XYZ LTD appliances division is flagship company of Godrej group. Godrej is pioneer company in consumer market for refrigerators & washing mcs, besides other various products.
XYZ LTD appliances are having 2 manufacturing set up; Mohali, Punjab & Shirwal, Maharashtra Shirwal Works details are as follows:
Location ~ Shirwal, Maharashtra
On the Golden quadrangle – Highway to the south of India.
200 kms from Mumbai l
50 kms from Pune
Land Area: 55 Acres (2.2 million sq feet)
Built up area: 11 Acres (0.44 million sq feet)
Main plant: 300,000 sq. ft.
Utility Bldg: 15,000 sq. ft
Admin Bldg: 35,000 sq. ft
Warehouses: 90,000 sq. ft
Washing machine started in 1996
Injection mouldjng facilities started in 1996
Refrigerator manufacturing started in 2005
Products (Total more than 100 SkUs)
Direct Cool Refrigerators
Frost Free Refrigerators
Washing Machine: Semi Auto & Automatic
Refrigerators: 0.8 million (3 lines)
Washing machines: 0.25 million (1 line)
110 white collar staff
350 blue collar staff
casual labour as and when required
Machines and Equipments
Injection Molding shop: ltaltech /Windsor make
18 Machines from 100 T to 1600 T capacity
Auto Material feeding through PIVON system ~
Ultrasonic Welding Machines — Stackmann make
Powder coating shop:
ITW — GEMMA powder coating booths
Colour Changeover time of 10 mins t
Machines and Equipments
Q5 — ltaly make linear vacuum forming machines
Asano- Japan make rotary vacuum forming machines
Door foaming and Cabinet foaming:
Cannon make foaming set — up
Gas charging and sealing:
Galileo charging machines
Lock ring technology — Vulkan Germany
Shirwal Other / ﬁnished goods storage Set Up
FG Warehouses – 65,000 sq feet of warehousing space. 25000 appliances can be stocked.
Input power —- 33kv express feeder line giving continuous power
DG back up power available
ETP & STP Plants — all the treated water is reused for gardening.
Other speciﬁc environment /Initiatives /efforts
Producer gas unit installed successfully for powder coating ovens; ﬁrst of its type in India (Reducing carbon
Plantation on all empty land
Introduced 5 star refrigerators series first in India
DSS score — safety champion — for assuring safety
Six sigma project for reducing warranty cost of washing machines (Mechanical assembly area)
During the Business review, Cost of poor quality – warranty cost of the various products were discussed; Warranty cost of washer was raised as a point of concern.
Total Washer warranty cost was observed to be Rs 98/washer which top management had given stretch targetto reduce by 50 % .(Mech assembly Rs _32 to Rs 10)
Six sigma methodology was applied to achieve targets:
Define: Reduce warranty cost of washer mechanical assembly from Rs. 32 to Rs 10 /washer.
CFT was formed & found from service data main contributors for defects as:
Water level sensor
After Pareto analysis, it was observed that the mechanical assembly contributes maximum in warranty cost.
Reasons reported from field for failures of mech assembly were:
2.Noise due to slippage
CFT studied the existing process & documented the present status.
Cp of present process was very low 0.55 & also 100 % inspection done to sort out defective components. Run out of cup A set observed to be in between 300 — 600 microns _& rejection of mech assembly is 5%.
After brain storming, root cause analysis was done & observed two processes needed specific improvements,
- Caulking process
- Slippage checking process.
- Caulking Process:
Caulking is a process in which Gear case & Case B — two components are joined & seals the gear assembly.
Single point Ca ulking: Single vertical rotating cam pressed against case B; with friction of rotating cam Gear
case gets locked in Case B.
Multiple Point Caulking:- Two Rotating caulking /forming rollers forces aluminum gearcase on Case b to form
Multiple caulking mc made operative & measured results ;proved better than single point caulking. (cp
.89).Started using multiple point caulking mc.
Though multiple point caulking mc capability improved over the present caulking, still to further improve, DOE
conducted to improve capability more than 1.5 to further reduce the runout in turn reduce vibration.
Variable parameters (X’s) decided & again full factor two level experiment conducted to get better results; Cp is nearer to 1.9.
