Operator Efficiency Improvement through Thinking 6 Sigma + Lean Problem SolvingApproach in IFX OP BE - A Case Study

Author: LIM Chun Pei, MOK Fock LinInfineon Technologies (Malaysia) Sdn. Bhd.

Category: 4. Manufacturing Technologies

Abstract

In this paper, Thinking 6 Sigma and Leanproblem solving techniques are adopted forthe solution development to improveoperator efficiency on Mal DS TSxP TapingProcess. The conventional Six Sigma toolsare replaced with lean tools to enhance theeffectiveness of the design, methodology,and approach for value and waste analysis.This newly developed approach hassystematically identified insufficient andpoorly defined operator job contents that arethe root cause for low personnel efficiency.The successful implementation of thistechnique has resulted in ~RM200k annualsaving without any unwanted implications onoverall machine performance, OEE andproduct quality.

This newly developed value and wasteanalysis approach has proven to be effectivefor personnel efficiency improvement atTSxP taping process. Also, it has opened upmany other opportunities at back-endmanufacturing environment to furtherstrengthen its competitiveness.

Introduction

Six Sigma, is a systematic problem solvingapproach originated from Motorola, focus onquality improvement through continuousreduction of process variation. The Sigma

(), is a Greek alphabet that represent

standard deviation in any process. Theprocess standard deviation of a Six Sigmaprocess level should be Six time smallerthan the range between process mean to theUpper or Lower Specification Limit. Resultingan estimated of 3.4 defect per million partsproduced, or equivalent to the process yieldof 99.99967% [2].

Six Sigma today, is a set of problem solvingtools that systematically applied five phasesof activities, namely Define, Measure,Analyze, Improve & Control (DMAIC) [1].Theproblem solving approach make known tothe world in 1988 after Motorola winning theBaldrige National Quality Award [3].

Lean Manufacturing System, also known asToyota Production System (TPS) that focuson customer value through continuouselimination of waste (Muda) throughout thewhole supply chain. Lean’s primary focus ison the customer, to address added and non-value added activities in the processoperation. The ultimate objective of creatingflow in lean manufacturing system is todeliver products and service just-in-time(JIT), in the right amounts, and the rightquality levels at the right place. A Six Sigmaprocess is necessary to enable the fulfillmentof Lean manufacturing concept. ValueStream Mapping (VSM), Standard Work, 5S,Single Minute Exchange of Die (SMED),Visual Management, Jidoka are the commontools applies to streamline the flow of LeanManufacturing System. [4] [5]

In the competitive semiconductor businessenvironment, only company deliver highestquality product and fast enough to adaptchange will survive. Integration of Six Sigma& Lean tools in problem solving become newapproach in the industries, whereby theclassical Six Sigma problem solving toolswere use to improve process quality byreduction of process variation, and LEANtools were use to focus on process efficiencyimprovement. Overall, the systematicDMAIC approach remains the key step in theprocess of problem solving.

Definition

Thinking Six Sigma + Lean (T6s+Lean) is anintegration of Six Sigma, Lean and 8Dproblem solving approach developed byInfineon. DMAIC remain the problem solvingsystematic in T6s+Lean. However, theapplication of problem analysis tools may bevaries depend on the case underinvestigation.

Personnel Efficiency (PE) as per equationbelow:

outputMonthly

hoursOThoursworkingMonthyPE

_

___

Where the PE value is lower, the better ofpersonnel efficiency is optimized

Project Objective

Main objective of this project is to study thefactors that causing low personnel efficiencyat TSxP taping line of MAL DS and developan improve solution through InfineonThinking 6 Sigma + Lean DMAIC problemsolving approach.

Methodology

Thinking Six Sigma + Lean (T6s+Lean) is anintegration of Six Sigma, 8D and Leanproblem solving approach. DMAIC remain asthe foundation of the problem solvingsystematic in T6s+Lean. However, theapplication of problem analysis tools may bevaries depend on the case underinvestigation as showed in Figure 1.

