analyze tollgate
DESCRIPTION
Lean Six Sigma project phase three.TRANSCRIPT
ANAL
YZE
Lean Six SigmaImproving FTX/STX2 Tank
Draw QualitySFC Henry, Don H. II
Project Initiation Date: 31/03/08
Analyze Tollgate Date: 03/07/08
AnalyzeAnalyze
v 2.0
Agenda Project Charter and Measure Phase Review
Critical X’s
Potential Root Causes Affecting Critical X’s
Reducing the List of Potential Root Causes
Root Cause Analysis (Qualitative)
Impact of Root Causes on Key Outputs (Y)
Prioritized Root Causes
Analyze Summary
Lessons Learned
Barriers/Issues
Next Steps
Storyboard
AnalyzeAnalyze
v 2.03
Analyze – Executive Summary
Improve tank maintenance quality by giving the 1/16 soldiers more time to perform maintenance during the draw.
The project starts at the FTX/STX2 T-6 IPR and ends when the tanks are ready for HETT transport. This project is contained within the Fort Knox Garrison and can transfer to other training support missions on Fort Knox.
We are feeling the pain in training and tank maintenance.
Soldiers fail to do a quality PMCS for the lack of time, training, and command emphasis.
AnalyzeAnalyze
v 2.0
Project Charter Review
Scope: this process begins with the T-6 IPR and ends when 1/16 loads the tanks on HETT’s.
Goal: Improve tank draw quality
Problem/Goal Statement
Tollgate Review Schedule
Business Impact
Core Team
State financial impact of project Expenses-none Investments-none Revenues-potential savings in time 819 hours per year
Non-Quantifiable Benefits are increased tank maintenance quality, soldiers morale, maintenance fault tracking, and less training time lost.
PES Name MAJ Mackey, Andre PS Name MAJ Mackey, Andre DD Name LTC Naething, Robert GB/BB Name SFC Henry, Don MBB Name Nathan SpragueCore Team Role % Contrib. LSS
Training CW2 Warren SME 20% none MAJ Aydelott SME 20% none MAJ Mackey SME 20% none SSG Jones SME 10% none CW4 Lucy SME 10% none SFC Henry BB 100% BB
Tollgate Scheduled Revised Complete
Define: 04/30/08 - 04/29/08
Measure: 05/14/08 04/06/08 04/06/08
Analyze: 06/13/08 07/13/08 XX/XX/08
Improve: 07/18/08 08/13/08 XX/XX/08
Control: 08/23/08 09/13/08 XX/XX/08
Reduce rework during tank draw from 90% to 45%, per FTX/STX2 by 1 October 2008.
Improve 5988-E fault tracking during tank draw from 10% to 85%, per FTX/STX2 by 1 October 2008.
Improve tank bumper number accuracy from 10% to 90%, during the T-2 preparation week by 1 October 2008.
Problem Statement Soldiers of 1/16 express dissatisfaction with the Unit Maintenance Activities M1 series tank quality prior to mission support. Currently, 90% of the tanks drawn require maintenance for mission readiness. Approximately 10% of faults listed on the 5988-E ‘s completed by soldiers are tracked by UMA. Lastly, tank bumper number accuracy during T-2 is currently at 10% which causes excess work in the last days of the mission support draw.
AnalyzeAnalyze
v 2.05
Baseline Data The current tank draw
process has a non-normal distribution
The mean time to draw one tank is .56 or 34 minutes
The tank draw range is .25 hours (15 minutes) to 2 hours (120 minutes) and the standard deviation is .4 (24 minutes)
The mean number of 5988-E’s updated by UMA is .1 or 10%
The average Tank Draw Time is 34 minutes +/- 24 minutes.
AnalyzeAnalyze
v 2.0
Baseline Data Cont.
33% of tanks presented to draw are not ready for issue.
10% of 5988-E faults annotated by soldiers during tank draw is updated by UMA clerks.
