how a single black belt project jump starts a successful lean six sigma effort
TRANSCRIPT
Executive Summary
Project Wins• Improved Bent/Scratched/Damaged Scrap $$$ from average
of $5850/month to $1362/month in Material Handling Department from August 2015 to December 2015 – over 76% Scrap Reduction – goal was 50%.
• Designed more efficient racks, improved employee cross-training, established Control Charts/Monitoring Plan to sustain improvements
• Improved outdated Standard Work Instructions with Visual Management and identified/implemented simple changes (spacer/arm covers/shavings/tiedowns) to the process.
• Created new storage area that eliminated 1,527 feet of wasted Motion/Transportation due to Process Bottleneck.
DMAIC Approach• Gathered Internal Customer Feedback using Affinity
Analysis• Identified Root Causes with a Fishbone Diagram
(Ishikawa)• Used a Process Walk to identify the 8 Wastes• Used LCS for Rack Design Change/Pilot• Put Controls in place for sustainment of improvements
Context• The flawed process was on pace for a 300% increase in
BSD Scrap. Identifying these root causes not only allowed us to improve the process – but also identify Best Practices to implement throughout our other facilities across the U.S. and Canada.
Project Start: 9/7/2015 Project End: 1/27/2016
$4,622.06
$8,114.10
$4,877.85
$4,942.94$5,653.64
$3,854.11
$6,412.27
$8,327.46
$6,177.00
$4,541.79
$2,252.09$1,362.02
$0.00
$1,000.00
$2,000.00
$3,000.00
$4,000.00
$5,000.00
$6,000.00
$7,000.00
$8,000.00
$9,000.00
Jan-15 Feb-15 Mar-15 Apr-15 May-15 Jun-15 Jul-15 Aug-15 Sep-15 Oct-15 Nov-15 Dec-15
2015- B/S/D- MATERIALHANDLING
SCRAP$$
Operational Definitions• Anodizing (ANO) – chemical coating process that converts metal stock
lengths into a decorative, durable, corrosion-resistant finish• B/S/D = Bent/Scratched/Damaged Material• Bent Material – Material bowed in excess of 1/32” for every 4.5” in width
(Max width is 9”)• Damaged Material – Any combination of Bent/Scratched metal in addition
to any damage that occurs during transit• Extrusion (EXT) – process where metal logs are pushed through a die to
create objects/shapes of a fixed profile• Paint (PNT) – process of pre-treating and adding full coats of primer and
desired paint color along conveyor• Scratched Material – Any length scratch appearing in an exposed area of
metal and/or scratch in excess of 3” in unexposed area of metal• Exposed Area – Any area of the metal that is on the exterior of the building
or visible through a lite of glass• Unexposed Area – Any area of the metal that is not visible upon installation
Define• Weighted Criteria Matrix• Project Charter• Collect VOC/Internal Customer Feedback (Affinity Analysis)• VOC Translation Matrix• High Level Map (SIPOC)• Process Walk Planning Checklist• Detailed Process Map (Swimlane)• Stakeholder Analysis• Relationship Map• Communication Plan• Pose and Answer
Weighted Criteria Matrix
Project CharterSubtitle
Scrap $$$ for the Material Handling Dept. has risen to $5850/month in 2015 (January-August) for material marked as Bent/Scratched/Damaged. The Scrap $$$ in 2014 for the same category was just $2141/month. While the costs obviously hurt the bottom line, the inability to push a good quality product in a timely manner impacts delivery to our Assembly/Glazing Plants and eventually; runs the risk of missing Customer Due Dates.
My organization's Strategy for 2015 was to create a culture of (1) Problem Solving and (2) Continuous Improvement. With B/S/D Scrap $$$ projected to jump almost 300% from 2014 to 2015, this cross-departmental project is our best opportunity to reduce scrap costs across the board while aligning our efforts with the Organizational Strategy. This project
Problem Statement
Goal Statement
Business Case & Benefits
The goal is to reduce Bent/Scratched/Damaged Scrap $$$ by 50% from $5850/month to $2925/month by the end of 4Q/2015.
TimelinePreliminary Plan: Timeline Planned Actual Define 9/28/15 9/27/15 Measure 10/26/15 10/27/15 Analyze 11/16/15 11/19/15 Improve 12/14/15 12/24/15 Control
Scope In/OutProcess : Material Handling Department handles the flow of all Metal traveling from (1) Extrusion Presses, (2) Anodizing Lines, (3) Paint Department, (4) Packing/Shipping. They are the only department that touches all four areas of our Metal Processing Plant.
