Decision Sciences-II

Case Solution: American Automobiles Limited: Production Planning

Presented to
Professor Trilochan Sastry

Indian Institute of Management, Bangalore

On
December 5, 2017

In
Partial Fulfillment of the Requirements for the
Decision Sciences II Course in the
Post Graduate Programme

By

(Group – 13, Section F)

Aninda Halder 1711385

Darpalli Gayathri 1711391
Kushal M B 1711403
Yagya Sharma 1711440
Zaidali Khoja 1711442

This study source was downloaded by 100000831850640 from CourseHero.com on 10-22-2023 02:36:02 GMT -05:00

https://www.coursehero.com/file/36765704/DS-Assignment-2pdf/
 Case Context

American Automobiles Limited (AAL) was faced with the challenge to optimize the
production costs of its stamping operations. The stamping shop consisted of six machines, each
with different shot and shift costs, and varying changeover times. So far, the allocation of parts
to machines has been driven by intuition and judgment. AAL, however, wants to move to an
objective production plan, keeping cost minimization in mind. For reference, the process flow
of stamping operations is given below:

Problem Definition

For the purpose of the issue, a set of 10 parts is to be produced across the 6 machines present
in the stamping division. However, the objective function for production is limited by the
following constraints:

1. Shot Costs which are dependent on the number of allocated parts to be produced, the demand
for a particular model and the cost incurred per shot.

2. Shift Costs which are dependent on the number of shifts per machine and the cost per shift
for particular press.

We, therefore, define the following variables for our problem:

Xij = Number of parts of model i produced using machine j, where i = 1…10 and j = 1…6
Cj = Shot cost for part j, where j = 1…6
Di = Demand for model i, where i = 1…10
Sj = Number of shifts for machine j, where j = 1…6

The cost function to minimize will therefore be:

The daily demand for each model, shot costs, shift costs and changeover times for each of the
machines are shown in the exhibits below:

Model No. Model ID Daily Demand

1 VM1002 53
2 VM1011 55
3 VM1030 79

This study source was downloaded by 100000831850640 from CourseHero.com on 10-22-2023 02:36:02 GMT -05:00

https://www.coursehero.com/file/36765704/DS-Assignment-2pdf/
 4 VM2011 39
5 VM1333 31
6 VM1153 65
7 VM1153 9
8 VM2341 71
9 MR128374 136
10 MR128361 7

Besides, we also have information about the production restrictions on the stamping shop. Our
objective function will therefore, be subject to the following constraints:

1. Hurdle Constraints

For our problem, there are certain flexibility constraints in that the number of parts of a given
model can be stamped only on a particular machine. A snapshot of the flexibility hurdle matrix
is given. A cell in the matrix indicates the minimum number of units of part i to be produced
by machine j.

The constraints and the flexibility hurdle matrix are shown below, we build the following
constraints:

X13 ≥ 6 X23 ≥ 3 X26 ≥ 3

X36 ≥ 3 X43 ≥ 4 X46 ≥ 3
X53 ≥ 2 X56 ≥ 3 X62 ≥ 2
X66 ≥ 3 X83 ≥ 2 X86 ≥ 3
X92 ≥ 1 X10,2 ≥ 1

This study source was downloaded by 100000831850640 from CourseHero.com on 10-22-2023 02:36:02 GMT -05:00

https://www.coursehero.com/file/36765704/DS-Assignment-2pdf/
 2. Sum of Parts

For each model, the sum of the number of parts allocated to each machine must be equal to the
total number of parts for that model. Therefore, we need to constrain the number of parts we're
allocating across each machine. Hence, the following constraint would hold.

where, Xij is the number of parts of model i produced using model j and P i is the total number
of parts for a particular model. For example, Pi for model 1 is 20. Hence, the parts for model 1
will be allocated across the 6 machines in a way that it equals P i.

The list of model and value of Pi for each model is given below:

Model No. Model ID Number of parts

1 VM1002 20
2 VM1011 25
3 VM1030 24
4 VM2011 23
5 VM1333 20
6 VM1153 24
7 VM1153 2
8 VM2341 22
9 MR128374 2
10 MR128361 1

This study source was downloaded by 100000831850640 from CourseHero.com on 10-22-2023 02:36:02 GMT -05:00

https://www.coursehero.com/file/36765704/DS-Assignment-2pdf/
 3. Number of Shifts and Shift Time

From the data given in Exhibit 4, we know that the maximum number of shifts that can be
performed by a machine in a day is 3. Hence, we limit the number of shifts for a machine to 3.

The constraint is therefore, 0 ≤ Sj ≤ 3

Also, we know that the shift time for a machine is the sum of the stamping time and the
changeover time.

Shift time = Stamping Time + Changeover Time

Therefore, the number of shifts is equal to the shift time divided by the length of a shift, rounded
up to the nearest integer. The linear programming constraint formulation for the same is shown
below:

Daily
Sl. No. Model
Demand
1 VM1002 53
2 VM1011 55
3 VM1030 79
4 VM2011 39
5 VM1333 31
6 VM1153 65
7 VM1153 9
8 VM2341 71
9 MR128374 136
10 MR128361 7

4. Non-negativity Constraints

The number of parts of model i produced using machine j (X ij) and the number of shifts
undertaken by machine j (Sj) are both non-negative integers.

