I declare that in submitting all work for this assessment I have read, understood

and agree to the content and expectations of the assessment declaration.

Course/Unit code Assignment Assignment Group/Session name (if applicable)

number due date

5pm Friday
OMGT2199 1 17th April, Wednesday 8am class
2020

Course/Unit name Program title

Operation Management

Lecturer/Teacher’s name Tutor / Marker’s name (if applicable)

Dr Reza Akbari

This statement should be completed and signed by the student(s) participating in preparation of the
assignment.

Declaration and statement of authorship:

1. I/we hold a copy of this assignment, which can be produced if the original is lost/damaged.
2. This assignment is my/our original work and no part of it has been copied from any other student’s work or from any
other source except where due acknowledgment is made.
3. No part of this assignment has been written for me/us by any other person except where such collaboration has been
authorised by the lecturer/teacher concerned and is clearly acknowledged in the assignment.
4. I/we have not previously submitted or currently submitting this work for any other course/unit.
5. This work may be reproduced and/or communicated for the purpose of detecting plagiarism.
 6. I/we give permission for a copy of my/our marked work to be retained by the School for review by external
examiners.
7. I/we understand that plagiarism is the presentation of the work, idea or creation of another person as though it is
your own. It is a form of cheating and is a very serious academic offence that may lead to expulsion from the
University. Plagiarised material can be drawn from, and presented in, written, graphic and visual form, including
electronic data, and oral presentations. Plagiarism occurs when the origin of the material used is not appropriately
cited.
8. Enabling plagiarism is the act of assisting or allowing another person to plagiarise or to copy your work.

Family name Given name Student number Student signature Date

Monday 13th
Pham Tuan S3695313
April, 2020

Further information relating to the penalties for plagiarism, which range from a notation on your student file to expulsion from
the University, is contained in Regulation 6.1.1 ‘Student Discipline’ www.rmit.edu.au/browse;ID=11jgnnjgg70y and Academic
Policy: ‘Plagiarism’ www.rmit.edu.au/browse;ID=sg4yfqzod48g1.

Assessor’s comments Grade School date stamp

(Office use only)

I. ZenoFex Case Study

 Perio MA WMA
d Semiconductor (n=4) ES (n=4) t*y t^2 LT
1 48       48 1 48.53
48.00
2 55     110 4 52.99
50.10
3 59     177 9 57.46
4 61   52.77   244 16 61.92
55.24
5 66 55.75 57.9 330 25 66.39
58.47
6 70 60.25 62 420 36 70.85
61.93
7 73 64 65.9 511 49 75.32
65.25
8 78 67.5 69.5 624 64 79.78
69.07
9 86 71.75 73.7 774 81 84.25
74.15
10 89 76.75 79.4 890 100 88.71
78.61
11 93 81.5 84.3 1023 121 93.18
82.92
12 99 86.5 88.9 1188 144 97.64
Sum 78 877       6339 650  

Simple moving average

Simple moving average is usually utilized to assess the present level of a time series, with this
figure being anticipated as a forecast for future observations (Hansun 2013).

n
∑ A t−i
Ft =MAn = i=1
n
where
F t =Forecast for time period t
MA n=n period moving average
At−1 = Actual value in period t−1
n= Number of periods in the moving average

In this case, n=4 so we will start forecasting from the period 5 and they will be calculated as
below:

F5 = MA4 = (A1 + A2 + A3 + A4)/4 = (48 + 55 + 59 + 61)/4 = 55.75

F6 = MA5 = (A2 + A3 + A4 + A5)/4 = (55 + 59 + 61 +66)/4 = 60.25

F7 = MA6 = (A3 + A4 + A5 + A6)/4 = (59 + 61 +66 +70)/4 = 64

The remaining periods are similarly calculated and shown in the table above.

Weighted Moving Average

Weighted Moving Average is frequently utilized for smoothing irregular changes in a time series
to allow the data analyst to identify the trend over time (Perry 2010).

F t =w t ( A t )+w t−1 ( A t−1 )+. ..+wt−n ( A t−n )

where
w t =weight for period t , wt−1=weight for period t−1, etc .
At =the actual value for period t , A t−1 =the actual value for period t−1, etc .
in above formula 0 ≤ w ≤ 1,
∑ wt =1
In this case, n=4 so we will start forecasting from the period 5. The figures .4 .3 .2 .1 are used
respectively from the most recent periods.