Team MembersTeam Lead: Jeff (Quality Engineer/Green Belt) - 25% Team Member 1: Michael (Process Owner) - 20% Team Member 2: Dexter M. (Mat. Handler/Driver) - 15% Team Member 3: Vincent L. (Maintenance) - 15% Team Member 4: Jaxson P. (Material Handler) - 15% Team Member 5: Kevin Y. (Fabrication) - 5%
VOC – AFFINITY ANALYSISCustomers of this Process-driven project are all internal departments (Extrusion, Anodizing, Paint). The feedback focused on three important factors:
VOC Translation MatrixCustomer Comment Gathering More Understanding Customer Requirement
Skipping Visual Checks - bad material being ignored
Customers want material reviewed at each process step
Each Dept. (EXT/ANO/PAINT) along with Material Handling to initial traveler showing QC check performed
Poorly conditioned (decaying/nails) racks scratching parts
Customers want better racks to protect parts (consider re-design)
Remove all decaying/defective racks from circulation - replace with new
Material arriving out of order, items left waiting, not rushing jobs that should be
Customers want to receive parts as needed and have a more organized scheduling system
Daily schedule for flow of Material/Orders - sync across EXT/ANO/PNT
Parts being stacked on exposed surfaces Customers want parts properly stacked to avoid scratches and double-handling
All employees to receive training - signoff on process of stacking parts
Metal spacers in circulation with no (or torn) sleeve covers
Customers want spacers properly covered to prevent scratches to material
Metal spacers to have kevlar sleeves replaced/stapled prior to loading parts
Employees not aware if material can be reworked - assumed scrap
Customer can prevent serious scrap $$$ by reworking material
Relocate potential rework to MRB area so Dept Managers can sign-off/scrap
Voice Of the Customer Translation Matrix
SIPOC
S I P O CSupplier Input Process Output Customer
Extrusion Logs Extrusion Press Metal Stock Lengths Anodizing (Internal)
Anodizing Anoding TanksMetal Stock Lengths anodized (chemical
coating)Anodized Coating Paint (Internal)
Paint Paintline Primers - Full Coats of Paint Applied Painted Metal Shipping (Internal)
Packing/Shipping Load Trucks Match Material to Orders Finished Goods
General Contractors - Construction Firms
(6) Relocate potential Rework to MRB area so Dept Managers can sign-off/scrapTraining: (5) All employees to receive training - signoff on process of stacking parts
SIPOC
Extruded thru Dies (shapes)
Customer Requirements:Quality: (1) Each Dept (EXT/ANO/PNT) and MH to initial traveler showing QC check (2) Remove all decaying/defective racks from circulation - replace with new (3) Metal Spacers have kevlar sleeves replaced/stapled prior to useOTIF: (4) Daily Schedule for flow of Material/Orders - sync across EXT/ANO/PAINT
Process Walk Planning
Activity Completed(Yes or No)
Develop a high level map of the process to be “walked” (SIPOC, Spaghetti) Yes
Create a list of all Process Walkers/Interviewers (Jeff, Michael) Yes
Create a list of all Interviewees (Dexter M., Jaxson P., Steven R.) Yes
Create a list of Stakeholders to be invited to the Stakeholder Presentation (Tim J., Paul K., Klay M., Wes B.)
Yes
Develop a Communication Plan for the event Yes
Create a high level agenda for the Process Walk Yes
Set a tentative date/time for Process Walk Orientation Yes
List tentative dates/times and location for the Process Walk and the Stakeholder Update Yes
List tentative dates/times and location for the Stakeholder Update (last day) Yes
Process Walk Planning Checklist
Notes: *Orientation/Walk/Interviews scheduled for 9/22-9/23/15
Swimlane Map
Stakeholder Analysis
Stakeholder/ Stakeholder
GroupImpact Level Level of Support Reason for
Resistance or Support
Action(s) to Address This Stakeholder
GroupContact
VP-MANUFACTURING DECISION AUTHORITY SUPPORTIVE FINANCIAL BACKING - WILL
HELP CLEAR BARRIERS
UPDATE/NOTIFY IF ANY PURCHASES OR MATERIAL
NEEDEDTIM J.
METAL PROC. MANAGER IMPACTS NEUTRAL
AWARE OF PROBLEM BUT UNSURE EMPLOYEES CAN
HANDLE CHANGES
CREATE WORK INSTRUCTIONS/TRAINING
OPPORTUNITIES EARLY IN THE PROCESS
KLAY M.
PLANT MGR AFFECTED RESISTANTAFRAID CHANGES COULD
LEAD TO NON-VALUE ADDED TIME TO CURRENT PROCESS
USE DATA TO SHOW EXPECTED COST SAVINGS AND POTENTIAL LOSSES IF
NOT ADDRESSED
WES B.
QUALITY TEAM DECISION AUTHORITY SUPPORTIVE
ANXIOUS TO SHOW PROCESS IMPROVEMENT TEAMS CAN
ADD VALUE TO THE ORGANIZATION
CREATE USEFUL TEMPLATES/DOCUMENT STRATEGIES - IDENTIFY
GOOD/BAD LESSONS LEARNED
PAUL K.
Stakeholder Analysis (Advanced) - DEFINE
Relationship MapProject:
Relationship Map
BSD Scrap Reduction Project
Decision Authority - Tim J. (VP-
Manufacturing)
Impacts Outcome -Klay M. (Metal
Processing Mgr)
Will Be Affected -Wes B. (Plant Mgr)
ProjectPaul K. (Quality
Dir)
Communication PlanStakeholder Group Objectives/Actions Desired Message Delivery
Method/Venue Timing
Communication Plan
VP - Manufacturing
Metal Processing Mgr
Plant Manager
Quality Team
Communicate any financial limitations or restrictions -
Support the Project
Have leads/supervisors give full access to employees and
help with training needs
Enforce countermeasures and action plans put in place once
root causes identified
Support team members and encourage other departments
to research future projects
Potential changes to the racks/sleeves need to be explored for $$$ Email Bi-Weekly
Allow time for Process Walks/Interviews to allow team to build
knowledge of the current state1-on-1/Email Weekly (Tues/Thurs)
Assist Process Improvement Activities in focused areas and review data ($$)
that backs up need for projectWeekly Dept. Meetings Weekly (Tues/Thurs)
Document Activities/Progression showing value of a culture change 1-on-1 Weekly
Pose and AnswerPose & Answer - Influence Strategy
Questions Answers
What is important about this project? It is a broken/flawed process that impacts all departments of our plant.
Why is this project important to do right now? The Scrap $$$ for BSD is in a freefall and on pace to triple our scrap from 2014 (thru 8 months).