67% of tank bumper numbers presented to 1/16 at T-2 by UMA is actually drawn for mission support.
6
These numbers take into account vehicles presented to draw but never actually drawn or PMCSed.
These numbers represent what was actually given, PMCS’ed, and drawn.
AnalyzeAnalyze
v 2.0
Critical X’s: Cause and Effect Matrix
Cause and Effect Matrix
Key Process Output Variables
Customer Importance
10
9 2 6 8
Customer Rank
1 2 3 4 5
1 2 3 4 5 6 7 8 910
11
12
13
14
15
Process Step KPIV
accurat
e t
ank li
st
5988-
E
QA/
QC
DA For
m
2062
Dispatch
Rank
Rating Total
Process Steps & Key Process Input
Variables
1 T-1 RATSS 1 9 9 9 1 2 7.533 171
2 PMCSTechnical Manual
9 1 1 9 9 1 10 227
3 QA/QCUMA inspector
9 1 1 3 3 3 6.3 143
4tank sign over
DA 2062 9 1 1 1 3 4 5.771 131
5tank dispatch
5987-E 9 1 1 1 1 5 5.066 115
###
#####
AnalyzeAnalyze
v 2.08
Potential Root Causes: C & E Diagram
Effect:
The tank draw takes too long.
ManMachine
Material Method
Spread thinly across multiple tasks
Shortage of UMA maintenance personnel
Deadlines, AOAP, Service Schedule, affect # of tanks available
Tanks already in use by other units/missions
BII draw uses excessive people and excess time
RATTS request is not referenced by UMA to assign accurate bumper number list
Tanks are PMCS’d
Tanks are QA/QC’d
Tanks are dispatched
Excessive delays from lack of UMA personnel
5988-E not updated by UMA
AnalyzeAnalyze
v 2.09
Potential Root Causes: FMEAProcess
Step / Input
Potential Failure Mode
Potential Failure Effects
SEVERITY
Potential CausesOCCURRENCE
Current Controls
DETECTION
RPN
What is the
process step and
Input under
investiga-tion?
In what ways does the Key
Input go wrong?
What is the impact on the Key Output Variables (Customer
Requirements)?
What causes the Key Input to go
wrong?
What are the existing controls and procedures
(inspection and test) that prevent either the cause or the Failure Mode?
T-1 bumper
number listnot accurate excessive delays 7
lack of organization
7 none 7 343
PMCS not updated rework 7 lack of personnel 6 Army Policy 5 210
QA/QC not timely rework 4lack of
maintenance4 EXSOP/Army policy 4 64
tank sign over
already issued rework 7lack of
organization2 Army Policy/EXSOP 2 28
tank dispatch
does not go wrong
no problems 1 no problems 1 EXSOP 5 5
AnalyzeAnalyze
v 2.010
Reducing List of Root Causes: Pareto Analysis
Track able causes contained over 90% of the Defects. Our project will focus on tracking vehicle maintenance status.
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
Deadlined Services Due QA/QC Needed Already Issued
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
AnalyzeAnalyze
v 2.0
Root Cause Analysis: Non-Value Add Analysis
1111
QAQC
Maintenance leader
Dispatch
Soldier
Issues bumper number list to soldier
Maintenance leader checks 5988-E and verifies
faults/makes repairs if needed
Hand receipt
Vehicle signed over to soldier
Avg.Delay
2 hoursAvg.Delay
15 min
Avg. Delay
90 min
Soldier conducts PMCS and completes 5988-E, turns it in to maintenance leader
Passes
QAQC
Receives signed QAQC sheet
Vehicle dispatched to
soldier
YESNO
NVA time is in dark blue
Total delay time is 3.75 hours
Retrieves info from
RATSS system
Notify UMA of the # of tanks
needed
AnalyzeAnalyze
v 2.012
Root Cause Analysis: HistogramThe outlier was a vehicle issued that was actually NMC and required 90 minutes to repair.
