Vous êtes sur la page 1sur 19

FINDING OUT THE MOST ACCURATE

METHOD FOR OPERATOR EVALUATION


Table of Contents
Abstract ......................................................................................................................................................... 3
Objective ....................................................................................................................................................... 3
Research paper under study ......................................................................................................................... 4
HYPOTHESIS .................................................................................................................................................. 4
Alternate Hypothesis ................................................................................................................................ 4
Null hypothesis ......................................................................................................................................... 4
Background behind the study ....................................................................................................................... 4
Objectives of developing and maintaining operator skill matrix .............................................................. 5
Methodology.................................................................................................. Error! Bookmark not defined.
Method of developing Skill Matrix............................................................................................................ 6
LINE BALANCING ....................................................................................................................................... 6
Allocation of Operator's (Line Balancing) ..................................................................................................... 7
Using linear regression analysis .................................................................................................................... 8
Experimental Data ........................................................................................................................................ 9
MODEL 1 ................................................................................................................................................... 9
Without using any skill matrix .................................................................................................................. 9
MODEL 2 ................................................................................................................................................. 11
Operator allocation using skill matrix .................................................................................................... 11
Results and discussion ................................................................................................................................ 14
FOR MODEL 1 .......................................................................................................................................... 14
Output of linear regression analysis ....................................................................................................... 14
Scatter diagram ....................................................................................................................................... 15
For 2nd model .......................................................................................................................................... 16
Output of LINEAR REGRESSION ANALYSIS ............................................................................................ 16
Scatter diagram ....................................................................................................................................... 17
Conclusion ................................................................................................................................................... 18
Abstract
Due to fast changing of styles, there is very little time available for the production
process, so the optimum uses of resources especially the available machinery &
operators who are adding value of the products is quite important.

The project deals with finding out the most accurate method for operator
evaluation . 1st part of project consist of creating a database of skills of workers by
time study using & also by PCSMOE system developed by Dr. Prabir jana & Dr.
Deepak Panghal , to understand the strength & weakness of each operator .
In 2nd part of project , workers will be allocated in the line using traditional
way(without using skill matrix), then by using time study skill matrix & third by
using PCSMOE system & then output of the workers will be compared to find out
the best operator evaluation method

Objective

Creating skill matrix using Time study method and also by PCSMOE system.

Allocating workers in line using following 3 models:-

1. Traditional method ( without any skill matrix)


2. By using time study skill matrix
3. By using PCSMOE skill matrix

Comparing efficiency and output of each operator to find out the most accurate
method for operator evaluation.
Research paper under study

HYPOTHESIS
Alternate Hypothesis
Line balancing using skill matrix gives more output in comparison to traditional
method

Null hypothesis
line balancing using skill matrix does not increase output in comparison to
traditional method

Background behind the study


During our study of the process of line setting before starting of a new style in
production line. We found that , at the time of line setting, operators are selected
based on their experience on operations by the supervisors which we found was
very random as calculated skill level of the operators on the operations is not
considered at all. As a result after couple of hours, high skilled operators start
sitting idle and low skilled operators stuck with their work. Thus line becomes
imbalanced and lot of productive time is lost as operators sit idle. To utilize
operators maximum capacity, work allocation must be done based on operators
potential performance level (efficiency) and work must be shared with operators
who has excess capacity.

To get a balanced line from day one, select operators for the operations based on
operators skill levels and capacity required for the task. So we needed to develop
a skill matrix of all operators.

Objectives of developing and maintaining operator skill matrix


1. To check operators skills in running different machines (in efficiency
percentage)

2. To check operators capability of doing various operations (If you maintain


operations wise record, you may skip recording machine record or keep
both) and their performance level (efficiency %)

3. To know list of operators who can do selected operations and their


performance level
Methodology

developing Skill Matrix

Develop template for data recording:


We will be using excel sheet for developing skill matrix. (Reason, Gaurav
International Pvt. Ltd. doesnt have software for keeping skill matrix records.)

To have all basic information in one table, you need to capture following data.

Date of data entry,

Operator name and employee code,

Operation name,

Machine name and

Operator efficiency on the operation

Evaluating workers by using PCSMOE system


Evaluating workers by using time study
Development of the skill database using above 2 methods & comparing them
to find out the most accurate method.
Allocating operators in the line to find out the impact of operator evaluation
above.

LINE BALANCING
Line Balancing is the distribution of work on the line in such a way that everyone
gets the same output of work in terms of time.
Objectives of line Balancing

To minimize the no. of workstation on line.


