set to
zero. The supply at this source is assumed to be

5.2 Assignment problem

The assignment problem is a special type of linear programming problem where

assignees are being assigned to perform tasks. For example, the assignees might be
employees who need to be given work assignments. Assigning people to jobs is a common
application of the assignment problem. However, the assignees need not be people. They could
be machines, or vehicles, or plants, or even time slots to be assigned tasks.
To fit the definition of an assignment problem, these kinds of applications need to be
formulated in a way that satisfies the following assumptions:

(i) The number of assignees and the number of tasks are the same (Balanced).
(ii) Each assignee is to be assigned to exactly one task.
(iii) Each task is to be performed by exactly one assignee.
(iv) There is a cost associated with assignee i performing task j
(v) The objective is to determine how all assignments should be made to minimize
the total cost.

The general assignment model with n assignees and n tasks is presented as follows:

Tasks
1 2 …… n 1
1 1
2
Assignee : : : : : :
: : : : : :
: : : : : :
n 1
1 1 1 1

The assignment model is actually a special case of the transportation model in which the
assignees represent the sources, and the tasks represent the destinations. In effect, the
assignment model can be solved directly as a regular transportation model. Nevertheless, the fact
that all supply and demand amounts equal 1 has led to the development of a simple solution

AHA – Lecture 5 8
algorithm called the Hungarain method. Although the new method appears totally unrelated to
the transportation model, the algorithm is actually rooted in the simplex method, just as the
transportation model is.

5.2.1 Mathematical Formulation of the problem

Let = 0, if the task not assigned to assignee

= 1, if the task assigned to assignee

The objective of the model is to determine the unknowns that minimize the overall cost:

Minimize

Subject to the constraints

for all i and j

5.2.2 The Assignment Algorithm – Hungarian method

The steps of the assignment algorithm are as follows:

Step 1: For the original cost matrix, identify each row’s minimum, and subtract it from all the
entries of the row.
Step 2: For the matrix resulting from step 1, identify each column’s minimum, and subtract it from
all the entries of the column.
Step 3: Assign the zeroes:
(a) Examine the rows of the current matrix successively until a row with exactly one
unmarked zero is found. Circle this zero, indicating that an assignment will be
made there. Cross out all other zeroes lying in the column of above encircled zero.
The crossed cells will not be considered for any future assignment. Continue this
manner until all the rows have been taken care of.
(b) Similarly for column.
Step 4: Check for Optimality: Repeat step 3 successively till one of the following occurs.
(a) There is no row and no column without assignment. In such a case, the current
assignment is optimal.
(b) There may be some row or column without assignment. In this case, the current
solution is not optimal. Proceed to next step.
Step5: Draw minimum number of lines crossing all zeroes as follows: If the number of lines equal
to the order of the matrix, then the current solution is optimal, otherwise it is not optimal.
Proceed to the next step.
Step6: Examine the elements that do not have a line through them. Select the smallest of these
elements and subtract the same from all the elements that do not have a line through
them, and add this element to every element that lies in the intersection of the two lines.
Repeat this until an optimal assignment is reached.

Example 5:

AHA – Lecture 5 9
 Joe’s three children – John, Karen and Terri – want to earn some money to take care of
personal expenses during a school trip to the local zoo. Joe has chosen three tasks for his
children: mowing the lawn, painting the garage, and washing the family cars. To avoid anticipated
sibling competition, he asked them to summit (secret) bids for what they feel was a fair pay for
each of the three tasks. The understanding then was that all three children would abide their
father’s decision as to who gets which tasks. The following table summarizes the bids received.

Mow Paint Wash

John $15 $10 $9
Karen $9 $15 $10
Terri $10 $12 $8

Based on this information, how should Joe assign the tasks?

Solution: Let and be row i and column j minimum costs as defined in step 1 and 2
respectively.
Mow Paint Wash Row minimum
John 15 10 9
Karen 9 15 10
Terri 10 12 8

Mow Paint Wash

John 6 1 0
Karen 0 6 1
Terri 2 4 0
Column minimum

Mow Paint Wash

John 6 0 0
Karen 0 5 1
Terri 2 3 0

The cells with underscored zero entries provide the optimum solution. The total cost is 9 + 10 +
8 = 27. This amount also will always equal .

Example 6:
Suppose that the situation discussed in Example 5 is extended to four children and four
tasks. The following table summarizes the cost elements of the problem.

Task
1 2 3 4
1 $1 $4 $6 $3
2 $9 $7 $10 $9
Child 3 $4 $5 $11 $7
4 $8 $7 $8 $5

Based on this information, how should you assign the tasks?

Solution:
Step 1:
Task

AHA – Lecture 5 10
 1 2 3 4
1 0 3 5 2
2 2 0 3 2
Child 3 0 1 7 3
4 3 2 3 0

Step 2:
Task
1 2 3 4
1 0 3 2 2
2 2 0 0 2
Child 3 0 1 4 3
4 3 2 0 0

The locations of the zero entries do not allow one task per child. For example, if we assign child 1
to task1, then column will be terminated, and child 3 will not have a zero entry in the remaining
three columns. This obstacle can be accounted for by drawing the minimum number of lines that
will covers all the zero entries – step 5.

Task
1 2 3 4
1 0 3 2 2
2 2 0 0 2
Child 3 0 1 4 3
4 3 2 0 0

The smallest un-shaded cell equals 1. This element is subtracted from all the unshaded cells and
added to the intersection cells.

Task
1 2 3 4
1 0 2 1 1
2 3 0 0 2
Child 3 0 0 3 2
4 4 2 0 0

There are no four lines drawn (shaded cells).The cells with underscored zero entries provide the
optimum solution. The assignment will be: Child 1 -Task 1, Child 2 -Task 3, Child 3 -Task 2 and
Child 4 –Task 4. Total cost is: 1 + 10 + 5 + 5 = 21. The same cost is also determined by summing
the ’s, the ’s and the smallest unshaded cell.

Example 7

Consider the problem of assigning five jobs to five persons. The assignment costs are given as
follows:

Jobs
Person 1 2 3 4 5

A 8 4 2 6 1
B 0 9 5 5 4
C 3 8 9 2 6

AHA – Lecture 5 11
 D 4 3 1 0 3
E 9 5 8 9 5

5.2.3 Unbalanced Assignment Problem

When the cost matrix of an assignment problem is not a square matrix (# of assignees is
not equal to the # of tasks), the assignment is called an unbalanced assignment problem. In such
problems, dummy rows or columns are added in the matrix so as to complete it to form a square
matrix.

AHA – Lecture 5 12