Figure 1 shows the suitable tools to bedeployed throughout DMAIC phase.

The Define phase is to ensure a clear, alignedunderstanding of the project topics and on theexpected outcome of the project. The Measurephase helps the team to identify potential sourcesfor the nonconformity in the process. In theAnalyze phase, the hypotheses are verified toconfirm that the real causes for problem. Also, toidentify respectively suitable levers to improvethe process performance. Purpose of phaseImprove is to develop and to tests solutions toachieve the targeted process performance.Finally, in Control phase, the process change isto be rolled out on a broader scale and to assessthe performance in a long term.

DEFINE PHASE

Objective of this phase is to describe the problembased on customers’ requirements and needs andto define the project scope and plan.

In this case, the object is Taping machine namedTHA and the non-conformity is unoptimizedheadcount efficiency where the man to machineration is 1:2.A final problem statement is clearlydefined by only facts: TSxP package run onTHA Taping machine with final PersonnelEfficiency of 2.54 which is above TCR target2.30 and current actual headcount is notoptimized where the man machine ration is only1:2 (Figure 2)

Figure 2: Object and Nonconformity areidentified with a final problem statement

A problem clarification consisting of 10elements is needed to provide sufficient

information and solid state to identify the

root cause and to define suitable solution,

e.g., The nonconformity was first noticed

since September 2011, with the potential

saving RM78,000 (Figure 3)

Figure 3: A problem clarification of 10 elements

SIPOC tool is used to provide a high level

process map to identify the business process

that needs to be improved. The five high

level process are: tested unit supply and

obtain F1 material; prepare lot traveler;

setup equipment and PC buyoff; machine

running and output send to incoming 100%

VI rack.

Next, Voice of Customer (VoC) is

undertaken to lead the team step by step to

a good understanding of what the customer

wants and needs and to identify the Critical

to Quality(CtQ)characteristics, which a key

attribute of products must be met to satisfy

the customer needs. As showed in Figure 4,

customer is DS Management and the

requirement is to improve the man to

machine ration from 1:2 to 1:3 so that the

personnel efficiency target could be achieved.

Figure 4: Voice of Customer (VoC )

The team concluded that there is no

containment action requires in this case as

the visual data analysis show the THA

Taping process is in control and show no

special cause in I-Chart as illustrated in

Figure 5.

Figure 5: Visual Data Analysis shows the

process is in control and no containment

action required

In the last step of the Define phase, the

team completed a project charter which

gives the Project Leader, project customers

a clear agreement on the project scope,

project schedule, and other relevant

elements. With the completion of the goals,

project can be continued to Measure phase.

MEASURE PHASE

Objective of this phase is to measure theactual process capability at the potentialsources of the nonconformity.

At Measure Phase, process mapping isemployed to identify the potential source ofthe low personnel efficiency at TSXPproduction. The sequence of the singleactivities performed in THA taping process isrecorded.

Figure 6: The result for the taping processmapping

In this case, VA/NVA & 7 Waste Analysis isused to identify and eliminate waste at thetaping process as showed in Figure 7.

Figure 7 details the VA/NVA analysis forthe TSXP taping.

The analysis indicated that some tasksinvolved transportation and motion waste.Operation Value Add (OVA) and Non ValueAdd (NVA) contributed more than 70%.These revealed that there are someimprovement opportunities at the tapingprocess. Operation Value Add (OVA) is theactivity that is not a value add tocustomer’s point of view but necessary foroperation to complete the process,example: Visual Inspection, StatisticalProcess Control (SPC)

As a result, activities such as scheduledJHAM, reels and pizza box transportation,manual pizza box folding, over-processingon lot traveler checking, manual lottraveler recording, equipment set-up,transportation on peel off forcemeasurement, transportation on test stripto quality check, visual inspection, motionon reel changed and wafer change assist,machine assist, transportation and motionon print material protocol andtransportation on good material submittedfor next process were categorized as nonvalue add activities.

