Textile News, Apparel News, RMG News, Fashion Trends
STB Projects

Controlling sewing output DHU by data visualization and root cause analysis

Abstract

In this competitive apparel market, uncontrolled Defect per Hundred Units (DHU) is killing company’s competitiveness and profit margin as well. High DHU% is causing low production efficiency and fall in quality. World apparel market totally changed throughout the last decade. Now retailers focus basically on rigid quality to take world apparel market share. Quality is the key driving force for most of the retailers worldwide. High end fashion brands are uncompromised with their product quality. For supplying fashion to the customers big brands are sourcing apparel with high quality for short fashion session. On the other hand on time shipment and no short shipment are the key performance indexes from buyer’s end. For both the things controlled DHU% can play major role to keep manufacturing companies safer and sustainable. This project focuses on standard operating procedure regarding quality management. Proper data collection and management in a standard way can bring the root causes of the defects out from the production line. The project concentrates on the implementation of Pareto chart to find out major defects, cause and defect diagram to find out root causes responsible for those defects. Then made monthly corrective action plan to achieve the planned DHU and weekly revised action plan and weekly follow-up data sheet. That’s how data visualization and data analysis is done in this project. This project consists of different lean tools, technics and lean mind set.

Key words: DHU% control, Lean Tools, Lean mind set, Data visualization, Pareto chart, Cause and effect diagram.

Introduction

Presently most of the garment industries facing problems with defective garment production and re-work for alter. These two things taking away money from factory owners and making profit margin very narrow. If DHU can be controlled it will be able to ship maximum quantity of goods produced. We all know maintaining cut to ship ratio is very crucial for all garment industry as most of the companies take order on FOB. And fabric costs approximately 60-65% of the total FOB. So there is a scope to maximize profit shipping maximum quantity of garment produced. And if re-work for defect production can be controlled it will be easy to keep the efficiency high. Efficiency is another parameter for profit. So many wastes lie in defect. Waste minimization is the key term of lean manufacturing. That means not producing defect is controlling waste. At the same time minimizing DHU we can manage finishing section easily and ensure on time shipment. If we compare cutting, sewing & finishing, finishing section suffers most before shipment. Most of the time factory needs to give extra over time to finishing section to prepare goods before final inspection. More or less DHU has impact on every section and every operation of garment industry.  So for good factory culture and good profit margin DHU% should be controlled in a proper way maintaining standard quality management procedure. That’s the main objective of this project.

Hypothesis

DHU% can be controlled through quality production. Quality production comes from quality mind set. So before we go for DHU% control we need to build a quality mind set among the respective workers and stuffs. Only after that we can apply tools to improve defect rate.

Objectives

  1. To build a quality mind set.
  2. To collect data from sewing line in a standardized way.
  3. To train the responsible quality people of the line of how we can collect data and give feedback after analysis.
  4. Set a quality goal. Here my goal is to keep DHU% within 4.
  5. To organize data to find out the major defect types that are killing factory.
  6. To find out root causes for those defects.
  7. To take corrective action plan daily and weekly basis to improve those areas.
  8. End of the project again check result through same data management.
  9. If goal achieved go to sustain that, if not re plan for improvement.
  10. Visualize sewing line defect data to the operators, inline stuffs and management.
  11. To create a smart defect management system working from the root.

Methodology

Main concept of this project to reduce defect percentage was “Right first time”. For that all possible machine folders and other attachment were used to make the operations standardized. It helps operator go through error proof environment.   As it was a project from lean perspective it’s done to eliminate every type of waste in lean way. From lean point of view re-work and over processing are wastes.

The total project work process is shown sequentially:

sewing output DHU-data visualization-total project work process

My factory is 8 lines ladies tops and shirt factory. Initially I’ve selected line no. 5 for my project. My factory’s average DHU% lies between 6-6.5%. My project target is to keep that within 4%. For that I need to know the present status of that line. First of all collected data of June 2021. Then sequentially progressed my project.

Pareto Analysis

To find out the present major defect based on present data is the initial task of all. For that I went through Pareto analysis or 80/20 rules. My project work data is shown below.

Sewing defects found by in-line quality inspectors are here.

Defect Analysis Line 5
Defect Name #Defects Cum.Total %
Raw Edge Exposure 158 12.38%
Uneven Stitch 153 24.37%
Pleat 152 36.29%
Uncut Thread/
Loose Threads
142 47.41%
Down Stitch 104 55.56%
Puckering 58 60.11%
Button Defects/ Missing 53 64.26%
Front Placcket
Up-Down
50 68.18%
Collar/Cuff Projection 49 72.02%
Lower Part Visible (Placket/OL) 48 75.78%
Label Slanted 45 79.31%
Open seam 39 82.37%
Sewing Mark 28 84.56%
Neck Loose 28 86.76%
Parts Number Mistake 25 88.71%
Skip Stitch 24 90.60%
Up-down at Seam Match Point 22 92.32%
Hole Slanted/
Position Mistake
21 93.97%
Roping Seam 19 95.45%
Sleeve Gamble
Tack Reverse
14 96.55%
Insecured Button/ Button half stitch 13 97.57%
Broken Stitch 12 98.51%
Shading 11 99.37%
Neck uneven from shoulder seam/Hala up-down 8 100.00%

From 42 types of sewing defects i found 10 defects are responsible for 80% of the total defects quantity. So I had to work on that 10 defects to find out root cause for those defects. Then took corrective action plan to reduce defect percentage.

Pareto Defect Analysis

Root cause analysis by Fish Bone Diagram

In sewing floor defect comes from material, machine, method & man. If we check these four things for every single type of defect the root cause will automatically come out. Here I’m presenting a standard fish bone diagram for my project.