Why is this project more important than other pressing projects?
Other projects can be handled in a 3-5 day Kaizen blitz - more of a quick fix. This project hits the bottom line more than others and has cross-functional impacts and best practices that can be gained.
What problems will this project solve? It will identify the root causes for a plant-wide issue - for which currently there are none.
How will this project help your colleagues? It will help specify process improvement activities in focused areas.
How will the project help your manager? It gives my manager the ammunition (proof) that our Process Improvement Teams can work hand-in-hand with all departments and add value to the organization.
In what ways does the business gain from this project? We will gather the lessons learned to fix the problem and put metrics together to prevent it from getting out of control.
Draft Business Case
What are some positive side effects of doing this project?
The ultimate value will be the Best Practices created that can be crossed over and implemented throughout our other facilities across the United States and Canada.
Other? My goal is to communicate (with proven results) to the shop employees, department managers, and upper level management that the Lean-Based Culture is something we need to completely adopt.
Measure• Data Collection Planning• Bent Material (Visual Management)• Scratched Material (Visual Management)• Damaged Material (Visual Management)• Checksheets• Calculate Sampling Plan• Measurement Systems Analysis (MSA)• Baseline Measure Chart• Process Capability
Data Collection Planning
Measure Data Type Operational Definition Stratification Factors Sampling Notes Who and How
BENT MATERIAL DISCRETE
MATERIAL BOWED IN EXCESS OF 1/32" FOR EVERY 4.5" IN WIDTH (MAX WIDTH IS 9")
(1) LOCATION (EXT/ANO/PAINT) (2) SHIFT (1st/2nd)
SAMPLE EVERY 10TH RACK OF FG1101M (25 PIECES PER RACK/3 RACKS MAX) PER SHIFT EACH DAY FOR 5 DAYS (10/19-10/23) - 750 TOTAL PIECES SIMULTANEOUS FOR BENT, SCRATCHED, AND DAMAGED
MATERIAL HANDLING TEAM - 1 RACK EACH AT EXT/ANO/PAINT - VISUAL INSPECTION - USE CHECKSHEET - FOLLOW MARKED RACKS FROM EXT TO ANO TO PAINT
SCRATCHED MATERIAL DISCRETE
ANY LENGTH SCRATCH APPEARING IN AN EXPOSED AREA OF METAL AND/OR SCRATCH IN EXCESS OF 3" IN UNEXPOSED AREA OF METAL
(1) LOCATION (EXT/ANO/PAINT) (2) SHIFT (1st/2nd)
SAMPLE EVERY 10TH RACK OF FG1101M (25 PIECES PER RACK/3 RACKS MAX) PER SHIFT EACH DAY FOR 5 DAYS (10/19-10/23) - 750 TOTAL PIECES SIMULTANEOUS FOR BENT, SCRATCHED, AND DAMAGED
MATERIAL HANDLING TEAM - 1 RACK EACH AT EXT/ANO/PAINT - VISUAL INSPECTION - USE CHECKSHEET - FOLLOW MARKED RACKS FROM EXT TO ANO TO PAINT
DAMAGED MATERIAL DISCRETE
ANY COMBINATION OF BENT AND SCRATCHED METAL IN ADDITION TO ANY DAMAGE THAT OCCURS DURING TRANSIT OF MATERIAL
(1) LOCATION (EXT/ANO/PAINT) (2) SHIFT (1st/2nd)
SAMPLE EVERY 10TH RACK OF FG1101M (25 PIECES PER RACK/3 RACKS MAX) PER SHIFT EACH DAY FOR 5 DAYS (10/19-10/23) - 750 TOTAL PIECES SIMULTANEOUS FOR BENT, SCRATCHED, AND DAMAGED
MATERIAL HANDLING TEAM - 1 RACK EACH AT EXT/ANO/PAINT - VISUAL INSPECTION - USE CHECKSHEET - FOLLOW MARKED RACKS FROM EXT TO ANO TO PAINT
Data Collection Planning
Bent Material
MATERIALBOWEDINEXCESSOF1/32"FOREVERY4.5"INWIDTH(MAXWIDTHIS9")
Scratched MaterialExposedArea– 2”scratch
ANYLENGTHSCRATCHAPPEARINGINANEXPOSEDAREAOFMETALAND/ORSCRATCHINEXCESSOF3"INUNEXPOSEDAREAOFMETAL
UnexposedArea– 6”scratch
Damaged Material
ANYCOMBINATIONOFBENTANDSCRATCHEDMETALINADDITIONTOANYDAMAGETHATOCCURSDURINGTRANSITOFMATERIAL
Strap/Tie-downDamage ForkliftTransitDamage
ChecksheetsB/S/DMETALCHECK DATE: B/S/DMETALCHECK DATE:
TEAM: PART# TEAM: PART#SHIFT: LENGTH SHIFT: LENGTHLOCATION: PIECES: LOCATION: PIECES:
TYPE COUNT QTY/RW COUNTTOTAL TYPE COUNT QTY/RW COUNTTOTALBent BentScratched ScratchedDamage/Transit Damage/Transit
B/S/DMETALCHECK DATE: B/S/DMETALCHECK DATE:
TEAM: PART# TEAM: PART#SHIFT: LENGTH SHIFT: LENGTHLOCATION: PIECES: LOCATION: PIECES:
TYPE COUNT QTY/RW COUNTTOTAL TYPE COUNT QTY/RW COUNTTOTALBent BentScratched ScratchedDamage/Transit Damage/Transit
B/S/DMETALCHECK DATE: B/S/DMETALCHECK DATE:
TEAM: PART# TEAM: PART#SHIFT: LENGTH SHIFT: LENGTHLOCATION: PIECES: LOCATION: PIECES:
TYPE COUNT QTY/RW COUNTTOTAL TYPE COUNT QTY/RW COUNTTOTALBent BentScratched ScratchedDamage/Transit Damage/Transit
B/S/DMETALCHECK DATE: B/S/DMETALCHECK DATE:
TEAM: PART# TEAM: PART#SHIFT: LENGTH SHIFT: LENGTHLOCATION: PIECES: LOCATION: PIECES:
TYPE COUNT QTY/RW COUNTTOTAL TYPE COUNT QTY/RW COUNTTOTALBent BentScratched ScratchedDamage/Transit Damage/Transit
*Markanyimpactedpiecewitha(B)Bent,(S)Scratch,or(D)Damagedinunexposedareaofmetal
Calculate Sampling Plan• Process Sampling (Live)• Stratified/Systematic• Ongoing – every 10th rack• 50% Estimate of Prop (worst case)• Stratification Factors
1. Location2. Shift
SamplingCalculation(Discrete)forRacksofBSDMaterial-MINITAB
Parameter Proportion Distribution Binomial Proportion 0.5 Confidence level 95% Confidence interval Two-sided Results Margin Sample of Error Size 0.04 623
With a Metal Processing Plant and an unlimited amount of samples at my disposal – I chose to use the 50% Estimate of Proportion for a Worst-Case Scenario. This gives me a minimum sampling size of 623 pieces.