To minimize the cycle times.
To maximize the workload smoothness.
To minimize the frequency of tools change.
To maximize the efficiency & effectiveness

Allocation of Operator's (Line Balancing)

Appropriate allocation of the skill operators for particular operations is the


keystone for any line balancing in the garment industry. Allocation also depends
on type of balance required. One approach of allocation is to find the closest
match between operators required and operators available. This type of
allocation results in intrinsic balance of line. Another approach is the effective
utilization of operator's skill so far he can do the best, this approach results in
dynamic balance of line.

OPERATOR PERFORMANCE REQUIRED WITH RESPECTIVE OPERATIONS


OPERATION SAM MACHINE TARGET OPERATORPERFOEMANCE
PIECES REQUIRED (%)
Collar making 1.13 SNLS 34 63
Cuff making 1.20 SNLS 68 114

Table 1
FROM SKILL INVENTORY OPERATOR PERFORMANCE AVAILABLE ARE
OPERATOR A B C D
COLLAR 66 75 72 48
MAKING
CUFF MAKING 85 89 130 65
Table 2

Dynamic balancing

From the Tables 1 and 2 for a setting the operators are selected as B & C because
the operational efficiencies of these operators are as per the required for the
particular jobs which will have the effective use of all operators' skills.

Using linear regression analysis

A linear regression model attempts to explain the relationship between two or


more variables using a straight line. Linear regression is a basic and commonly
used type of predictive analysis

When you are conducting a regression analysis with one independent variable,
the regression equation is Y = a + b*X where Y is the dependent variable, X is the
independent variable, a is the constant (or intercept), and b is the slope of the
regression line.

For example in this case


Linear Regression analysis will done on all 3 models to find out following:-

How will a given model fits data.


Analyzing the data using a scatter plot.
The linear regression gives us equation of the best fit line.
The line represents the trend of the data so we can make predictions.

Experimental Data

Operational Bulletin of Jacket

Capacity study

MODEL 1

Without using any skill matrix


Process Name Name M/C Performed Allocated Output @ Efficiency Output @
Type SMV SMV 100% off % performed
efficiency %

Back part panel Jarna O/L 0.64 0.62 96.77 96 93.26


join

Back panel ts tc Naher F/L 0.35 0.3 20.00 87 173.08

Left and right Hasina O/L 1.72 0.62 96.77 36 34.95


panel join

Left and right Eima O/L 1.28 0.62 96.77 48 46.88


panel join
Process Name Name M/C Performed Allocated Output @ Efficiency Output @
Type SMV SMV 100% off % performed
efficiency %

Left and right Mena F/L 0.62 0.55 109.09 89 97.30


panel ts and tc

Front part panel Ronju O/L 0.78 0.43 120.00 64 76.60


join and tc

Front part panel ts Monzilla L/S 0.53 0.5 200.00 57 113.92


and tc

Left and right Amina O/L 0.82 0.3 96.77 776 73.47
panel join

Left and right Sahina O/L 0.79 0.62 96.77 78 75.95


panel join

Left and right Nurbanu F/L 0.53 0.62 109.09 103 112.50
panel ts and tc

Shoulder panel Lima O/L 0.94 0.55 139.53 46 63.60


join

Shoulder panel ts Alima F/L 0.88 0.43 142.86 48 68.18

Shoulder join Rojina O/L 0.92 0.42 101.69 64 65.45

Sleeve panel make Momina O/L 0.64 0.45 133.33 70 93.26

Sleeve panel ts and Sokina F/L 0.47 0.49 122.45 104 126.76
tc

Sleeve hem Mormina F/L 0.42 0.3 200.00 71 141.73

Collar make Pervin L/S 0.79 0.63 95.24 80 75.95

Collar twill tape Rebeka L/S 0.26 0.22 272.73 85 230.77


make

Collar serving Roksana O/L 0.43 0.39 153.85 90 138.46

Zipper twill tape Sabina L/S 0.88 0.42 142.86 48 68.44


make

Zipper twill tape Rojina L/S 0.87 0.5 120.00 57 68.97


attach

Zipper holding Bobita L/S 0.66 0.3 200.00 45 90.91


tuck
Process Name Name M/C Performed Allocated Output @ Efficiency Output @
Type SMV SMV 100% off % performed
efficiency %