After VA & NVA and seven waste analyses,the team used Cause and Effect Matrix asshowed in Figure 8 to help to find therelationship between process elementsand the nonconformity by giving a grade

from 1 to 10 on the relationship betweennon value add process elements and thenon-conformity in taping process. The topthree non-conformities identified are jobbalancing for operator is not optimized,inconsistent of process method, andabnormal process issues.

Figure 8: Cause & Effect Matrix

Pareto Analysis to identify and confirm thetop three possible causes for thenonconformity by 80/20 Rule. Withbrainstorming ideas from the team, aFishbone Diagram in Figure 9 is to identifythe potential root cause that contributes tothe high OVA & NVA contribution

Figure 9: Fishbone Diagram

The team later setup a data collection planwith the purpose to summarize the relevantinformation to ensure the right data iscollected and usable. Figure 10 shows thatcollection method, sample size required,data collection ownership, how is the datacollection to be displayed for each possibleroot causes that will be carried forward tothe Analyze phase.

Figure 10:Data Collection Plan

Produ Proce

Date August 10 2012 Rev 1 Page 1/1

No Operation Description Freq VA 7 Time Distance Responsible

Total Time

(min)

A. Workstation Check

1 Perform JHAM activities 1 OVA 7.5 7.5 0.0 Operator 1

2 Clean chuck table 2 OVA 4.0 8.0 0.0 Operator 1

3 Clean sealing 2 OVA 0.2 0.3 0.0 Operator 1

B. Material Preparation

4 Travel to wafer rack to get the sawn frame 0 NVA 0.0 0.0 0.0Line

Assistant

6Travel to cabinet to get the reel spool and

pizza boxes2 NVA T 1.1 2.1 8.0 Operator 1

7 Insert cover tape 2 VA 2.2 4.4 0.0 Operator 1

8 Insert carrier tape 2 VA 3.1 6.1 0.0 Operator 1

9 Insert Paper Tape (Bottom) 2 VA 2.8 5.6 0.0 Operator 1

10 Fold pizza boxes 18 NVA M 0.5 9.0 0.0 Operator 1

C. Lot Preparation

13check lot traveller and device, edikette

2 NVA M 0.3 0.5 0.0 Operator 1

14Load empty reel on the holder

2 VA 0.5 1.0 0.0 Operator 1

15Paste correct edikette on the empty pool and box

15 VA 1.0 15.0 0.0 Operator 1

16Fill In lot travellerlike device type, lotnumber, lot size, quantity , reel ID at 2 OVA 2.4 4.9 0.0 Operator 1