MSA – Gage R&R (Discrete)1. Standard/Reference – BSD Material referenced in Data Collection Planning
Operational Definitions (Pg 16)2. Appraisers/Operators – Dexter M. (1st Shift)/Jaxson P. (2nd shift)3. Parts – WW400 – 20 total (14 good/6bad) – blind and unbiased (highlights)4. Trials – Twice for each operator5. Measurement/Response – 0=Defect, 1=Good
GOOD
DEFECT
MSA - ContinuedAttribute Agreement Analysis for ASSESSED RESULT
Date of study: 10/12/2015Reported by: Jeff OwensName of product: WW400Misc: BB Project - Discrete MSA Analysis
Within Appraisers
Assessment AgreementFleiss’ Kappa Statistics
Appraiser Response Kappa SE Kappa Z P(vs > 0)A 0 0.856631 0.223607 3.83097 0.0001
1 0.856631 0.223607 3.83097 0.0001B 0 0.687500 0.223607 3.07459 0.0011
1 0.687500 0.223607 3.07459 0.0011
Each Appraiser vs Standard
Assessment AgreementFleiss’ Kappa Statistics
Appraiser Response Kappa SE Kappa Z P(vs > 0)A 0 0.803971 0.158114 5.08476 0.0000
1 0.803971 0.158114 5.08476 0.0000B 0 0.722250 0.158114 4.56791 0.0000
1 0.722250 0.158114 4.56791 0.0000
Between Appraisers
Assessment AgreementFleiss’ Kappa Statistics
Response Kappa SE Kappa Z P(vs > 0)0 0.825708 0.0912871 9.04518 0.00001 0.825708 0.0912871 9.04518 0.0000
All Appraisers vs Standard
Assessment AgreementFleiss’ Kappa Statistics
Response Kappa SE Kappa Z P(vs > 0)0 0.763111 0.111803 6.82547 0.00001 0.763111 0.111803 6.82547 0.0000
InregardstotheAppraisersvs.Standard– Fleiss’ Kappareadingsremainedmarginal(between0.70-0.90)
BA
100
95
90
85
80
75
70
65
Appraiser
Perc
ent
95.0% CIPercent
BA
100
95
90
85
80
75
70
65
Appraiser
Perc
ent
95.0% CIPercent
Date of study: 10/12/2015Reported by: Jeff OwensName of product: WW400Misc: BB Project - Discrete MSA Analysis
Assessment Agreement
Within Appraisers Appraiser vs Standard
Baseline Measure Chart
After running the baseline data – projections showed we had still made no progress towards our goal of reducing scrap to
$2925/month
$2,141.04
$5886.83
$5427.23
$0.00
$1,000.00
$2,000.00
$3,000.00
$4,000.00
$5,000.00
$6,000.00
$7,000.00
2014 Jan-Sept2015 10/15Baseline
BSDMonthlyScrap- MaterialHandling
Process Capability• DPU = # Defects/# Units = 422/750 = 0.56
• DPMO = 422/3000 x 1,000,000 = 140,667 (4 Opportunities)
• Sigma Level = 2.5 (rounded down), 84.1% Yield (158,655)
• Final Yield (Y Final) = # Good Units/# Total Units = 639/750 = 85.2%
• First Pass Yield (YFP) = 750-422 = 328, 328/750 = 44%
• Rolled Throughput Yield (YRTP) – EXT (61%), ANO (88%), PNT (93%)– .61 x .88 x .93 = 50%
Analyze• 8 Wastes Checksheet• Fishbone Diagram• Data for Statistical Testing• Hypothesis Statement – One Proportion • Hypothesis Statement – Two Proportion• Hypothesis Statement – Chi-Square• Hypothesis Testing Plan Results
8 Wastes ChecksheetDate: 9/23/2015
Waste Definition Examples Type "High", "Medium", "Low" Description of Issues
- Inaccurate applications
- Broken parts
- Missed deadlines
- Extra copies of reports
- Redundant storage (hard & soft)
- "Reply All" on emails
- Waiting for approvals
- Waiting for equipment
- Waiting for large batches
- Employees unable to make decisions
- Employees not fully trained
- Skilled employees doing unskilled tasks
- Hand-offs between functions
- Multiple reviews
- Sending, resending emails
- Stockpiling supplies
- Information piling up for data entry
- Keeping data longer than necessary
- Repetitive key strokes
- Walking between equipment
- Switching applications
- Extra formatting, extra fields
- Extra features, excess detail
- Extra report information
The 8 Wastes Checksheet
M MotionAny movement by people that is not of value to the
customer
Material between Extrusion and Anodizing is kept across street as in Transportation - badly need
storage in these areas
Poor Layout makes Inventory seem worse than it actually is
TransportationUnnecessary movement of
materials, information or equipment
I Inventory
Accumulation of parts, information, applications,
etc. beyond what is required by the customer
Only source of storage is across the street and the racks transporting material are in terrible
shape - too much double-handling
Material Handling throughout EXT/ANO/PNTProcess Area:
D DefectsInformation, products and
services that are incomplete or innacurate
O OverproductionMaking more of something - making it earlier or faster-
than it's needed
W WaitingWaiting for information,
equipment, materials, parts or people
There are a few Value-Enabling steps but overall not a significant issueE Extra-
Processing
Any steps that do not add value in the eyes of the
customer
Low
High
Low
High
N Non-Utilized Talent
Not properly utilizing people's experience, skills,
knowledge or creativity
T
Defects occurring at every stop (EXT/ANO/PNT) due to inability to follow work instructions,
inspections, and safely/securely handle material transfer
Not currently a problem - waiting seems to be as a result of the bottleneck in the Extrusion Ovens
Mentioned above, cycle time for Extrusion is higher than Anodize and Paint combined. Work
schedules would be better off altered.
Lack of standard work and cross-training stands out - changing this could perhaps build more
ownership among the employees
High
Low
High
Medium
Fishbone Diagram
Data for Statistical Testing
Hypothesis Statement – One Proportion Test
According to past performance (2014), BSD Defect Rates were at 7% for Extrusion/Anodizing/Paint. The team suspects it’s getting worse due to the conditions of the wooden racks, which typically rot out within 3 years. Recent data has been collected and we want to know if BSD Defect Rates have risen. Note that current racks in circulation were built in early 2013. (X) = Time Period.
NULL: Ho P = .07 (7%)
ALTERNATE: Ha P ≠ .07 (7%) or P › .07 (7%)
Test: One Proportion (Proportion compared to standard). In this case, we are determining if there is a difference between a historical level of defects and current defect levels.
One Proportion Test Analysis
Conclusion: BSD Defect Rates have truly gotten worse, jumping from 7% in 2014 to 11% in 2015. The P Value (0.000) is low (less than 0.05)
so the Null must be rejected. The result of this test is True.
TestandCIforOneProportion (MINITAB)Test of p = 0.07 vs p ≠ 0.07Sample X N Sample p 95% CI Z-Value P-Value1 69 630 0.109524 (0.085138, 0.133910) 3.89 0.000Using the normal approximation.
0%
2%
4%
6%
8%
10%
12%
2014 2015
BSDDEFECTRATES– EXT/ANO/PNT
BSDDEFECTRATES
Hypothesis Statement – Two Proportion Test
The team suspects that inadequate training to 2nd shift employees has resulted in more defects being allowed to pass on 2nd shift than 1st shift. This test will help us determine if there is a difference between the two. (X) = Work Shifts. P1 = 2nd shift, P2 = 1st shift.
NULL: Ho P1 - P2 = 0
ALTERNATE: Ha P1 - P2 ≠ 0
Test: Two Proportion (Proportion of a data set against another) to see if one area is underperforming. Each shift will review the same sample size lot of 630 pieces in the Extrusion Area.
Two Proportion Test Analysis
Conclusion: BSD Defect Rates are significantly different between the two shifts with 2nd shift allowing 14% of defects compared to 9% for 1st shift while reviewing the same lot of material. The P Value (0.006) is low (less than 0.05) so the Null must be rejected. The result of this test is True.
TestandCIforTwoProportions:2ndShift,1stShift(MINITAB)Event = GoodVariable X N Sample p2nd Shift 542 630 0.8603171st Shift 573 630 0.909524
Difference = p (2nd Shift) - p (1st Shift)Estimate for difference: -0.049206395% CI for difference: (-0.0843421, -0.0140706)Test for difference = 0 (vs ≠ 0): Z = -2.74 P-Value = 0.006
Fisher’s exact test: P-Value = 0.008
0% 2% 4% 6% 8% 10% 12% 14%
2ndShift
1stShift
BSDDEFECTRATES
BSDDEFECTRATES
Hypothesis Statement – Chi-Square
The team suspects that better inspection is being done in the Paint Department compared to the Extrusion and Anodizing Departments. We will test Defect Rates in each department to see if there is truly a difference in the three. (X) = Departments. P1 = Paint, P2 = Extrusion, P3 = Anodizing.
NULL: Ho P1 = P2 = P3
ALTERNATE: Ha P1 ≠ P2 ≠ P3
Test: Chi-Square (2 or more Proportions against each other). In this case, we are determining if there is a difference between defect rates in different departments (EXT/ANO/PNT) in order to find Best Practices for Inspection.