Zipper patch Shahanara L/S 0.45 0.57 105.26 127 133.33


attach at bottom

Collar join Sabina L/S 1.77 0.78 76.92 44 33.90

Collar join Shahila L/S 1.99 0.78 76.92 39 30.15

Zipper tuck with Naher L/S 0.36 0.42 142.86 118 168.22
body

Zipper join with Halima L/S 0.58 0.5 120.00 86 103.45


left side

Zipper join right Orchona L/S 0.76 0.69 86.96 90 78.60


side

Zipper facing join Beauty L/S 0.66 0.5 120.00 76 90.91

Zipper facing o/l Sharmin O/L 0.36 0.3 200.00 83 165.14

Max Theoretical Output 76.92 Practical 30 pieces


Output

MODEL 2

Operator allocation using skill matrix

Process Name Name M/C Type Performed Efficiency % Output @


SMV performed
efficiency

Back part panel join Ronju O/L 0.64 48 46.45

Back panel ts tc Alima F/L 0.35 48 96.00

Left and right panel join Jarna O/L 1.72 96 92.90


Process Name Name M/C Type Performed Efficiency % Output @
SMV performed
efficiency

Left and right panel join Roksana O/L 1.28 90 87.10

Left and right panel ts and tc Nurbanu F/L 0.62 71 77.45

Front part panel join and tc Rojina O/L 0.78 64 76.80

Front part panel ts and tc Sabina L/S 0.53 44 88.00

Left and right panel join Amina O/L 0.82 76 73.55

Left and right panel join Morina O/L 0.79 70 67.74

Left and right panel ts and tc Mena F/L 0.53 87 94.91

Shoulder panel join Sharmin O/L 0.94 83 115.81

Shoulder panel ts Sokina F/L 0.88 89 127.14

Shoulder join Sahina O/L 0.92 78 79.32

Sleeve panel make Elima O/L 0.64 64 85.33

Sleeve panel ts and tc Naher F/L 0.47 103 126.12

Sleeve hem Momina F/L 0.42 104 208.00

Collar make Rebeka L/S 0.79 86 81.90

Collar twill tape make Shahila L/S 0.26 39 106.36

Collar serving Lima O/L 0.43 46 70.77

Zipper twill tape make Orchona L/S 0.88 48 68.57

Zipper twill tape attach Halima L/S 0.87 57 68.40

Zipper holding tuck Rojina L/S 0.66 45 90.00


Process Name Name M/C Type Performed Efficiency % Output @
SMV performed
efficiency

Zipper patch attach at bottom Bobita L/S 0.45 85 89.47

Collar join Naher L/S 1.77 118 90.77

Collar join Shahanara L/S 1.99 127 97.69

Zipper tuck with body Sabina L/S 0.36 57 81.43

Zipper join with left side Monzilla L/S 0.58 76 91.20

Zipper join right side Pervin L/S 0.76 90 78.26

Zipper facing join Beauty L/S 0.66 80 96.00

Zipper facing o/l Hasina O/L 0.36 36 72.00


Results and discussion

FOR MODEL 1

Output of linear regression analysis

F significance is too less

P value is less
than 0.05%
Scatter diagram

TREND LINE EQAUTION FOR MODEL 1


y = 0.324x + 41.42
For 2nd model

Output of LINEAR REGRESSION ANALYSIS

F signifiance is 2% which basically


means there is only a 2% chance
that the Regression output was
merely a chance occurrence.

P value is less than 0.05


This means this
intercept is correx
approx 99.99 %
Scatter diagram

TREND LINE EQAUTION FOR MODEL II


y = 0.470x + 56.70
Conclusion

Model I II
Trendline y = 0.324x + 41.42 y = 0.470x +
equation 56.70

For X=1 41.744 57.17


For x=2 42.068 57.64
For x=3 42.932 58.11

For every x , y is always greater in Model II in comparison to model I


So output will we always greater in model II

We found out that MODEL II has higher output than Model I which proves our
null hypothesis is wrong and our Alternate hypothesis is correct i.e Line balancing
using skill matrix produces more output in comparison to traditional method.

Also it proves, Skill matrix helps in allocating right person for the right job which
helps in achieving desired performance level. It keeps record of all operations an
operator had done in the past and efficiency level in each operation. Engineers /
line supervisors need minimum time to find and select most efficient operators
for an operation from the pull of operators.
For line balancing, operators can be selected according to work content. When
someone is absent, supervisor can easily find suitable person from the skill matrix
table and replace. To analyses the skill availability and distribution throughout the
factory. This can be compared with the skill requirement for a particular time
period and shortage/excess skill availability to achieve at the training
requirement. So productivity can be achieved by allocating skill & semi-skilled
workers to the right place and unskilled operator should be trained properly.
Also after line is set using skill matrix ,capacity study is done to further find
bottlenecks in the line . the main job is to reduce/eliminate WIP between
operations to improve the output.

Similarly ,MODEL III will be tested in our Graduation project

Vous aimerez peut-être aussi