D. Equipment Setup

17 Setup parameter 2 VA 1.5 3.0 0.0 Operator 1

E. Equipment PC buyoff

Perform manual inspection undermicroscope

2 OVA 1.5 3.0 0.0 Operator 1

24Paste two units and used cotton stick onto

the recording book2 OVA T 0.5 1.0 110.0 Operator 1

F. Taping Operation

22 Insert frame on chuck table 2 VA 1.5 3.0 0.0 Operator 1

23 Load the leadframe into track 2 VA 2.0 4.0 0.0 Operator 1

24 Key in program and scan the edikette 2 VA 1.5 3.0 0.0 Operator 1

27check and confirm test yield and reject fallout

2 OVA 1.0 2.0 0.0 Operator 1

28 Reel Change Assist 2 OVA 1.3 2.5 0.0 Operator 1

29 Wafer Change Assist 2 OVA 1.5 3.0 0.0 Operator 1

30 Machine Assist - stoppages 24 NVA 1.3 30.0 0.0 Operator 1

31 Remove Frame from chuck table 2 VA 0.5 1.0 0.0 Operator 1

32 Remove foil from frame 2 VA 0.2 0.4 0.0 Operator 1

33 Pack the taped reel into the pizza box 15 VA 0.5 7.5 0.0 Operator 1

G. Post Taping Documentation

34 Walk and Print protocal result 2 NVA T 1.5 3.0 50.0 Operator 1

35Fill into lot traveller ( record the result,

collect all the reject unit into red plastic bag,2 OVA 4.5 9.0 0.0 Operator 1

37 Write down date, device, QTY tested into 2 OVA 1.5 3.0 0.0 Operator 1

38Ensure no stray unit and edikeet at

workstation - 5S2 OVA 1.5 3.0 0.0 Operator 1

H. Post Taping Material Logistic

39 Submit good material to next process 2 NVA T 5.0 10.0 180.0 Operator 1I. Communication

40Shift briefing time

1 OVA6 6 0

Operator 1

41Communication time between shift start & shift end

1 OVA3 3 0

Operator 1

Total 63.2 164.8 348.0

VA 15.8 54.0 32.99%

OVA 35.8 55.2 33.68%

NVA 9.6 54.6 33.33%

Total 61.2 164

Original Workflow

ANALYZE PHASE

Objective of this phase is to test the

hypothesis of the nonconformity by data

and facts.

At Analyze Phase, the three hypotheseswere evaluated based on the non-conformities identified at the previousstage. Figure 8 shows the threehypotheses developed and the summaryof findings.

Figure8: Hypothesis is developed

Hypothesis 1: High percentage of NVAactivities in Taping operation causing lowpersonnel efficiency.

Ho = NVA activities cause nosignificant impact to low personnelefficiency

H1: NVA activities cause significant impactto low personnel efficiency

Propositional test method used todetermine the p value. If p value smallerthan 0.05, this revealed that there is asignificant difference and the hypothesisaccepted. Based on the proportion testcarried out, p value obtained for highpercentage of NVA activities in tapingoperation which cause low personnelefficiency was 0.025 and smaller than 0.05,this two proportion test indicate NVAcontribution is significantly higher than theVA activity and therefore the NVA operatoractivities needed to be reduced.Therefore, H1 is accepted.

Hypothesis 2: Operator to operator manualactivities processing time is not consistentresulting low personnel efficiency

Ho: No significant difference betweenoperator manual activities processing time

H1: Significant difference between operatormanual activities processing time

Data collection method is 2 data samplepoint for each operator for each manualactivities processing time:

1. Do the manual inspection and sendto Q buy off

2. Perform peel force test3. Paste two units and used cotton stick

onto the recording book4. Pack processed spool5. Equipment Setup – Key in info

manually on computer display6. Obtain F1 material from F1 material

roomBased on Normality Test and I control Chart,each manual activities processing timeshow normal and stable in the process andcan be further proceed with one variabletest against IE MOST STD Time.Based on one variable test, p valueobtained for operator to operator manualprocessing time is not consistent resultinglow personnel efficiency was bigger than0.05. Therefore, Ho is accepted.

Hypothesis 3: High frequency of VisualInspection Processing Time causing lowpersonnel efficiency

Ho: High frequency of visual inspectionProcessing Time cause no significant impactto low personnel efficiency

H1: High frequency of visual inspectionProcessing Time cause significant impact tolow personnel efficiency

The frequency of visual inspection processbetween SOP (Standard of Procedure) andcurrent is the same.Therefore, it was concluded that highpersonnel of visual inspection processingtime causing low efficiency is not valid.

With three hypotheses tested, the teammove forward to Improve phase.

IMPROVE PHASE

Objective of this phase is to develop and totest solutions to achieve the targetedprocess performance.

At Improve phase, suggestion of regroupingand re-consolidating the value added andnon-value added of taping activities withthe idea of removing the NVA task from the

,;

operator was proposed to improve multi-machine operations. Compared to thecurrent state, the future state of processintroduced "machine operator and "butterflyoperator" concept by re-designing jobcontent of operator into two differentgroups of operators. “Machine operator” isresponsible for value add tasks such asmaterial preparation, e.g.: pizza box, reels,carrier tapes, attend machine stoppageswhereas the ‘butterfly operator’ isresponsible for non value add tasks such astransportation, quality check as illustratedin Figure 9

Figure 9: Consolidation of value add andnon value add tasks

Based on the new future state of processsteps, the workflow step for THA taping wasdivided into two separate parts. There arethe "machine operator" activities and"butterfly operator" activities. The non valueadded activities of carried out by themachine operator was reduced by31.5% andthe walking distance of operator wasreduced by 57%. By introducing butterflyoperator, it shares some of these non valueactivities which original own by the machineoperator. As a result, the machine operatorwas able to take-care additional equipmentsince the job load significant reducescompare to the initial state. The new taskallocation of machine operator and butterflyoperator are illustrated in Figure 10.