Controlling sewing output DHU-oot cause analysis-Fish Bone Diagram

First of all those defects I checked man, machine, material & method which one is responsible for. Then segregated those defects according to that. After that went for final root causes considering different factors of that four main factors.

Defects from man

Operator’s problems that needs to follow-up

·         Raw Edge Exposure
·         Button Defects/ Missing
·         Front Placket Up-Down
·         Collar/Cuff Projection

·         Pleat

Operator’s problem that need training

  • Uneven stitch
  • Down stitch
  • Puckering
  • Pleat
  • Lower part visible (Placket or over lock machine operations)

Defects from machine

  • Uncut thread or loose thread

Root causes found

Operator’s problems that needs to follow-up:

Sewing supervisors are mostly effective here. They need to check those operators time to time and show them how to perform those operations right. For some special machine operations, like button attach button hole etc, operators need to be provided with standard operating procedure. Such as, for button attach machine after every power off it’s needed to attach first button on an extra piece of fabric. If found ok then go for production. Power off may happen for electric supply problem, load shedding, break for going to toilet or wash room. And alongside this operator needs to check button in every certain period of time whether there is any half stitch or not.

Operators needed in-line training:

Operators responsible for those alters are marked unskilled. They needed in-line training by supervisor and production technician. After doing that for a certain period it needed to make an assignment on those operators. If any operator found doing same defect again that one needed to be taken under training center. There they got training by training supervisor. After they get training for 2/3 days, one by one after selecting ok, sent back to sewing floor.

Defect from machine

Here this problem may happen for two reasons, machine set-up or operators awareness. First of all IE officer should sure that all machine’s auto trimming option is set-up where possible. After setting up that responsible mechanic should check body from output in a regular interval. If any thread found that mechanic should go to the specific operation and check the machine again. But if mechanic found no machine trimmer possible to cut the thread IE officer should knock the responsible supervisor to ask the operator to cut thread manually. After taking that two corrective actions the supervisor & IE officer should follow up to sustain that action to get better result.

This action plan was for the month of July, 2021. And plan was under the supervision of respective e IE officer. He monitored the total execution with sewing supervisor and mechanic. While executing the project IE officer needed to check the present progress, that’s why there was a weekly monitoring and analyzing file. After analyzing data of every week he revised the action plan that demander the situation. Every week he shared data with production team and training supervisor as well. Here is the weekly corrective action follow up format for reference.

Daily Defect Control Actions Monitoring
Kaniz Fashion Ltd., Group QA
Date:
Team:27
Operator’s Skill Check Analyzed DHU% –
Operator Name Operation DHU% against
total quantity
 

 

Mahine Set-up Check

Analyzed DHU% –
Operation Set-up point DHU% against
total quantity
Changed Method Result Check Analyzed DHU% –
Operation Previous Method New Method DHU% against
total quantity
………………… …………………. ………………… ………………..
IE QS F.Incharge Manager, IE

This format is filled up analyzing every week data. Then sit together with respective people to take actions. There is another daily defect awareness sheet also, top three report, which is pined up on the board in front of the line. This task is done by roving quality inspector (Line quality). Top three defects report of first day of July is shown below.

Root cause analysis-corrective-action-plan

Result

AT the end of the month July, 2021 again analyzed data got from in-line quality inspectors. We got month DHU 4.14%, which is lower than last month but little bit higher than our expectation. The analysis result is shown below.

Defect Name Defect% Cum.Total %
Down Stitch 12.8% 12.82%
Raw Edge Exposure 9.9% 22.71%
Puckering 9.5% 32.23%
Neck uneven from shoulder seam/Hala up-down 8.1% 40.29%
Shading 8.1% 48.35%
Pleat 7.7% 56.04%
Button Defects/ Missing/Half stitch 7.7% 63.74%
Collar/Cuff Projection 7.0% 70.70%
Broken Stitch 6.6% 77.29%

Here total number of defects responsible for 80% of the defect quantity is 9 than is 1 less than before. 5 defects are common for both the months, which means 5 defects rate decrease significantly. That’s effective result of the project.

Causes of result deviation from expectation

Last month DHU was 5.61% and expected DHU below 4%. After doing the project work we got DHU 4.14%. We the IE team did this project great and with good effort. But result is little bit below from our expectation. Getting the result we sit together and tried to find out the reasons. This time factory KPI dash board was our source of root cause analysis data. There we found two things that causes deviation from expected result; huge labor turn over & too many style changes. Usually KFL’s labor turnover rate is 5-6%, number of average style changes is 16. July has been short month as due to eid vacation total number of working days was 16, but style changes was 15. And labor turnover rate has been 9.5% which is not considerable.

Conclusion

Our project has been a great experience for all of us though result is little bit below from our expectation. There we saw a great change in mind set of all manpower related to that line. I think that’s the spirit to bring about a change, Lean also says the same thing. From that point of view we did great there.

Md. Abdullah Al-Hadi-Kaniz Fashion Ltd
Author: Md. Abdullah Al-Hadi, Sr. Executive, IE, Kaniz Fashion Ltd., Group QA.

Recommendation

Only quality tools can’t control DHU% alone. Some other factors are also responsible for that. If management policy is made to control labor turnover and absenteeism percentage that enhance quality also as labor is the main driving force of all operations. At the same time material quality control procedure should be firmed to get better result along with inline sewing defect control. DHU% control procedure should be long term initiative rather than one or two month. One or two month may be initial time only.

If anyone has any feedback or input regarding the published news, please contact: info@textiletoday.com.bd

Related posts

Resolving bottlenecks of garment sewing floor in possible ways

Ibrahim Khalil

Bangladesh to embrace a new era of technology with Coats Digital

Textile Today

RFT dyeing to meet perfect quality and gain profit

Textile Today

Latest Publications

View All