F5 = WMA4 = 0.1A1 + 0.2A2 + 0.3A3 + 0.4A4 = 0.1 x 48 + 0.2 x 55 + 0.3 x 59 + 0.4 x 61 = 57.9

F6 = WMA5 = 0.1A2 + 0.2A3 + 0.3A4 + 0.4A5 = 0.1 x 55 + 0.2 x 59 + 0.3 x 61 +0.4 x 66 = 62

F7 = WMA6 = 0.1A3 + 0.2A4 + 0.3A5 + 0.4A6 = 0.1 x 59 + 0.2 x 61 +0.3 x 66 +0.4 x 70 = 65.9

The remaining periods are similarly calculated and shown in the table above.

Exponential smoothing

Exponential smoothing forecast gives bigger weights to later observations, and the weights
decline exponentially as the observations become progressively distant (Ostertagova & Ostertag
2012).
 Ft =F t−1 +α ( A t−1−F t−1 )
where
F t =Forecast for period t
Ft−1 =Forecast for the previous period
α =Smoothing constant
A t−1=Actual demand or sales from the previous period
0≤ α ≤1

In this case, the forecasted value of period 2 is the actual value of period 1, which is 48. The
following periods will be calculated as below, given alpha = 0.3:

F3 = 48 + 0.3*(55-48) = 50.1

F4 = 50.1 + 0.3*(59 – 50.1) = 52.77

F5 = 52.77 + 0.3*(61 – 52.77) = 55.24

The remaining periods are similarly calculated and shown in the table above.

Linear trend

n ∑ ty−∑ t ∑ y
b= 2
Ft =a+bt n ∑ t 2−( ∑ t )
where ∑ y−b ∑ t
F t =Forecast for period t a= or y−b t
n
a=Value of Ft at t=0 where
b=Slope of the line n=Number of periods
t=Specified number of time periods from t=0 y =Value of the time series

n ∑ ty−∑ t ∑ y
b= 2
n ∑ t 2 −( ∑ t ) = 4.465

a=
∑ y −b ∑ t
n = 44.061

 y= 4.465x + 44.061

Forecasting methods comparison:

 Forecasting Methods Comparison
120

100
f(x) = 4.47 x + 44.06
80

60

40

20

0
1 2 3 4 5 6 7 8 9 10 11 12

Semiconductor Linear (Semiconductor)

MA (n=4) ES
WMA (n=4)

As it can be seen from the above graph, the Linear Trend line is the closest line to the
Semiconductor line while the Exponential Smoothing line is the furthest line to the
Semiconductor line. In other words, LT method is the most suitable one to predict the demand
while ES method is the least accurate one in this case.

MA WMA
  (n=4) ES (n=4) LT
MAD 11.25 12.04 9.05 1.16
MSE 147.57 170.25 96.20 2.00
MAPE 13.82 15.79 11.12 1.67

Forecasting Methods Comparison

180.00
160.00
140.00
120.00
100.00
80.00
60.00
40.00
20.00
0.00
MA (n=4) ES WMA (n=4) LT

MAD MSE MAPE

MAD, MSE and MAPE can be used as a tool to put different forecasting methods into
comparison. For all three measures, smaller values often show a more appropriate trend
(Klimberg et al. 2010).

It can be seen from the table that LT has the smallest values in all three types of measurement.
Therefore, LT should be utilized to forecast the demand for this product. On the other hand, ES
is the least appropriate as it has the biggest figures in all three types of measurement.

II. FORTech Inc. Case Study

1)

Scott obtained 20 random samples of 15 pieces each. So, in this case, it can be stated that the
sample size (n) will be 15 and number of samples (k) will be 20.

Sampl
e Mean Range UCL LCL
1 45.01 0.85 45.172 44.786
2 44.99 0.89 45.172 44.786
3 45.02 0.86 45.172 44.786
4 45 0.91 45.172 44.786
5 45.04 0.87 45.172 44.786
6 44.82 0.9 45.172 44.786
 7 44.79 0.86 45.172 44.786
8 45.04 0.89 45.172 44.786
9 45 0.85 45.172 44.786
10 44.79 0.9 45.172 44.786
11 44.97 0.91 45.172 44.786
12 45.11 0.84 45.172 44.786
13 44.96 0.87 45.172 44.786
14 45 0.86 45.172 44.786
15 44.92 0.89 45.172 44.786
16 45.06 0.87 45.172 44.786
17 44.94 0.86 45.172 44.786
18 45 0.85 45.172 44.786
19 45.03 0.88 45.172 44.786
20 45.09 0.91 45.172 44.786
         
  X Bar R Bar    
  44.979 0.876    

First set of samples

X bar 44.979
R bar 0.876
Standard
Deviation 0.187
n 15
k 20
A2 (n=15) 0.22
Process
Capability 1.373
UCL 45.172
LCL 44.786

Applying the data into the formula to find 𝜎:

𝐴2𝑅̅ ≈ 4 𝜎 / √n
=> 𝜎 ≈ (𝐴2𝑅̅ * √n) / 4

=>𝜎 ≈ 0.187

The process standard deviation, 𝜎, is calculated, so next we will calculate the capability of the
process

Cp = (UTL-LTL)/ 6𝜎 = 1.54 / (6* 0.187) = 1.373

=> As it can be seen, the Capability index (1.373) is bigger than 1.33, so it can be concluded that
the process is capable.