Figure 10: Improved process step byredesigning job content of operator andbutterfly operator

The trial run lasted for two months to ensurethat the changes of man to machine ratiofrom 1:2 to 1:4 and 1:16 brings no impactto Overall Equipment Effectiveness (OEE)and Mean Time To Repair (MTTR) .With thetotal number of operators to run fourteenTHA taping machines is optimized fromseven to five, fourteen THA Tapersperformance and operator availability torepair the machine process to be maintainedin normal and stable state.

Statistically, two sample T-Test show nosignificant difference of before and after ofMMO change, I control show the before andafter of MMO change is stale. Lastly, the boxplot also showed that there is no extremecompared before and after of MMO change.The result can be showed in Figure 11.

Figure11: Solution is verified throughseveral of statistic tools.

In conclusion, the result showed that changeof MMO does not bring negative impact toOEE and MTTR. With the successfulimplementation at Improve phase, the teammoves forward to Control phase.

CONTROL PHASE

The objective of this phase is to roll out theprocess improvement and assess the longterm performance for sustainability.

The solution from Improve phase wassuccessfully rolled out and gone live intaping production with some action itemneeded including formal shift briefing acrossall the shifts, documentation of change on IEMMO database. Besides this, long term

performance verification is monitored andthe result showed there is a significantpersonnel efficiency improvement from 2.54reduced to 1.94 which is achieving the TCRtarget of 2.30 as well (Figure 12)

Figure 12: Long Term PerformanceVerification of Personnel Efficiency Tracking

RESULT

With the successful project completion, theteam decided to review the business case.As showed in the table 1, the achievementof the project is showing double up of theproject target from RM 78,000 to RM174,000 with the man to machine ratio isimproved from 1:2 to 1:4 and the headcountrequired is improved from twenty one tofifteen daily. Also, the solution deploymentwas successfully rolled out in Wuxi DS andother sites of IFX.

Table 1: Compare project achievement VSproject target

CONCLUSION

This study applies the Thinking 6 Sigmaand Lean problem solving approach to lowpersonnel efficiency in TSxP packages onTHA taping process. Redesigning the jobcontent of operator by using the VA & NVAanalysis was able to improve headcountproductivity without implications ofmachine performance and product quality.Headcount productivity is improved andoperation cost is minimized. This

systematic approach can be applied byother thinking 6 sigma and leanpractitioners to conduct lean activities inother manufacturing sectors.

Acknowledgements

Special thanks to Mok Fock Lin, ChristianKeiling, Koh Chee Wei, Sim Cheng Whatt,Alan Ng Kam Choi, All TSxP TestingProduction Heads for their contribution onthis project.

Reference1. Thomas Pyzdek Dahlgaard, “The Six

Sigma Handbook” Mc Graw-HillR.2003.

2. Park, S.H.,2003 Six Sigma for qualityand productivity, Asian ProductivityOrganization.

3. Ronald D. Snee “Lean Six Sigma –getting better all the time”International Journal of Lean SixSigma. Vol 1 No. 1. 2010 pp 9 – 29.

4. Ali M. AL-Khouri. “Succeeding WithTransformational Initiatives: PracticalApproaches For Managing ChangePrograms” Management ResearchAnd Practice. Vol. 2 Issue 1 (2010) p:108-131.

5. Management Thinking, Issue 14.1998Drive Change or Cultivate It.

6. Christian Keiling, Whal & Wagner,“Thinking 6 Sigma Navigator”, EditionFeb 2012.