21st ASEMEP

National Technical Symposium

APPLICATION of WORK SAMPLING and ECRS (Eliminate, Combine, Re-lay out and Simplify)
PRINCIPLES of IMPROVEMENT at TO1 ASSEMBLY
Fritzie Ann A. Miranda
Manufacturing Department /Cleanroom, Section 1
SANYO Semiconductor Manufacturing Philippines Corporation (SSMP)
(An ON Semiconductor Company)
LIP-SEPZ, San Miguel, Tarlac City
Fritzie_ann_miranda @ sanyo.com
ABSTRACT
To continuously improve productivity, Improvement
Cycle was developed, starting from the process and work
content analysis down to the execution of process
improvement using various IE methodologies.

1.1 Problem objective

To come up with an improvement concerning manpower
productivity and process efficiency

The WORK SAMPLING Technique and ECRS

(Eliminate, Combine, Relay-out & Simplify) principle of
improvement was employed during the study in Transistor
Outline Package (TO) assembly that resulted to a 25 %
reduction of manpower allocation.

1.2 Problem statement

With the current set-up of machines, distance of end to end
machines are far from each other which makes it difficult
for both operators and inspectors during error assist,
product and material handling. The issue concerning the
distance travelled contributes to high man time.

1.0 INTRODUCTION

STK

Guided by the improvement cycle, each production line

was studied. One of the areas identified with improvement
opportunity is the Transistor Outline Package (TO1)
assembly line. The Line comprised of Die bonding and
Wire bonding machines. There are four (4) assigned
manpower responsible for Die bonding operation, Die
bonding on-line inspection, Wire bonding operation and
Wire bonding on-line inspection. See figure1.0.

WB016
UTC-50
TO7H

WB123
UTC-50
7H, 5H

DB023
STC-40
7H

WB141
UTC-50
7H, 5H

WB086
UTC-50
7H, 5H

STC-40
TO220
500 MIL

UTC-50
7H, 5H

WB042

STK

UTC-50
7H, 5H

WB016

DB026
STC-5
7H

PS01
2
DB035

DB019

STC-40
7H

EDP

STC-40
7H

DB016

WB086
UTC-50
7H, 5H

DB020
STC-40
TO220
500 MIL

WB140
UTC-50
7H, 5H

WB042
UTC-50
7H, 5H

TOTAL NO.
OF MAN FOR
3 SHIFTS

Die bonding
operator

Die bonding
on-line
Wire bonding
operator
Wire bonding
on-line
TOTAL

12

UTC-50
7H

UTC-50
7H, 5H

UTC-50
7H, 5H

NO. OF
MAN PER
SHIFT

WB142

WB041

WB141

STC-40
SMP5, 5H, 7H

WB079

UTC-345
7H. SMP5,
5H

DB037
STC-40
7H, 5H

STC-5
7H

DB037
STC-40
7H, 5H

WB033

UTC-345
7H, SMP5,
5H

WB122

UTC-50
7H. SMP5,
5H

DB025
STC-5
7H

STC-40
TO5H/7H

UTC-345
7H. SMP5,
5H

Machine B

The project is mainly focused at Transistor Outline

Package (TO1) assembly line. Contributing factors such as
man, machine, method and materials from the affected line
will be subjected to analysis.
2.0 EXPERIMENTAL SECTION

WB033

WB122
UTC-50
7H. SMP5,
5H

2.1 Materials and Equipments

Legend:
2.1.1 Production check sheets/lot b card
2.1.2 Stopwatch,
2.1.3 Time & motion worksheet.
2.1.4 Work instructions

Die bonding machines

DB025
STC-5
7H

-12 Wire bonding machines

ED P

STC-40
TO5H/7H

WB079

UTC-345
7H. SMP5,
5H

1.3 Problem scope

WB077

DB019

STC-40
SMP5, 5H, 7H

*Walk time from machine A to B: 11 seconds

UTC-345
7H, SMP5,
5H

-8
DB026

UTC-50
7H

WB077

FUNCTION
DB023

WB142

UTC-50
7H, 5H

WB140

UTC-50
7H, 5H

STC-40
7H

DB016

DB020

WB123

PS01
2
DB035

WB041

Fig.1.0 Transistor Package (TO) ASSEMBLY layout:

UTC-50
TO7H

Machine A

UTC-345
7H. SMP5,
5H

21st ASEMEP

National Technical Symposium

2.1.5 Die bond and Wire bond Machines

2.2 PROCEDURE

STANDARD TIME = Normal Time x {100/(100Allowance in Percent)}

2.2.1 TIME AND MOTION STUDY

2.2.2 WORK SAMPLING AND STRUCTURED
ESTIMATING

TO1 Die bonding and Wire bonding assembly process was

analyze into work elements to easily cater the application
of ECRS (Eliminate, Combine, Relay-out & Simplify)
principle. A time and motion study was conducted to
determine the standard time to do each job. Shown below
are the steps performed during the study:

After the standard time was determined, Work sampling

technique was used to easily discern how much time is
spend waiting for work, or performing paperwork tasks, or
even performing activities that are not included in their job
descriptions.