Xij, Sj ≥ 0

This study source was downloaded by 100000831850640 from CourseHero.com on 10-22-2023 02:36:02 GMT -05:00

https://www.coursehero.com/file/36765704/DS-Assignment-2pdf/
 Assignment Questions

1. What is the most cost-effective allocation of the parts to the press machines?

The most cost-effective allocation matrix for parts of each model to each machine is shown
below. Based on this allocation, we find the total cost to be 3879.1. For arriving at this solution,
we used the Solver package provided by Microsoft Excel to solve the linear program with the
given objective function and constraints. However, due to lack of sufficient computing power,
we had to restrict the number of iterations by limiting the number of iterations performed by
Solver. Based on this configuration, we arrived at the following solution.

Allocation Matrix

Optimal Total
M1 M2 M3 M4 M5 M6
Allocation

VM1002 2 0 6 5 7 0 20
VM1011 0 0 3 7 12 3 25
VM1030 19 0 0 0 2 3 24
VM2011 0 0 4 2 14 3 23
VM1333 0 0 2 0 15 3 20
VM1153 3 0 2 7 9 3 24
VM1153 0 0 0 0 2 0 2
VM2341 13 0 2 0 4 3 22
MR128374 0 1 0 0 1 0 2
MR128361 0 1 0 0 0 0 1

Objective Function Values

Shots cost 3699.1

Shift cost 180
Total cost (Z) 3879.1

Number of Shifts Per Machine

Machine M1 M2 M3 M4 M5 M6

Number of Shifts (Sj) 3 1 2 1 3 2

This study source was downloaded by 100000831850640 from CourseHero.com on 10-22-2023 02:36:02 GMT -05:00

https://www.coursehero.com/file/36765704/DS-Assignment-2pdf/
 2. How does the optimal solution compare with the current allocation?

We observed that with our solution, the total value of the objective function stood at 3879.1
while that for the old production plan was 4424.41. The higher cost can be explained by the
higher shot cost and higher number of shifts allocated to machine 2. In our plan, only one shift
is allocated to machine 2 while in the old plan, 3 shifts were allocated to machine 2.

Allocation Matrix (Old Production Plan)

Current
M1 M2 M3 M4 M5 M6 Total
Allocation

VM1002 7 2 6 2 3 0 20
VM1011 7 4 3 3 5 3 25
VM1030 3 6 2 10 0 3 24
VM2011 7 4 4 3 2 3 23
VM1333 8 4 2 0 3 3 20
VM1153 6 4 7 2 2 3 24
VM1153 0 0 2 0 0 0 2
VM2341 1 3 2 13 0 3 22
MR128374 1 1 0 0 0 0 2
MR128361 0 1 0 0 0 0 1

Objective Function Value (Old Production Plan)

Shots cost 4229.41

Shift cost 195
Total cost (Z) 4424.41

Number of Shifts Per Machine (Old Production Plan)

Machine M1 M2 M3 M4 M5 M6

Number of Shifts (Sj) 3 3 2 2 1 2

3. What costs, other than the shot and shift costs, are involved?

For our operations, there are several fixed costs and variable costs involved. A breakdown of
the costs is shown below:

Fixed Costs: Since a lot of machinery is involved in stamping operations, depreciation of

machinery would be a major cost involved, irrespective of the number of units produced. Other
fixed costs involved would be administrative expenses, such as salaries of managers and other
indirect labor, support staff and other overheads. Besides, if the company is financed using

This study source was downloaded by 100000831850640 from CourseHero.com on 10-22-2023 02:36:02 GMT -05:00

https://www.coursehero.com/file/36765704/DS-Assignment-2pdf/
 debt, it will also incur interest costs on that debt. Also, if the land and building of the plant is
not owned by the company, rent costs will be incurred.

Variable Costs: Since, the operations in consideration pertain to manufacturing, raw material
costs are a major expense. Besides, since a batch operation process was involved, material
handling and other direct labor costs will be incurred. Also, since some amount of raw material,
work-in-progress and finished goods inventory is held, inventory holding costs will also be
incurred.

4. Can other types of data help refine the decision?

For solving the problem at hand, we have only considered the stamping and changeover costs
involved. However, the optimal allocation could possibly change if details for other costs (as
mentioned above) are provided. Another critical parameter in the automobile industry is the
quality of the parts produced, which may vary with respect to each machine. Since, each of the
machine is of a different age, taking quality into consideration could impact the optimal
allocation of parts to each machine.

This study source was downloaded by 100000831850640 from CourseHero.com on 10-22-2023 02:36:02 GMT -05:00

https://www.coursehero.com/file/36765704/DS-Assignment-2pdf/
Powered by TCPDF (www.tcpdf.org)