With the improved runout(O-150 microns ) vibration of mc whenymeasured found less than 1mm at spinningRPM
of 900 at front portion with full load capacity.
Though the process is now in control ; we monitor the process by implementing control chart & training operators for maintaining the same; new process parameters updated in quality plan.
Other improvements kaizens: ’
- Slippage checking procedure improvement:
Slippage force initially was checked by attribute method; now new method with torque wrench with speciﬁed
torque is assured.
- Hydraulic holding used to reduce operator fatigue.
Turbobrush Mech assembly
Reducing warranty cost of washing mc due to mechanical assembly from Rs 32 to Rs 10.
In our business review meeting at HO, chaired by our C00, cost of poor quality (warranty cost) is regularly reviewed amongst other operational parameters. Post doing a benchmark study within the industry, we learn that our washing mc warranty cost was higher than the best in the industry.
Hence in one of the meeting, it was agreed upon to take up the warranty cost reduction project on priority.
Godrej is facing ﬁerce competition in appliance market in India. Godrej is competing with Big multinational giants like Samsung, LG, Whirlpool who are not only having a technological advantage but being big MNCs, also have a very strong financial backup and access to the world market.
Functional quality of pro duct is no more a competitive advantage, IT IS A MUST. In all consumer durable markets, its only the aesthetics and product features that differentiate the product. Reliability and quality has to be the best and at par with industry or we cannot su rvive in the market.
Due to competitive market situations all companies are under a tremendous pressure to deliver maximum value to the consumer at a very competitive price. A higher warranty cost not only impacts the bottom line of the company but it also erodes its brand value in the market and affects its customer loyalty.
In fully automatic washing mcs category ,when analysed service calls it was observed that Vibrations contributes maximum & in that mechanical assembly replacements are maximum .50 we had taken project to work upon mechanical assembly warranty cost reduction.
When analysed total service calls it is observed that 45% of total calls in auto mc are for vibrations.
When vibrations measured as compared with competitors mcs ,competitors mc‘s vibration was found to be less than
1.5mm at front portion at 300 mm from top which is 2.5mm to 3.00 mm for our mcs though mcs functionality is not
affected. Service technician’s feedback is also taken as they always complain that Godrej washers are having more vibrations as compared to competitors; ( customer feedback to them ).
About 5 % calls observed from service call data is for slippage] noise .
Though customer is not knowledgeable & aware about slippage but it results in heavy noise & dislocating CUP A SET from its original position ;thus causing non functioning of washing mc .
when Cup A set assembly checked with torque meter found Torque at which this slips varies from 400 kgcm to 900 kgcm; The defective mechanical assembly which are returned from service found slippage torque 50 kgcm to 250 kgcm only.
Above two points are taken as CT (1 as these are very important from customer point of view.
Following two points are taken as priority CT Q for project as it contributes maximum in warranty cost.
2. Slippage /noise
Other CT Q ‘s considered are:
2. Functional problems of various components in washing mc
The washing mc warranty cost comprises of failures due to various elements in year 2008-9 which is follows:
- Vibrations (due to Mechanical assembly) Rs 32
- Operational 8: Operating defects (due to Control card), Rs 25
- Electrical Failures (Motor), Rs 12
- Components Failures (lnlet valves, water level sensors, gear motor) R515
- Damages (plastic components / cabinets) 15 ‘
For above ,service material warranty cost data is taken as reference.
CFT formed studied the process AS IS & documented the present status.
Though the ﬁeld report suggests max service calls for vibration; it is perceived feeling of customer. ‘
So actual mc vibrations are measured on line & compared the same with competitor.
Data is collected about service calls first indicating vibrations as measure concern followed by slippage or noise.
As no standards specification available on measurement of vibration it was decided to measure the vibration of
competitor’s mc & then compare the same with our machines.
Comparative study was done with 2 competitor’s mcs & values measured.
To measure vibrations of our mcs it was decided to measure 8 readings daily after every one hour for 8 days.
Above data collection is adopted to randomise the readings Readings are taken at random to avoid bias in data collection;
(Biasness: To avoid biasness in data & get exact representation of population, samples were randomly selected.)
Gauge R & R study was carried out for measuring the Vibrations of auto washing MC on front portion of cabinet &
found that AS the GRR is within 10%; Present measurement system is acceptable for vibration measurement
Data stratification, and sampling plans:
Data stratiﬁcation is done by operator wise /shift wise /hour wise.