A. Making the time study

A. WORK SAMPLING PROCEDURE

The stop watch method was use in securing and recording
the operation and operator being studied.

The following steps used for WORK SAMPLING were

carried out:

B. Determining the number of observation

A.1 Establish the Purpose of the studythe objective of the study should be established.

A decision must be made as to the confidence level and the

desired accuracy that are to be used in determining the
number of observations to make. A 95% confidence level
and 5%precision are commonly used in the time study.

A.2 Identify the Subjectsthe people performing the task must be identified

N =

40

NX ) (X)
(X)
2

A.3 Identify the Measure of Output

the identification of the measure of the output produced or
the types of activities performed on the jobs being studied.

5% Precision

N =

20

NX ) (X)
(X)
2

A.4 Establish a Time PeriodStarting and stopping points for the study must be defined
as well.

10% Precision

Where:
N= actual number of observations of the element
X= each stop- watch reading or individual observation
= sum of individual readings

A.5 Define the Activitiesactivities that are performed by the people under study
must be define.

Codetermining Rating Factor and Normal Time

-Direct Productive Work- work that is focused on the

product and machine operations.

The WESTINGHOUSE system of rating was use to rate

the operators performance. (a) SKILL, (b) EFFORT,(c)
CONDITIONS, (d) CONSISTENCY.

-Indirect Productive Work-work that is complementary to

the product or machine operations.

-Non-Productive Work- actions that are not needed in the

product or lot.

Normal Time was measured to determine the time that a

qualified operator would need to perform the job if he
worked at a normal pace.

-Out of Work: operator is not within the area or line. They

are furthered classified as ASSIGNED (deployed at an
area) or UNKNOWN.
-Break time: Scheduled (Company Mandated Break time)
Personal

NORMAL TIME=Observed Average Time X rating

Factor
C. Determining Allowances and Standard Time

-Working Time = Direct and Indirect ProductiveWork and

Out of Work (Assigned).

Allowances were considered during the study depending

on the work content; such interruptions were classified as
follows: (1) Personal Allowance, (2) Fatigue Allowance,
or (3) Delay Allowance. STANDARD TIME computation
includes all elements in the operation and all necessary
allowances. Standard time is equal to normal time plus the
allowances.

-Idle Time = Out of Work (Unknown) and Breaktime.

21st ASEMEP

National Technical Symposium

A.6 Determine the Number of Observations Needed

Table 1:Table of Random Numbers

A.6.A Using the statistical method

A.6.2 Using the nomogram methodan easier way to determine sample size is to read off the
number of observations needed directly from a nomogram

A.8 Inform the Personnel Involved-the personnel

involved should be informed about the objective of the
study and the methodology that will be employed
A.9 Record the Raw Data
A.10 Summarize the Data-after the data collection, result
of the study must be summarized. Shown below is the
summary of the actual study conducted.

Work sampling result In terms of work category
Out of Area
( Unknown)
3%
Breaktime
4%

Out of Area
( Assigned)
2%

NonProductive
Work
11%

A.7 Schedule the Observations

Once the number of required observations has been
determined, the actual observations must be scheduled
using random number table .Refer to table shown:

Direct
Productive
Work
37%

Indirect
Productive
Work
43%

Analysis: In terms of type of work distribution, 43 % of

the proportion of time spent on indirect productive work
activities that are complementary only to the product or
machine operations.

21st ASEMEP

National Technical Symposium

Work sampling result Pareto of activities

A.ELIMINATE: Redundant inspection was applied for

removal. One task found to be with unnecessary inspection
is the after Change gold wire.