Chi-Square Test Analysis
Chi-SquareTest– MINITAB
Rows: Worksheet rows Columns: Worksheet columnsEXTRUSION ANODIZE
PAINT SUMM SUMM SUMM All1 566 555 558 1679
559.67 559.67 559.670.07167 0.03891 0.00496
2 64 75 72 21170.33 70.33 70.33
0.57030 0.30964 0.03949
All 630 630 630 1890
Cell Contents: CountExpected countContribution to Chi-square
Pearson Chi-Square = 1.035, DF = 2, P-Value = 0.596Likelihood Ratio Chi-Square = 1.046, DF = 2, P-Value = 0.593
Conclusion: BSD Defect Rates showed variation but are not statistically different between the three departments. The P Value (0.596) is high (more than 0.05) so
we fail to reject the Null. The result of this test is False.
7%
8%
9%
10%
11%
12%
13%
14%
15%
16%
PAINT EXTRUSION ANODIZING
BSDDEFECTRATES
BSDDEFECTRATES
Hypothesis Testing Plan Results
Possible X Root Cause Hypothesis Hypothesis Test Results
Time Period BSD Defect rates were lower in the past when newer racks were being used to transport material One Proportion TRUE
Work Shifts 2nd Shift Material Handlers are responsible for allowing more defects than the 1st Shift Two Proportion TRUE
DepartmentsMaterial Handlers in the Paint Department are doing a better job of inspecting material than the Extrusion and Anodizing Departments
Chi-Square FALSE
Hypothesis Testing Plan
Improve• Root Causes to Address (Fishbone)• Rack Design Change• Poor Lighting• Poor Layout• New Storage Area• Spacers• Dolly/Rack Covers• Shavings• Tie-Down Material• Cross-Training Matrix• Updated Communication Plan• Updated Stakeholder Analysis• Implementation Plan
Root Causes to Address
The Team referenced our Fishbone Diagram to brainstorm solutions that address our root causes. The highlighted root causes (aside from the Racks) above allowed for the most impact
with the least amount of effort. A new Rack Design is our featured change.
Rack Design ChangeBased on our Hypothesis for a One Proportion Test in the Analyze Phase, we were able to show an increase in Defect Rates from 7% in 2014 to 11% in 2015 due to the poor conditions of wooden racks used to transport material. The pictures below give you an idea of the battle our Material Handling Department has of not damaging material in transit.
Within 2-3 years, the wooden racks begin to rot and expose nails that results in Scratched Material. In addition, the wood eventually loses its structure and will bottom out on other
material, making it susceptible to scratches and bowed metal (Bent Material).
Rack Design ChangeWith a game-changing issue such as Rack Design, we decided to add a few team members and used another DMAIC tool (LCS) to get ideas from other departments upstream and downstream that also have to deal with the racks. This allowed us to overhaul the Racks using metal that was originally scrapped with the help of our Brake Metal and Welding Shop. Our newest team members helped suggest a preset lifting length that could be setup for all forklifts and therefore be used throughout the entire campus. The new design would also carry enough strength/stability for our longest material lengths (28 feet) while maintaining features mandated by our Safety Department (weighing under 75 lbs so two team members can move).
Rack Design Change PilotArmed with a new design and taking advantage of material that had been left for scrap – we had 15 racks built and deployed five each to Extrusion, Anodizing, and Paint to gain feedback of potential improvements to the Material Handling Process. After carefully assessing the risk of process changes, we would heavily use them in circulation for the next two weeks before making any large scale plans. The supportive feedback was unanimous on both shifts and we made it known to our Stakeholders that we wanted to fully replace our 800+ wooden racks already in circulation. Since the previous process involved purchasing the wood and building from within, the changed process of using already scrapped metal and welding on-site was an actual push in terms of Return on Investment ($$$).
Poor LightingOne root cause that was easily identified as a “Quick Fix” or simply just needing implementation was the poor lighting in the Inspection Areas near the Anodizing Lines. We found this to be a routine problem throughout the plant and added Maintenance to our list of Stakeholders to help ensure this doesn’t re-occur.
Poor LayoutThe layout of our plant remains the biggest challenge we cross every day. With our final step of the Metal Processing lifecycle (Paint) and our Warehouse being located across the street, it leaves our Material Handling Department open to multiple risks in transit –not to mention a horrible violation of the 8 Wastes (Motion/Transportation).
Poor LayoutAfterreviewingthewastedmotiongoingbackandforthtotheWarehousearea,ourteamwasabletonegotiatesomeunused spacebetweentheExtrusionandAnodizingareasforextrudedmaterialawaitingAnodizing.Thisknocked downthe1,527feetduetoabottleneckto lessthan50feetandfreesuptheWarehouseformorepressingneeds.
NewStorage
New Storage Area
SpacersOur process walk during the Define Phase shed the light on a problem that we’ve been familiar with but simply was not being implemented in the Extrusion area. All metal spacers should have stapled kevlarcovers on them. Uncovered spacers lead to Scratched Material. It’s already in the Work Instructions but was not being followed. After being told they were not getting much life out of the old style kevlar – we got the Purchasing Dept. involved to make sure we could get a better product with a longer life expectancy.
ScratchedMaterial
Dolly/Rack CoversAlong the lines of the Spacer Covers was an issue with the arms of the dollies/racks rubbing up against material and causing scratches. Our Purchasing Dept. was able to use the same company from the Spacer Covers for these kevlar-coated Rack Covers.
ShavingsShavings left on extruded metal can only result in one thing – Scratched Material. This is another implementation item that is called out in all Work Instructions but not being followed by all employees despite air guns at the Extrusion Presses. We had all department supervisors reiterate in their pre-shift meetings the importance of this and will also make it more visual in the Work Instructions.
