MODI-METHOD

The following method of solving a transportation problem is known as MODI Method.

1. Determine the initial basic feasible solution of the given transportation problem by using any method (viz. Row
Minima or Column Minima or Matrix Minima or Vogel’s Approximation Method).
2. Calculate all the net-evaluations corresponding to non-basic cells. If all the net-evaluations are non-positive
quantities at any iteration, the solution is optimum and the minimum cost can be calculated.
3. If the solution is not optimal, determine the entering cell and the leaving cell (wrt the basis) and construct a new
transportation table with readjusted basic variables. Repeat step (2) until an optimum solution is obtained.

Example: Determine the optimal solution to the given problem:

D1 D2 D3 D4 ai

O1 19 30 50 10 7

O2 70 30 40 60 9

O3 40 8 70 20 18

bj 5 8 7 14 34

TABLE 1: Initial Basic Feasible Solution using the Matrix Minima Method

D1 D2 D3 D4 ai

O1 19 30 50 7 10 7

O2 2 70 30 7 40 60 9 2

O3 3 40 8 8 70 7 20 18 10 3

bj 5 8 7 14 7 34

Special Note on the step-wise allocation of units made in the above table:

1. Cell (3,2) has the minimum cost (i.e. 8), the min(a3 ,b2) = 8 is allocated to the cell. This infers that the
requirement of D2 is fulfilled and no more units will be allocated to the 2nd column and the availability
of O3 changes to 10.
 2. Cell (1,4) has the minimum cost(i.e. 10), the min(a1 ,b4) = 7 is allocated to the cell. This infers that the
availability of O1 is nil and no more units will be allocated to the 1st row and the requirement of D4
changes to 7.
3. Cell (3,4) has the minimum cost (i.e. 20), the min(a3 ,b4) = 7 is allocated to the cell. This infers that the
requirement of D4 is fulfilled and no more units will be allocated to the 4th column and the availability
of O3 changes to 3.
4. Cell (2,3) has the minimum cost(i.e. 40), the min(a1 ,b4) = 7 is allocated to the cell. This infers that the
requirement of D3 is fulfilled and no more will be allocated to the 3rd column and the availability of O2
changes to 2. (Here allocation may be done otherwise to cell (1,3) if one wishes to. )
5. Finally Cell(3,1) is allocated 3 and Cell (2,1) is allocated 2 in a similar manner.

TABLE 2: Checking Optimility of the Initial BFS obtained by using the UV-Method a.k.a. MODI-Method

D1 D2 D3 D4 ui

19 30 50 𝟕−𝝐 10 -10
O1 +𝜖

11 - 32 -50

70 30 40 60 30
O2 2 7

8 -10

𝟑−𝝐 40 8 70 20 0
O3 8 𝟕
+𝝐
-60

40 8 10 20
vj

Procedure for finding the net-evaluations::

1. The values of u,v’s are found by using the cost in the basic cells (i.e. the cells with allocation).
2. As the 3rd row has the most number of allocations, we take the corresponding value of u3 =0.
3. Noting down that for the basic cell (the cells with allocation) )s, 𝑢! + 𝑣" = 𝑐!" , the other values of u’s and v’s are
calculated.
4. Then the net-evaluations are calculated for the non-basic cells by using the formula
𝒛𝒊𝒋 − 𝒄𝒊𝒋 = +𝒖𝒊 + 𝒗𝒋 . − 𝒄𝒊𝒋

Conclusion from the Table 2: Since all the net-evaluation are not non-positive (see cell (1,1) and cell (2,2)) , the
allocation in Table 2 is not optimum. Hence, we are to construct another table by readjusting the allocation to
Table 2..
 Procedure for Readjustment to construct Table 3:

1. Since the max positive evaluation (i.e. 11) is corresponding to the cell (1,1), we take cell (1,1) to be the
entering cell and assign a non-negative value, say 𝜖 to it.
2. To ensure that problem remains the same, in the sense that the requirement and availability, a
loop(shown in dotted lines in Table 2) is bound to be formed.
3. The cell that leaves the basis is then taken to be (3,1) as it has the least allocation out of all the cells
which include a negative epsilon.

Table 3: Readjusted Transportation Table

D1 D2 D3 D4 ui

19 30 50 10 0
O1 3+𝜖 4−𝜖
- 32 -50

70 30 40 60 51
O2 2−𝜖 +𝜖 7

23 -5

40 8 70 20 6
O3 8−𝜖 𝟏𝟎 + 𝝐
-15 -75

19 2 -11 4
vj

Conclusion from Table 3:

1. Since, the non-basic cell (2,2) has a positive net-evaluation, the solution is not optimal.
2. In this regard, we take take the cell (2,2) to be the entering cell and the cell (2,1) to be the cell leaving the basis
(as it is the maximum value epsilon can take.
Table 4: Readjusted Transportation Table

D1 D2 D3 D4 ui

19 30 50 10 0
O1 5 2
- 32 -42

70 30 40 60 32
O2 2 7

-11 -18

40 8 70 20 10
O3 6 12

-11 -52

19 -2 8 10
vj

Since the net-evaluations for all the non-basic cells are non-positive, this is the optimum allocation. And the
optimum (minimum ) cost is :

5*19 + 2*10 + 2*30 + 7*40 + 6*8 + 12*20 =

---------------------------------------------------------------------------------------------------------------------------------------------------
P.S. The same problem is done in Example 13.4.2 in Karak’s book by finding the initial BFS by using VAM
Method.