SUMMARY: PARETO OF ACTIVITIES

500

150%

450

130%

400

BEFORE

110%
90%

300
250

70%

200

50%

150

Symbol

%cumulative

frequency

350

30%

100

10%

50

-10%
Filling up of checksheet
Inspection
Loading of leadframes
Load and unload magazine
Product info and qty check
Machine cleaning/5S
Wait for Machine to Finish
Unknown
Speak Unnecessary
Printing of lot b cards
Get lot can and plate numbers
Error Resetting
EDP input
Setting up the machine
Get wafer lot
Get leadframe and validate
Unloading wafer lot
Minor machine adjustment
Filling up of lot card
Walk Empty
Silver paste preparation
Marking of defective units
Getting of indication board
Scheduled
Personal
Change/cleaning collet
Change paste
Assigned
Speak necessary
Cleaning dispense nozzle
Arrange stocks
Transportation
Process Turnover

Work elements
Bond 1 pellet
Machine monitor /microscope inspect ion
Bond 1 leadframe
Microscope inspection using machine
Normal sampling inspection using machine

BEFORE

operator

Utilizing the result of work sampling, the ECRS approach

was use for the improvements.

Is it possible to change
the timing?

Normal sampling inspection using machine

BEFORE AFTER DIFFERENCE

3
2
1
4
3
1

B.COMBINE:
The operation and inspection function were merged.
Operators assigned at Transistor Package (TO) assembly
were cross trained to on-line inspection. Also same
inspection items were combined.
.

2.2.3 APPLICATION OF ECRS

ReRe-lay out

Microscope inspection using machine

-Operation
-Inspection

Analysis: It was found out that the average utilization rate

of operators is 82% only.
Source: Work sampling and structured estimating -Internet

Is it possible to
Eliminate a specific
work?

Bond 5 leadframes

Summary:

82%
18%

Eliminate

Microscope inspection using machine

AVERAGE
UTILIZATION RATE
OF OPERATORS
IDLE

Bond 1 leadframe

AFTER

Computed Utilization rate of operators

WORKING

Machine monitor /microscope inspect ion

Symbol

Analysis: Filling up of check sheets and inspection were

the top activities observed.

Work elements
Bond 1 pellet

inspector

AFTER

Operator
& inspector

C.RELAY OUT:
An efficient lay out is vital to achieved a smooth
production flow. Transistor package assembly was
subjected to major lay out from a scattered type (mixed
machines) to Clustered type (machines grouped per
process).The main objective of machine re-lay out is to be
able to reduced walk time of operators and to reduce the
man machine assignment.

Combine

Is it possible to integrate
or separate?

Simplify

Shorten the distance.

Lessen the load.
Is it possible to simplify?

STK

Before

After

WB016
UTC-50
T O7H

WB123
UTC-50
7H, 5H

DB023
ST C-40
7H

PS01
2
DB035
STC-40
7H

WB142

UTC-50
7H

WB041
UTC-50
7H, 5H

DB016
WB141
UTC-50
7H, 5H

Each work elements were analyzed for the possibility of

ECRS application. However the top activities from the
work sampling result were prioritized. Shown below are
samples of the improvements done per improvement type:

WB086
UTC-50
7H, 5H

DB020
ST C-40
TO220
500 MIL

WB140
UTC-50
7H, 5H

WB042
UTC-50
7H, 5H

DB026
STC-5
7H

STC-40
SMP5, 5H, 7H

WB079

UT C-345
7H. SMP5,
5H

Wire
bonding

Die
bonding

DB037
STC-40
7H, 5H

WB033

UTC-345
7H, SMP5,
5H

WB122

UTC-50
7H. SMP5,
5H

Die
bonding

Wire
bonding

DB025
STC-5
7H

WB077

DB019
EDP

ST C-40
TO5H/7H

UTC-345
7H. SMP5,
5H

Scattered type
lay out

Clustered type
lay out

21st ASEMEP

National Technical Symposium

D.SIMPLIFY
To lessen the load of operators the following items were
done:
D.1 Machines with manual loading of frames and wafers
were replaced to automatic machines.

Table.3 Number of machine to assign:

Function

Number of
machines
Operator 1
8
Operator 2
6
Operator 3
6

MANUAL LOADING --

 AUTOMATIC

* The number of machine is computed using the

prescriptive symbolic model of multiple activity chart.

D.2 Since it was found out that most of the time the
operator spends on check sheets filling up, automation
of daily production record was implemented.

3.0 RESULTS AND DISCUSSIONS

The application of Work sampling and ECRS principle
yielded the following results aligned with our company
policy of:

Manual

Automation

3.1 HIGH QUALITY

3.1.1Cases of swapped lots will be avoided since machines
are clustered according to process. No case of swapped lots
(Oct09-to present.)

x
2.2.3 MAN AND MACHINE COMPUTATION
From the determined standard time and after the
implementation of improvements, man machine
requirement was computed. See figure below:
Table 2: Man requirement:
Funct ion
Die b on d in g
op erato r
Die b on d in g o n lin e
Wi re b on d in g
op erato r
Wi re b on d in g
on -lin e
T OT AL

Actual Man Ma n need D ifference

deployed
(computed)
1

0 .9 5

0.0 5

0 .6 3

0.3 7

0 .5 8

0.4 3

0 .8 5

0.1 5

3 .0 1

3.1.2 Due to merged machines, immediate assist on errors

is expected. Defect produced by errors will be
minimized.