Sampling plan: Vibrations of 1 mc measured every hour for seven days.
Data is collected in below format: Readings are in attached excel sheet, Sr No Date.Operator Shift Time Vibration
Vibrations are measured in mm of displacements No of washing mcs failed against standard/specification are measured in % & in DPMO. Also the Warranty cost of washers are measure of present base line performance (service data).
DPMO For Washers including with criteria of vibration (max 1.5 mm): 8823 Warranty cost /washer: Rs 32/ washer(due
to meclwassembly). ‘
ROOT CAUSE DIAGNOSTICS
Fish bone diagram is prepared in Minitab for Y = vibration; attached in excel sheet. Major X’s are mechanical assembly,
spin basket, tub, leveling, suspension set & cabinet.
Pareto analysis is done for service data to prioritize
When a nalyzed 100 auto mcs ﬁeld failure data following are the ﬁndings:
With Pa reto analysis, mechanical assembly is maximum contributing factor in vibrations
Analysis of Vibrations field complaints: 100 me analysis:
Data is again collected to further analyze the root cause of failure of mechanical assembly.
Pareto chart is used to further analyze the Defects in mechanical assembly:
|Defect||actual failures||% failures|
|High R un out||62||81.6|
|CSS pressing distance variation||6||7.9|
|Case B & css Undersiie||2||2.6|
Cross fu nctional team meet & suspect main reason of vibration source attributed to RUN OUT of mechanical assembly.
Detailed why why analysis is done; attached in excel sheet.
Immediate corrective actions taken & kaizen projects implemented to take benefit of low hanging fruits; some Kaizens initiated are;
1.Slip page checking mechanism by torque wrench instead ofjudgment & feel
2.Low Cost Automation for slippage checking in house
3.Go no gauge for vendor developed to assure uniform height of bottom support
4.To take half shot in lnjection Molding Department to ensure uniform flow of materials at 2 hour intervals
5.As Cft had decided to study effect of one way clutch ,data is collected for 50 readings for vibration first with one way clutch & then without one way clutch. With T test results were validated & as P va|ue(0.975 ) is more thansignificant value Nullhypothesis is fail to reject (no difference due to One way clutch.).During this trials as one way clutch was removed, the drum rotates both way in turn better friction between clothes & better wash-ability.
When this shared with design & marketing ,this phenomena is again validated at R& D & we could introduce new mechanical assembly design not only by adding additional feature for the product but also reduce BOM cost by Rs 35 /washing mc.
Minitab session results
Two-sample T for vibration with one way clutch vs vibration without one way clutch
|vib with one way||60||2.378||0.636||0.082|
|vib without one||60||2.375||0.623||0.080|
Difference = mu (vibration with one way clutch) – mu (vibration without one way clutch) ‘
Estimate for difference : 0.003333
95% Cl for difference : (-0.224344, 0.231011)
T-Test of difference = 0 (vs not =) : T-Value = 0.03 P-Value = 0.977 DF = 118
After above kaizen improvements CFT again meet; Brain storm & suspect main reason of vibration source attribute to RUN OUT of mechanical assembly.
Cross functional team brainstormed & with the experience decided to narrow down vibration affecting parameters from 6 to minimum possible; during this experiment other factors are kept constant.
Full factorial (3 factor 2 way) experiments conducted;
Factors & its 2 levels are as follows:
|Mechanical assembly run out||0.3||mm|
|Spin basket run out||2.5||mm|
|Cabinet diagonal distance||3||mm|
Results of full factorial experiment as follows:
As P value is 0 for MA; major contributing factor is mechanical assembly. minitab session in excel sheets.
5 = 0.306186 R-Sq = 97.92% R-Sq(adj) = 96.35% With DOE as is clearly observed, P value for Mechanical assembly runout is 0; clearly indicating most important parameter to work upon. Though P value is 0.01 for spin baskets, kaizens implemented in basket manufacturing process had improved the accuracy levels of spin baskets.
As with technical expertise & experience, CFT attributed Main reason of vibrations to Mechanical assembly run out; So team had decided to work on Run out of mechanical assembly.