2)

Sampl
e Mean Range UCL LCL
1 45.01 0.85 45.172 44.786
2 44.99 0.89 45.172 44.786
3 45.02 0.86 45.172 44.786
4 45 0.91 45.172 44.786
5 45.04 0.87 45.172 44.786
6 44.82 0.9 45.172 44.786
7 44.79 0.86 45.172 44.786
8 45.04 0.89 45.172 44.786
9 45 0.85 45.172 44.786
10 44.79 0.9 45.172 44.786
11 44.97 0.91 45.172 44.786
12 45.11 0.84 45.172 44.786
13 44.96 0.87 45.172 44.786
14 45 0.86 45.172 44.786
15 44.92 0.89 45.172 44.786
16 45.06 0.87 45.172 44.786
17 44.94 0.86 45.172 44.786
18 45 0.85 45.172 44.786
19 45.03 0.88 45.172 44.786
20 45.09 0.91 45.172 44.786
         
  X Bar R Bar    
  44.979 0.876    
 45.3
X bar Control Chart (First set of samples)
45.2
45.11 45.09
45.1 45.04 45.04 45.06
45.01 44.99 45.02 45 45.03
45 45 45
45 44.97 44.96
44.92 44.94
44.9
44.82
44.79 44.79
44.8
44.7
44.6
44.5
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Mean UCL LCL

Sampl
e Mean Range UCL LCL
1 44.96 0.42 45.151 44.851
2 44.98 0.39 45.151 44.851
3 44.96 0.41 45.151 44.851
4 44.97 0.37 45.151 44.851
5 45.02 0.39 45.151 44.851
6 45.03 0.4 45.151 44.851
7 45.04 0.39 45.151 44.851
8 45.02 0.42 45.151 44.851
9 45.08 0.38 45.151 44.851
10 45.12 0.4 45.151 44.851
11 45.07 0.41 45.151 44.851
12 45.02 0.38 45.151 44.851
13 45.01 0.41 45.151 44.851
14 44.98 0.4 45.151 44.851
15 45 0.39 45.151 44.851
16 44.95 0.41 45.151 44.851
17 44.94 0.43 45.151 44.851
18 44.94 0.45 45.151 44.851
19 44.87 0.38 45.151 44.851
20 44.95 0.41 45.151 44.851
21 44.93 0.39 45.151 44.851
22 44.96 0.41 45.151 44.851
23 44.99 0.4 45.151 44.851
24 45 0.44 45.151 44.851
25 45.03 0.42 45.151 44.851
26 45.04 0.38 45.151 44.851
27 45.03 0.49 45.151 44.851
 28 45.14 0.41 45.151 44.851
         
  X Bar R Bar    
  45.001 0.406    

X bar Control Chart (Second set of samples)

45.2
45.15 45.14
45.12
45.1 45.08 45.07
45.04 45.0345.0445.03
45.05 45.0245.03 45.02 45.0245.01
45 44.99 45
45 44.98 44.97 44.98
44.96 44.96 44.9544.9444.94 44.95 44.96
44.95 44.93
44.9 44.87
44.85
44.8
44.75
44.7
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

Mean UCL LCL

● Putting the first and second set of samples into comparison, the first one has a lower
number of samples, which leads to a less stable trend of X bar value. The difference of
the biggest figure and the smallest one of the first set of samples is 0.32.

● Whereas, the second set of samples has more samples. Therefore, the trend of X bar value
of it seems to be more stable. The difference of the biggest figure and the smallest one of
the second set of samples is 0.27, which is smaller than that of the first set of samples.

As it can be seen in two X-bar graphs above, clearly the second set of samples can illustrate the
cycling of the process. To be explicit, in the first graph, the figures spread out from the centre
line with an enormous range which is considerably more varied. This is due to when loads of
data gathered, it can distinguish even little, insignificant changes in the process. In this way, a
null hypothesis showing the data from a regularly distributed population is dismissed. Therefore,
enormous data set might demonstrate bogus alerts as it can misrepresent some modest changes in
some cases. For example, it can make a few figures seem to be out of control while they are still
in and the trend is difficult to accurately determined. Whereas, in the second graph, the data
points are engaged to the mean and less scattering. Consequently, the general trend is simpler to
be distinguished.