Tie-Down MaterialAs you’ve seen with the Spaghetti Map, material travels quite a distance from our Anodizing Department and across the street to Paint and the warehouse for storage. If a batch of material is not properly tied down, it will shift while in transit and lead to Damaged Material. This has also led to the need to upgrade the road conditions for the material transfer. This will be an added Training feature and part of the updated Work Instructions.
ProperlyTied
Cross-Training MatrixIn our Two Proportion Hypothesis Test, we were able to show a statistically significant difference in the defect rates between 1st and 2nd shift due to inadequate training. We created a Cross-Training Skills Matrix to ensure both shifts would be equally trained.
NAME
NOT TRAINED
IN TRAINING - OPERATES UNDER SUPERVISION
CAPABLE WITH MINIMAL SUPERVISION
OPERATES INDEPENDENTLY TO STANDARDS
FULLY COMPETENT - CAN TRAIN OTHERS
TEAM LEADER OR CAN FILL IN FOR TEAM LEADER
MATERIALHANDLINGCROSS-TRAININGSKILLSMATRIX
READ PRINTS
STACKING
PROCEDURES
DETECT REWORK
LOADING AND
UNLOADING
EPICS
KNOWLEDGE
DEFECT CODE
KNOWLEDGE
FORKLIFT
CERTIFICATI
ON
PALLET JACK
TRAINING
CRANE/HOIST
TRAINING
SKILLS
UNDERSTAND
WORK
INSTRUCTIONS
MATERIAL
INSPECTION
INTERPRET
B/S/D
MATERIAL
Updated Communication Plan
Stakeholder Group Objectives/Actions Desired Message Delivery Method/Venue Timing
Communication Plan
VP - Manufacturing
Metal Processing Mgr
Plant Manager
Quality Team
Inform of project findings and report back potential
costs/expected savings
Inform of project findings, foster buy-in and ownership,
participate in roll-out
Inform of project findings, foster buy-in and ownership,
participate in roll-out
Document Project Timeline for Best Practices
Translate ROI for racks and costs of sleeves after discussions with
Purchasing - inform of pilot resultsEmail Within 1 week
Inform of pilot results, work instruction overhaul, training and implementation
plan
Schedule meeting along with Plant
ManagerWithin 2 weeks
Inform of pilot results and implementation plan
Maintenance
Purchasing
Safety
(1) lighting at inspection areas (2) road conditions for material transit
(3) preventive maintenance as identified
Purchase sleeves (covers) for spacers and dolly/rack arms
Review new Rack design
Schedule meeting along with MP Manager Within 2 weeks
Use project results to help identify/teach other departments about
the importance of a Lean CultureEmail After Project Storyboard
is Complete
Need areas of the plant setup immediately - order safety stock Email Today
Research/compare suppliers Email Today
Inform of any challenges due to added weight of new rack design - apply in next month's safety training topics
1-on-1 This week - prior to pilot
Updated Stakeholder AnalysisStakeholder/ Stakeholder
GroupImpact Level Level of Support Reason for
Resistance or Support
Action(s) to Address This Stakeholder
GroupContact
VP-MANUFACTURING DECISION AUTHORITY SUPPORTIVE FINANCIAL BACKING - WILL
HELP CLEAR BARRIERS
UPDATE/NOTIFY IF ANY PURCHASES OR MATERIAL
NEEDEDTIM J.
METAL PROC. MANAGER IMPACTS NEUTRAL
AWARE OF PROBLEM BUT UNSURE EMPLOYEES CAN
HANDLE CHANGES
CREATE WORK INSTRUCTIONS/TRAINING
OPPORTUNITIES EARLY IN THE PROCESS
KLAY M.
PLANT MGR AFFECTED RESISTANTAFRAID CHANGES COULD
LEAD TO NON-VALUE ADDED TIME TO CURRENT PROCESS
USE DATA TO SHOW EXPECTED COST SAVINGS AND POTENTIAL LOSSES IF
NOT ADDRESSED
WES B.
QUALITY TEAM DECISION AUTHORITY SUPPORTIVE
ANXIOUS TO SHOW PROCESS IMPROVEMENT TEAMS CAN
ADD VALUE TO THE ORGANIZATION
CREATE USEFUL TEMPLATES/DOCUMENT STRATEGIES - IDENTIFY
GOOD/BAD LESSONS LEARNED
PAUL K.
MAINTENANCE IMPACTS SUPPORTIVEWANT TO PREVENT ANY TYPE
OF DOWNTIME AND LOST PRODUCTIVITY
PROVIDE LINE ITEMS OF NECESSARY IMPROVEMENTS
AND KEEP SAFETY STOCKVINCENT L.
PURCHASING IMPACTS SUPPORTIVEPOTENTIAL NEW SUPPLIERS
COULD BENEFIT OTHER AREAS OF OUR ORGANIZATION
RESEARCH/STOCK COVERS USED FOR SPACERS AND
DOLLY/RACK ARMSDEREK J.
Stakeholder Analysis (Advanced) - IMPROVE
Implementation PlanAction Item
(List steps required to implement solutions)
Responsible(List person(s) responsible
for action steps)
Due Date(Indicate when action items
must be completed)
Outline Pilot Strategy (Timeframe/Dept/Adjustments) Jeff/Daniel/Kevin 12/8/2015
Order supply of spacer/arm covers with enough safety stock - apply immediately Michael/Purchasing 12/11/2015
Update Work Instructions with Process Changes Jeff/Dexter 12/23/2015
Adjust/Replace Lighting in Inspection Areas Vincent 12/11/2015
Create cross-training matrix and setup training for 2nd shift Jeff/Michael 12/22/2015
Create P-Chart to show Improvement over Baseline Jeff 1/6/2016
Set-up Safety Topic to discuss changes and potential lifting hazards (heavier/wider racks) Jake 12/3/2015
Implementation Plan
Control• Improvements from Baseline (P-Charts)• Data compared to Goal• Monitoring Plan• Response Plan• Visual Management• Innovation Transfer• Lessons Learned
P-Chart (Measure Phase)Process Capability
P-Chart for Process Capability during Measure Phase in October 2015
P-Chart Revisited
28252219161310741
0.4
0.3
0.2
0.1
0.0
Sample
Prop
ortio
n
_P= 0.1347
UCL= 0.3395
LCL= 0
P Chart of DEFECTS
Testsareperformedwithunequalsamplesizes.