* The man requirement is computed using standard

time, total time to operate and the time available per
day.

Before

After

Time per error assist 12.3 secs 5.7 secs

Analysis:
Computation shows that there is a difference of 1 man
from the actual manpower deployed.

21st ASEMEP

National Technical Symposium

3.1.3 Early detection of on-line inspector is expected since

machines are group together per process.

3.2.2 Machines with manual loading of frames and wafer

were line out from the line as shown in Table 4.
Table 4

Die
bonding

.
Loading of leadframes
Loading of wafer
Average

Wire
bonding

BEFORE
(Manual)

AFTER
(Automatic)

SAVINGS
(mins)

100%
100%
100%

80%
79%
80%

20%
21%
21%

80% Decreased in
assist time/lot

3.2 HIGH SPEED

3.3 LOW COST

3.2.1Applying the clustered type of lay out, walk time of

operators from one machine to another during error assist
will be reduced. Also the distance traveled by On-line
inspectors will be lessen. Table A shows the time saved
per job function.

3.3.1 Clustered type of lay out with improvements on

work elements resulted to reduction of man assignment.
BEFORE

STK

Manpower assignment
12
9
Php 1,091,988 Php 1,455,955
Manpower cost

WB016
UTC-50
TO7H

WB123
UTC-50
7H, 5H

DB023
STC-40
7H

AFTER

PS01
2
DB035
STC-40
7H

3.3.2 Increase in multi skilled rate by 27%.

WB142

UTC-50
7H

WB041
UTC-50
7H, 5H

DB016
WB141
UTC-50
7H, 5H

WB086
UTC-50
7H, 5H

DB020
STC-40
TO220
500 MIL

WB140
UTC-50
7H, 5H

WB042
UTC-50
7H, 5H

DB026

Multi skilled rate

WB079

63%

100%

SAVINGS
/year
27%

DB037
STC-40
7H, 5H

3.3 .3 Thru the implementation of automated daily

production record cost of paper and printing was
eliminated. Thus the company saves money.

WB033

UTC-345
7H, SMP5,
5H

WB122

UTC-50
7H. SMP5,
5H

DB025
STC-5
7H

EDP

3.4 OTHERS
After proving the effectiveness of Work Sampling and
ECRS principle of improvement, this methodology was
implemented to other companys production process. The
implementation contributes a good result on the attainment
of companys goal.

UTC-345
7H. SMP5,
5H

Before

After

Table 3.

Function
Die bonding
operator
Die bonding onWire bonding
operator
Wire bonding onAverage

AFTER

UTC-345
7H. SMP5,
5H

WB077

STC-40
TO5H/7H

BEFORE

STC-40
SMP5, 5H, 7H

STC-5
7H

DB019

SAVINGS
/year
3
2.69

BEFORE

AFTER

100%

57%

SAVINGS
(mins)
43%

100%
100%

69%
61%

31%
39%

100%

61%

39%

100%

62%

38%

4.0 CONCLUSION
The application of ECRS Principle is therefore effective to
increase man efficiency thereby reducing manpower cost.

62% Decreased
in walk time/lot

21st ASEMEP

National Technical Symposium

5.0 RECOMMENDATIONS
The use of IE methodology such as Work sampling and
ECRS fundamentals are best recommended to any
manufacturing work stations to improve productivity and
cost.
`
6.0 ACKNOWLEDGMENT
This project shall not be realized without the help of the
following who shared their expertise, time and ideas
unreservedly. Thus, my sincere thanks to all of them!

 Manufacturing shift in-charge for ensuring proper

implementation of the new system
 Die bonding and Wire bonding operators, inspectors
and Technicians for complying the set standard
 Edwin Buscas and Jinky Andres for providing the
training needs
 Arnel Hulipas and Dennis Ayson for the support on
relay out activities
 Louie Dizon & Jason Capati for the unending
support in time, resources and management
intervention for the realization of the project
7.0 REFERENCES
1. Time and Motion study- Author: Ralph Barnes
2. Work sampling and structured estimating-internet
8.0 ABOUT THE AUTHOR

Fritzie Ann A. Miranda is a graduate of BS Industrial

Engineering from Saint Louis University Baguio City.
She has been with SSMP since June 2001 and currently
assigned as Manufacturing Process Engineer of Process
Improvement from Dicing to Wire bonding process.