Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 30

Statistical Analysis of

Operations at an Automobile
Manufacturing Line
Purpose
 Interpretation through Graphical Representation
of Production data
 Descriptive Statistical Analysis to compute the
mean , median , mode of the ungrouped data and
grouped data
 Studying the Normal Distribution with the help
of data.
 Hypothesis testing for Production Cycle time
 Regression Analysis of the Production Figures as
dependent variables against independent
elements.
 Forecasting of Production Figures and
comparison them with the actual figures.
The Data
 The model details of each vehicle produced from 1st April
2017 to 31st Oct 2018 with roll out time.
 The Total Down time for Production Line reported in
minutes each day for the said period.
 The down time further divided into two categories:
 Equipment/Machine Downtime
 Operational Loss(Man, Material, Method)
 The Equipment Down time is further distributed over type
of equipment.
Graphical Representations and
Interpretation
Production Share of Models
D
0%
A
4%
B
19%

C
77%
Variant Distribution and Production
Share
60 59
43
40
Numbers

18
20
2
0
A B Model C D

A41 B5 B11
14% 9% 8%

A40 C8
6% 31%
B13 C14
A34 23% 30%
A39 18%
12%
B3
30%
A35
10%
A36 B17
18% B2 8%
6%
Production Volume : Month wise
30000 30

Number of Working Days


25000 25

20000 20
Vehicles

15000 15

10000 10

5000 5

0 0
Vehicle
Produced
Number of
Working Days

Total Line down time (in Min)


30000 9000
8000
25000
7000
20000 6000
Vehicles

5000
15000
4000
10000 3000
2000
5000
1000
0 0

Vehicle Produced

Total Line Down


Time( In minutes)
9000
Down time : Month wise
8000
7000
Down Time in Minutes

6000
5000
4000
3000
2000
1000 Equipment/Machinery Loss

0
Operational Losses

100%
90%
80%
70%
60%
50%
Equipment/Machinery Loss
40%
30% Operational Losses
20%
10%
0%
Contribution of Equipment Types in Down
time
8000
Electrical Control Systems Power Supply Systems
7000 Cooling systems Sealing Systems
Door Machines Conveyors and Body Transfer Systems
6000 Jigs and Fixtures Welding Equipments
5000
Minutes

4000

3000

2000

1000

2500 100%
Monthly Average(Min)

90%
2000 80%
70%
1500 60%
50%
1000 40%
30%
500 20%
10%
0 0%
Welding Jigs and Conveyors andDoor Machines Sealing Cooling Power Supply Electrical
Equipments Fixtures Body Transfer Systems systems Systems Control
Systems Systems
Descriptive Statistical Analysis
Vehicles Produced per day
Mean 1,013
Standard Error 4
Median 1,031
Mode 1,032
Standard Deviation 90
Sample Variance 8,164
Kurtosis 4
Skewness (2)
Range 594
Minimum 588
Maximum 1,182
Sum 447,661
Count 442
Production Figures per day : Distribution
100

80
Frequency

60

40

20

Production Figures
Down time in Minutes
Total Downtime Equipment Downtime Operations Downtime

Mean 173 Mean 127 Mean 46

Standard Error 5 Standard Error 5 Standard Error 2

Median 140 Median 92 Median 37

Mode 112 Mode 74 Mode 25


Standard
Standard Deviation 102 Standard Deviation 97 Deviation 36

Sample Variance 10,415 Sample Variance 9,338 Sample Variance 1,329

Kurtosis 4 Kurtosis 5 Kurtosis 6

Skewness 2 Skewness 2 Skewness 2

Range 641 Range 618 Range 245

Minimum 34 Minimum 22 Minimum -

Maximum 675 Maximum 640 Maximum 245

Sum 76,604 Sum 56,064 Sum 20,540

Count 442 Count 442 Count 442


Down Time per day : Distribution
250
200 Total Downtime per day

Frequency
150
100
50
0

250
Equipment Losses per day 200
200 Operational Losses per day
Frequency

150
150 Frequency
100 100

50 50
0 0
Normal Distribution
Study
 As per Central Limit Theorem, the Sample Means of the samples (n>30)
taken from Production Figures must be normally distributed with

&

 We took around 410 Sample means with each sample size of n=38 using
RANDBETWEEN formula in Excel
120

Frequency 100
80
60
40
20
0

Sample Means

Sample Means Population Mean


Mean 1013.081 Mean 1,013
14.7597 Also
same as
Standard Deviation 90/Sqrt(38) Standard Deviation 90
Hypothesis Testing for
Production Cycle time
 It has been stated that Managers have done improvement activities in the
month of March 2018 to reduce down the production cycle time for the
next fiscal year over previous fiscal.