The improvement from our baseline back in October was dramatic, dropping our Proportion Defective for Bent, Scratched, and Damaged Material from 0.5627 to 0.1347, shown in the Minitab P-Chart below. This translated to only 101 defects over a lot of 750 pieces, compared to 422 defects during our baseline measure of 750 pieces. I chose the P-Chart due to our racks of material having unequal sample sizes, ranging from 20-30 pieces per rack.
Shattering the GoalOur goal was to reduce Bent/Scratched/Damaged Scrap $$ by 50% from an average of $5850/month (thru August) to $2925/month by the end of the year. We blew that away with savings of over 76% by reducing it to just $1362.02 by the end of 2015. The graph below shows the dramatic decline.
$4,622.06
$8,114.10
$4,877.85
$4,942.94
$5,653.64
$3,854.11
$6,412.27
$8,327.46
$6,177.00
$4,541.79
$2,252.09
$1,362.02
$0.00
$1,000.00
$2,000.00
$3,000.00
$4,000.00
$5,000.00
$6,000.00
$7,000.00
$8,000.00
$9,000.00
Jan-15 Feb-15 Mar-15 Apr-15 May-15 Jun-15 Jul-15 Aug-15 Sep-15 Oct-15 Nov-15 Dec-15
2015- B/S/D- MATERIALHANDLING
SCRAP$$
Monitoring PlanMeasure Control Limits/Specs Checking Item Checking Frequency Who
Bent, Scratched, and Damaged Material (Defects)
No greater than 0.3395 Proportion Defective - Upper Control Limit
Checksheets created in Data Collection Plan
Each rack of material (ranges from 20-30 pieces) in Extrusion, Anodizing, and Paint
Material Handling Team
Monitoring Plan
We want the communication to be open so this Monitoring Plan will be a standard slide used in the MDI (Managing Daily Improvement) walks performed each week for all levels of Plant leadership along with Pre-Shift Meeting documentation for all operators. The expectations are that with the attention this has received and the pride of the employees taking ownership – sustainment of these improvements will come easy.
Response Plan
Damage Control Process Adjustment Effectiveness Assessment Continuous ImprovementConfirm all Employees Trained on Operational Definitions (and Visuals) of Bent, Scratched, Damaged Material
Process Owner to Audit Material on Each shift with employees
No greater than 0.3395 Proportion Defective per rack - UCL setup in the Monitoring Plan
Offset employee turnover with continued Cross-Training and pair new employees with seasoned veterans to ensure proper training habits
Check to see if Racks need repairs or if spacers/arms are properly covered
Gather a safety stock for all maintenance items used for racks
Replace rack if defective and store two months of safety stock (kevlar covers) at point-of-use
Create weekly Preventive Maintenance schedule for Material Handling Teams to sign-off on
Response Plan (If Proportion Defective dips above 0.3395)
Visual ManagementPictures identifying definitions like Bent, Scratched, Damaged Material along with Checksheets to use for Data Collection (Pages 17-20 of this PowerPoint) have been added to the Standard Work Instructions. We’ve also included pictures of the spacer/arm covers, shavings, tiedowns for a visual explanation within the Work Instructions. The next step is to have the Control Charts/Monitoring Plans posted in each department for everyone to see.
Innovation TransferProcess Innovation/Addition/Change/Removal
(Describe the implemented solution)
Process Benefits(Describe the both measurable and intangible benefits
of the change)
Area/Department/Business Unit(Indicate the process area that could
benefit from innovation transfer)
New Rack Design/Spacer Covers/Arm Covers Eliminates B/S/D Opportunities All Metal Processing Facilities in
U.S. and Canada
Cross-Training Matrix Creates Employee Ownership and Strengthens workforce depth All Departments
Control Charts/Monitoring Plans Identifies Trigger Measures (UCL/LCL) that lead to variation All Departments
Visual Aids added to Work InstructionsQuicker training method and immediate
understanding of Operational Definitions like Bent, Scratched, and Damaged Material
All Metal Processing Facilities in U.S. and Canada
New Checksheets for Data Collection An easy-to-use sheet that tracks each rack of material against our goal of Proportion Defective
All Metal Processing Facilities in U.S. and Canada
Innovation Transfer Opportunities
Lessons Learned• Important to drive inspection earlier in the process where Rework is
an Option. The more VA to the Metal – the harder it is to Rework –and the less $$ you get for defective material in a scrapyard
• Too many blind handoffs – wanted to create opportunities where items (Checksheets) had to be initialed to create more ownership
• Newspaper Items– Dig deeper into Scrap Reporting Efficiency– Use Project to show value of adding a 2nd Age Oven in
Extrusion to serve as Workload Balancing and eliminate the obvious bottleneck.
– Sell Lean Culture benefits at the Management Level for ultimate buy-in on Process Improvements
– Identify more KPIs/Metrics for Tracking (Rework/FPY)