The vehicle that required 60 minutes was actually dispatched to another unit.
Two of the five that required 45 minutes of work were deadline with a third needing a QAQC from UMA
5.25 hours were spent doing rework that is non value added
AnalyzeAnalyze
v 2.0
One-Way ANOVA of Time and Defects
The data is distributed non-normally with an outlier shown here
The variance in the data is also constant but there are no systematic effects due to collection order or time.
AnalyzeAnalyze
v 2.0
One-Way ANOVA of Time and Defects Data
14
Source DF SS MS F P
DEFECT 3 4915 1638 5.04 0.012
Error 16 5199 325
Total 19 10114
S = 18.03 R-Sq = 48.59% R-Sq(adj) = 38.95%
Individual 95% CIs For Mean Based on Pooled StDev
Level N Mean StDev -------+---------+---------+---------+--
D 4 63.75 37.50 (-------*------)
I 1 60.00 * (--------------*--------------)
N 14 26.79 8.68 (---*---)
Q 1 45.00 * (--------------*--------------)
-------+---------+---------+---------+--
25 50 75 100
Pooled StDev = 18.03
The R-squared value of 48.59% is statistically significant meaning the model predicts nearly half of the variation causing increased tank draw times as being caused by defects.
Therefore we reject the null hypothesis.
AnalyzeAnalyze
v 2.0
Mood’s Median Test The medians
may tell a more complete story. The outlier falsely inflates the averages, this test omits outliers.
Based on the P value, two or more medians are significantly different and we reject the null hypothesis
Mood Median Test: TIME versus DEFECT
Mood median test for TIMEChi-Square = 13.37 DF = 1 P = 0.000
Individual 95.0% CIsDEFECT N<= N> Median Q3-Q1 -----+---------+---------+---------+-D 0 4 45 56 *------------------------)I 0 1 60 Not UsedN 13 1 30 15 (----*Q 0 1 45 Not Used -----+---------+---------+---------+- 30 60 90 120
Overall median = 30* NOTE * Levels with < 6 observations have confidence < 95.0%
AnalyzeAnalyze
v 2.0
Tukey’s Pairwise Comparison
16
One-way ANOVA: TIME versus DEFECT
Tukey 95% Simultaneous Confidence IntervalsAll Pairwise Comparisons among Levels of DEFECT
Individual confidence level = 98.87%
DEFECT = D subtracted from:
DEFECT Lower Center Upper --------+---------+---------+---------+-I -61.47 -3.75 53.97 (----------*-----------)N -66.23 -36.96 -7.70 (-----*----)Q -76.47 -18.75 38.97 (----------*-----------) --------+---------+---------+---------+- -50 0 50 100
DEFECT = I subtracted from:
DEFECT Lower Center Upper --------+---------+---------+---------+-N -86.65 -33.21 20.22 (---------*----------)Q -88.01 -15.00 58.01 (--------------*--------------) --------+---------+---------+---------+- -50 0 50 100
DEFECT = N subtracted from:
DEFECT Lower Center Upper --------+---------+---------+---------+-Q -35.22 18.21 71.65 (----------*---------) --------+---------+---------+---------+- -50 0 50 100
Statistically significant factors are in RED
legendI= issued alreadyN= no defectsQ= need QAQCD=deadlined
Deadlined tanks are statistically significantly different in terms of time and defects
Tanks with no defects are statistically significantly different in terms of time and defects
AnalyzeAnalyze
v 2.0
Current Process Capability
The average issue time was 36 minutes
The target issue time was 30 minutes
The variability of the process is greater than the specification limits (Cpk<1.33)
The process is not meeting customer expectations
AnalyzeAnalyze
v 2.0
One-Way ANOVA Boxplot Deadlines
were the largest time wasters with a mean time of 63 min.
Deadlines also represented the largest range of values
The mean time for the tank that was already issued had a mean of 60 min.