The root cause identified by DOE was validated by Regression analysis:
When the validation is carried out Run out against vibration as Rsq adjusted is 95.9 %; it is clear that 95.9%times variations in vibrations can be explained by variations in Run out.
The regression equation is
Vibrations = 0.0745 + 0.00466 Runout
S = 0.135375 R—Sq = 96.0% R—Sq(adj) = 95.9%
CFT brainstormed on following altemative solutions,
- Improve present caulking mc to improve Run out
- Make Two rollers caulking mc (unused till now ) operative & tryout for improvement in Run out
- Try out altemate process for caulking
Detailed evaluation matrix prepared & conclusion was as follows
(As concluded by CFT team ):
- Single point caulking mc is constraint from design point of view
- Purchase altemative mc which is costly affair: instead it is better to work upon advantages of Two point caulking mc with
less investment & try to obtain better results with experimentation method ).
CFT had agreed & decided to work on Two rollers caulking process; with less investment & also decided to conduct experiments for ﬁne tuning the results on this two point caulking mc.
when CFT worked on project ofimproving Run out of 2 rollers caulking mc ;process parameters on Two rollers caulking mc were decided by full factorial, 4 factor & 2 levelexperiments with help of MlNl’l’AB.
After conducting the DOE on two rollers caulking mc RESULTS are checked & found improvement in results .
with DOE on two rollers caulking mc optimum parameters set as follows:
- Rollers angle : 35 degree
- Rollers speed : 600 mm
- Booster pressure : 5 kg/sq cm
- Vertical travel : 52 mm
With above though the Tolerance limit made half of original (0 to 500 microns initially now 0 -300 microns)resu|ts are very consistent & Run out average reduced to 143 micron which was more than 400 micronsbefore improvement.
Validation whether solution works:
- Ho: Both process ( Before single point caulking & after — two rollers caulking) are having same variances
- Ha: Both process ( Before single point caulking & after- two rollers caulking) are not having same variances
Test statistic = 18.35, p—va|ue = 0.000
AS P value is less than the signiﬁcance level, reject the null hypothesis & accept the alternate hypothesis
i.e. Both variance are not equal.
Test statistic = 18.35, p—value = 0.000
AS P value is less than the signiﬁcance level, reject the null hypothesis & accept the altemate hypothesis
i.e. Both variance are not equal.
Validation by T Test:
- – Ho: Both means are same
- Ha: Both means are not same
AS P value is less than the significance level, reject the null hypothesis & accept the alternate hypothesis.
Both mean & variance improved with new process
When the better results achieved continuously & consistently for 3 months; SOP prepared & other operators were given training about the change process.
There was some resistance initially from operators but when the importance of the project was explained & also the impact of the same on our washing mc business; the resistance reduced.
Also no human beings like rejections & reworks, so when some of them were also become project team members, they also contributed in finding solution; implementing different alternatives & also initiated kaizens. Person working in Mech assembly, Mr Ganesh Bobade was awarded with best kaizen award for Slip checking mechanism, low cost automation manufactured inhouse.
SUSTENANCE OF GAINS
Operators were asked to monitor the performance & also asked to maintain the control chart to gauge the process continuously.
As only one washing mc factory of Godrej similar operations can not be duplicated but when Some vendors from China visited our factory; they appreciated our Foolproof Slippage checking methodology; they also video shoot our operation to show in their factory.Also similar vibration issue is observed in compressor; so similar analysis & project carried out for compressor also.
- Warranty cost due to mechanical assembly reduced from Rs 32 to Rs 8.50/washer .
Above project contributed Rs 23.5 /washer reduction (Total Rs 58.75 lacs warranty cost saving ).
- Also internal rejection cost reduced from Rs 4/washer to Rs 1 /washer for mech assembly in turn saved Rs. 7.5 lacs/annum.
- Due to introduction of new mechanical assembly BOM saving /washer is Rs 35, For 50 k washers of this type total saving achieved is Rs 17.5
lacs Other Beneﬁt
Company could introduce new model without any investmerit with additional feature of Turbobrush (both way drum
rotation “during wash) to capture additional new market segment which is competitive edge over competitors without any investment.
- improve morale & motivation of people.
- Always not technology but team work helps to achieve Results
- Involvement & commitment of all; from operators up to top management is must for achieving high target results.