 In order to prove it , first average cycle time is calculated from April 2017 to
Feb 2018 (Population Mean).
Total
Loss Due
Total Line working
Produced to Operations Total Line
Down Time Time(16 Tact time Tact time
Vehicles( in Equipment Losses ( In Running
Reported ( hours=960 in minutes in Seconds
numbers) ( In minutes) Time
In minutes) minutes a
minutes)
day)
T=(960-
P L E O 960 =960-L L)/P Ts=T*60
 The average of Tact Time(Ts) is taken from 1st April 2017 to 28th February
2018 and similarly Standard Deviation is calculated for the population

Tact Time from 1st April 2017 -28th February 2018

Mean 45.78
Standard Deviation 6.65
 Now Hypothesis testing is performed taking any random sample (n=35)from 1st
April 2018 to 31st October 2018 ( a finite Population N= 165).
Tact time in
Date Sno. Seconds
Random Number Random Sample
=VLOOKUP(D2,$B$2:$C$166,2,FALSE
4/2/2018 1 47.8203928905519 =RANDBETWEEN(1,165) )
=VLOOKUP(D3,$B$2:$C$166,2,FALSE
4/3/2018 2 48.0428954423593 =RANDBETWEEN(1,165) )
=VLOOKUP(D4,$B$2:$C$166,2,FALSE
4/4/2018 3 56.1800219538968 =RANDBETWEEN(1,165) )
=VLOOKUP(D5,$B$2:$C$166,2,FALSE
4/5/2018 4 47.5862068965517 =RANDBETWEEN(1,165) )
=VLOOKUP(D6,$B$2:$C$166,2,FALSE
4/6/2018 5 47.4832535885167 =RANDBETWEEN(1,165) )
 The test is one tail .
H0 µ=45.78
Ha µ<45.78

 Since sample size (n=35)> 5% of (N=165)


 Considering α =0.05 , the critical z-Value is 1.65

 Observed value for the sample


Random Sample
Mean 47.64167678
Zobserved 0.7635

 The above values fall in the non- rejection region , hence fail to reject the
null hypothesis. The improvement activities has not resulted in cycle time
reduction so far
Regression Analysis
Production Figures per day as dependent variable ( Y) and Equipment Loss
(X1) & Operation Losses (X2) as independent variables.

 Correlation :

Produced Equipment Operations


Vehicles( in Losses ( In Losses ( In
numbers)(Y) minutes)(X1) minutes)(X2)

Produced Vehicles( in
1
numbers)(Y)

Equipment Losses (
-0.482670049 1
In minutes)(X1)

Operations Losses (
-0.200591534 -0.035622219 1
In minutes)(X2)
Multiple Regression Model
Regression Statistics
Multiple R 0.53
R Square 0.28
Adjusted R Square 0.28
Standard Error 76.82
Observations 442.00

ANOVA

Significa
df SS MS F nce F

Regression 2.00 1,009,801.99 504,901.00 85.56 0.00

Residual 439.00 2,590,704.66 5,901.38

Total 441.00 3,600,506.65

Standard Lower Lower Upper


Coefficients Error t Stat P-value 95% Upper 95% 95.0% 95.0%

Intercept 1,096.10 7.73 141.73 - 1,080.90 1,111.30 1,080.90 1,111.30


Equipment Losses ( In
minutes) (0.46) 0.04 (12.11) 0.00 (0.53) (0.38) (0.53) (0.38)
Operations Losses ( In
minutes) (0.54) 0.10 (5.38) 0.00 (0.74) (0.34) (0.74) (0.34)

Y=1096.10- 0.46*X1- 0.54*X2


 The model explains only roughly 28% of variability. Also SSE is very high as compared to SST.
 This implies there are other explanatory variables as well for Production Figures
Forecast of Production Figures for
November 2018 using Exponential
Smoothing Method
Forecast of Production Figures for November 2018 using
Exponential Smoothing Method

Ft+1 = αXt+ (1-α) Ft

α =0.301 for minimum error in this case


Month Number of Working Days Vehicle Produced Daily Average Figures ExSM Err %Err

April-17 22 23922 1,087

May-17 26 21455 825 1,087

June-17 19 19160 1,008 1,008 0.03 0%

July-17 24 23817 992 1,008 16.07 2%

August-17 25 25118 1,005 1,004 1.11 0%

September-17 24 23587 983 1,004 21.15 2%

October-17 22 21279 967 998 30.35 3%

November-17 24 24514 1,021 988 32.98 3%

December-17 19 19633 1,033 998 34.95 4%

January-18 23 23867 1,038 1,009 28.81 3%

February-18 24 25765 1,074 1,018 55.98 6%

March-18 25 26888 1,076 1,034 41.11 4%

April-18 24 24932 1,039 1,047 7.95 1%

May-18 27 28323 1,049 1,044 4.61 0%

June-18 19 19807 1,042 1,046 3.30 0%

July-18 26 26324 1,012 1,045 32.32 3%

August-18 24 25531 1,064 1,035 28.74 3%

September-18 23 22483 978 1,044 66.18 6%

October-18 22 21256 966 1,024 57.60 6%

Nov-18 1,006
Thanks

You might also like