AnalyzeAnalyze
v 2.0
A Different View The same
data classified as tanks that are ready “R” and not ready “NR”
The total NR time is 340 minutes
The total ready time is 480 minutes
The defects make up 40% of the time spent on the tank draw!
AnalyzeAnalyze
v 2.0
DELETE ME after tollgate review
N= no defect D= deadlined Q=need qa/qc I=issued already
DEFECT TIME
N 15
N 15
N 15
N 15
N 30
N 30
N 30
N 30
N 30
N 30
N 30
N 30
N 30
D 45
D 45
Q 45
D 45
N 45
I 60
D 120
Data used in minitab to get calculations, I later changed the D,Q, & I variables into Defects to create different comparisons of defects vs no defects
AnalyzeAnalyze
v 2.021
Impact of Root Causes on Y
AnalyzeAnalyze
v 2.0
Prioritized Root CausesEffect (Y) Root Cause
(X)Hypothesis for Relationship
In/Out of Team’s Control1
Impact2 Score (Control x Impact)
Priority of Effort
rework In-accurate bumper numbers
Accurate bumper numbers will increase throughput
3 9 27 2
Poor tank maintenance
PMCS not completed correctly
Correctly performed PMCS will improve tank draw quality
9 9 81 1
Poor tank maintenance
5988-E’s are not regularly updated
Regularly update 5988-E’s will improve tank draw quality
3 9 27 3
rework Tanks are issued that are not ready for issue
Rework will be reduced if the tanks are ready for issue at the time they are to be issued to units
3 3 9 4
1 In team’s Control = 9; In team’s sphere of influence = 3; Out of team’s control = 12 High impact = 9; Medium impact = 3; Low impact = 1
AnalyzeAnalyze
v 2.023
Analyze SummaryImpact of Root Causes:
Hypothesis Tests
Tools Used
Reducing List of Root Causes
Prioritized Root Causes / Effects Root cause #1: No visual tracking method
Effect-in-accurate bumper numbers
Root cause #2: 5988-E’s not completed correctly
Effect-poor tank maintenance
Root cause #3: 5988-E’s are not updated regularly
Effect-poor tank maintenance
Root cause #4: Tanks issued that are not ready for issue
Effect-rework
Value Add Analysis Pareto Plot Histogram One-Way ANOVA
C&E Matrix Cause & Effect
Diagram FMEA Process Capability
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
Deadlined Services Due QA/QC Needed Already Issued
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
AnalyzeAnalyze
v 2.024
Lessons Learned Application of Lean Six Sigma Tools
Communications
Team building
Organizational activities
Other
AnalyzeAnalyze
v 2.025
Barriers/Issues/Project Action Log Resources
Unexpected delays
Team or organizational issues
Updated risk analysis and mitigation plan
Revised project scopeLean Six Sigma Project Action Log
Last
Revised: 10/15/2007
No Description/ Recommendation
Status Open/Closed/
Hold
Due Date Revised Due Date
Resp. Comments / Resolution
1 Leave of critical team member 14 May 4 May
2
3
4
5
AnalyzeAnalyze
v 2.026
Next Steps Outline activities for Improve Phase
Planned Lean Six Sigma Tool use
Barrier/risk mitigation activities
AnalyzeAnalyze
v 2.0
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
Deadlined Services Due QA/QC Needed Already Issued
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Analyze StoryboardDefine
Problem: Poor tank quality at issue
Goal: Improve tank quality at issue, Reduce rework, improve fault tracking
Non-quantifiable BenefitsMorale, tank maintenance, fault tracking
Project CharterT-6 IPRT2T
RATSS
T-5T2T
T-4T2T
T-3T2T
T-2IPR vehicleBumper #s
Given to unit
T-1Tank draw,
HETT,5988-E update
Measure
BII draw measured Tank draw measured 5988-E updates measured
RIE Baselines collected
Critical X’s Identified Potential Root Causes Identified Root Causes Prioritized
Analyze
AnalyzeAnalyze
v 2.0
Sign Off
• I concur that the Measure phase was successfully completed on 03/07/08.