3)

Assuming the problem can be found and corrected, the standard deviation (σ) will reduce.
Consequently, it will precisely enhance the process capacity (Cp) and the process will be more
capable as:

specification width
Cp=
6. σ

The professor recommended acquiring another set of samples, this time utilizing smaller sample
size and taking more samples. Jane then consulted with Scott and they concurred that he would
take 28 samples of eight observations each. So, in this case, it can be stated that the sample size
(n) will be 8 and number of samples (k) will be 28.

Sampl
e Mean Range UCL LCL
1 44.96 0.42 45.151 44.851
2 44.98 0.39 45.151 44.851
3 44.96 0.41 45.151 44.851
4 44.97 0.37 45.151 44.851
5 45.02 0.39 45.151 44.851
6 45.03 0.4 45.151 44.851
7 45.04 0.39 45.151 44.851
8 45.02 0.42 45.151 44.851
9 45.08 0.38 45.151 44.851
10 45.12 0.4 45.151 44.851
11 45.07 0.41 45.151 44.851
12 45.02 0.38 45.151 44.851
13 45.01 0.41 45.151 44.851
14 44.98 0.4 45.151 44.851
15 45 0.39 45.151 44.851
 16 44.95 0.41 45.151 44.851
17 44.94 0.43 45.151 44.851
18 44.94 0.45 45.151 44.851
19 44.87 0.38 45.151 44.851
20 44.95 0.41 45.151 44.851
21 44.93 0.39 45.151 44.851
22 44.96 0.41 45.151 44.851
23 44.99 0.4 45.151 44.851
24 45 0.44 45.151 44.851
25 45.03 0.42 45.151 44.851
26 45.04 0.38 45.151 44.851
27 45.03 0.49 45.151 44.851
28 45.14 0.41 45.151 44.851
         
  X Bar R Bar    
  45.001 0.406    

Second set of samples

X bar 45.001
R bar 0.406
Standard
Deviation 0.106
n 8
k 28
A2 (n=8) 0.37
Process
Capability 2.421
UCL 45.151
LCL 44.851

Applying the data into the formula to find 𝜎:

𝐴2𝑅̅ ≈ 4 𝜎 / √n

=> 𝜎 ≈ (𝐴2𝑅̅ * √n) / 4

=>𝜎 ≈ 0.106

The process standard deviation, 𝜎, is calculated, so next we will calculate the capability of the
process

Cp = (UTL-LTL)/ 6𝜎 = 1.54 / (6* 0.106) = 2.421

=> As it can be seen, the Capability index of the second set of samples is 2.421, which is much
bigger than that of the first set of samples. It can be concluded that the second set of samples has
a more capable process compared to the first set of samples.

4)

Previously, it seemed that smaller sample size can reveal something that the large one might not.
However, it does not always work for every case. Large sample size implies that more data is
gathered. Though the standard deviation does not diminish as sample size expands, the standard
error does decline when the sample size expands (Barde & Barde 2012). As that being said, more
information will be acquired and uncertainty will be diminished. Therefore, the output will be
progressively solid and exact. Additionally, the size of the example depends on what precision
level conductors would like to have. In some uncommon cases, when doing a survey for
propelling a new product which requires high precision level, the conductor will calculate the
expense of inspection and the cost of obtaining a sample (Wetherill 2013). In the event that
inspection cost is small, even if the cost to take a sample is higher, using large sample size will
be a good choice.
References

Barde, M.P & Barde, P.J 2012, ‘What to use to express the variability of data: Standard deviation
or standard error of mean?’, Perspectives in clinical research, vol. 3, no. 3, p. 113.

Klimberg, R.K, Sillup, G.P, Boyle, K.J & Tavva, V 2010, ‘Forecasting performance measures—
What are their practical meaning?’, Advances in business and management forecasting, vol. 7,
pp. 137-147.

Ostertagova, E & Ostertag, O 2012, ‘Forecasting using simple exponential smoothing method’,
Acta Electrotechnica et Informatica, vol. 12, no. 3, pp. 62.

Perry, M.B 2010, ‘The Weighted Moving Average Technique’, Wiley Encyclopedia of
Operations Research and Management Science.

Wetherill, G.B 2013, ‘Sampling inspection and quality control’, Springer, vol. 129.

Hansun, S 2013, ‘A new approach of moving average method in time series analysis’, In 2013
conference on new media studies (CoNMedia) (pp. 1-4), IEEE.