• I concur the project is ready to proceed to next phase: Improve
CW4 David LucyResource Manager/Finance
COL Leopoldo Quintas Deployment Director
SFC Don H. Henry II Black Belt
Nathan SpragueMaster Black Belt
MAJ Andre L. Mackey Sponsor / Process Owner
AnalyzeAnalyze
v 2.029
Analyze Tollgate Checklist
Has the team examined the process and identified potential bottlenecks, disconnects, and redundancies that could contribute to the problem statement?
Has the team analyzed data about the process and its performance to help stratify the problem, understand reasons for variation in the process, and generate hypothesis as to the root causes of the current process performance?
Has an evaluation been done to determine whether the problem can be solved without a fundamental ‘white paper’ recreation of the process? Has the decision been confirmed with the Project Sponsor?
Has the team investigated and validated (or devalidated) the root cause hypotheses generated earlier, to gain confidence that the “vital few” root causes have been uncovered?
Does the team understand why the problem (the Quality, Cycle Time, or Cost Efficiency issue identified in the Problem Statement) is being seen?
Has the team been able to identify any additional ‘Quick Wins’?
Have ‘learnings’ to-date required modification of the Project Charter? If so, have these changes been approved by the Project Sponsor and the Key Stakeholders?
Have any new risks to project success been identified, added to the Risk Mitigation Plan, and a mitigation strategy put in place?
Tollgate ReviewTollgate Review
StopStop Has the team identified the key factors (critical X’s) that have the
biggest impact on process performance? Have they validated
the root causes?
Deliverables: List of Potential Root
causes
Prioritized List of Validated Root Causes
Additional “Quick Wins”, if applicable
Refined Charter, as necessary
Updated Risk Mitigation Plan
Green Belt/Black Belt Actions:
Deliverables Uploaded in PowerSteering
Deliverables Inserted into the Project “Notebook” (see Deployment Director)
AnalyzeAnalyze
v 2.0
DMAIC Methodology - Analyze
Identify PotentialRoot Causes
Reduce List ofPotential Root
Causes
PrioritizeConfirmed Root
Causes
Estimate Impactof Root Causeson Key Output
Metrics
ConfirmRelationship
between PotentialRoot Cause andOutput Metrics
Refine RiskManagement
Plan
List ofConfirmed
Root Causes
Risk Mitigation Plan
Conduct 'Analyze'Tollgate Review
Quantitative Data Analysis Cause & Effect Matrix/FMEA Qualitative Process Analysis
Brainstorming Cause & Effect (Fishbone)
Diagrams 5 Whys Value Stream Map
Analysis Spaghetti Diagram
Statistical Analysis Comparison of Means - t-test - Paired t-test - ANOVA Comparison of
Proportions - Chi Square - 1-Proportion & 2-
Proportion Test Simple & Multiple Linear
Regression Components of Variation DOE Taguchi Methods
Other Analyses Time Study Visual Inspection Process Complexity Queuing Theory Process Knowledge Process Cycle Time/
Efficiency Operation Load Analysis Labor Skills Flexibility
Risk MitigationSpreadsheet
Simulation Benchmarking Industry Stds Estimates Based
on ExtensiveProcessKnowledge
Limited Pilot DOE Taguchi Methods Regression
Analysis
Root CauseConfirmation
PlanReduced
List of PotentialRoot Causes
List ofPotential Root
Causes
Prioritized Listof ValidatedRoot Causes
Estimated IndividualReductions
in Key Y MetricsThrough Elimination
of IndividualRoot Causes
Additional‘QuickWins’
Root Cause Analysis Pareto Analysis Statistical Analysis
Refine ProjectScope, asNecessary
Project Charter Project Tracking
System (PTS)
RefinedCharter in PTS
Revised Charter Project Presentation Storyboard Risk Mitigation Plan