D M A I C
Premium Reduction & Delivery Improvement
Champion –Sponsor –
Tollgate Review
Contract Project Title: Premium Reduction
Department PC&L Duration August 2004 - February 2005 Sponsor: Rachel Heidenreich Team Members Function
Eric Crump Ken Dillon Marne Cario Chris Kushmaul Jim Altier Tom Catheline William Katko John Salerno Donnell Conner PC&L Manager - Project Leader PC&L Scheduler - Project Leader PC&L General Supervisor Master Black Belt Master Black Belt Shipping Supervisor Quality Engineer Tool Engineer MFG Supervisor
D M A I C
Total Months: 6 Champion: Marybeth Cunningham Role
Black Belt Candidate Green Belt Candidate Green Belt Re-Certification Project Coach Manufacturing Support Green Belt Candidate Green Belt Candidate Green Belt Re-Certification Green Belt Candidate
Customer (internal/external): internal: PC&L CTQ addressed: Premium / IPM (Internal & External) Business Driver(s) Impacted: Premium Cost Reduction / IPM Internal And External Delivery Improvement Project Scope: In Scope - Reduction of all Premium from Pt 11
Preliminary Problem Statement: Text: Plant has been experiencing high premium transportation over the past 2 years. 2003 premium was over $220,000 and 2004 premium is over $250,000 thru July. In addition to high premium, Plant 11 has experienced internal and external delivery issues to several customers. Delivery Performance has created several concerns from the DCS Business Line because of the impact to their final customers. The goal of this project is to reduce Premium by 40% and reduce IPM by 45% Aug Define actual Measure actual Analyze actual Improve actual Control actual Control and Response Plan ` Process Mapping, Cause and Effect Diagram, Data collection Plan, MSA DPMO/Capability, Hypothesis Testing/Root Cause Analysis Solution generate, Pilot on Department 1108, Moods Median Test, Test for Equal Variance Sep Oct Nov Dec Jan Feb Mar Apr Tools Used Contract / SIPOC / Graphs
Tollgate Review
Plant Premium 200x Detail D M A I C
$1 83 ,4 11 $1 9, 57 1
$200,000 $180,000 $160,000 $140,000 $120,000 $100,000 $80,000 $60,000 $40,000 $20,000 $0
Behind Schedule Internal Quality Supplier Quality Trans / Customs Deliver Miss Late Entry Mislabeled Cust. Req / Inv. Adj Req / Inventory Adj. Part Cert. Delay
$1 1, 18 1
$5 ,5 8
4
$1 ,2 2
4 $7 74
$2 00
$1 08
$6 4
Data was obtained from the Premium Web Site Tollgate Review
$1 24
Plant Premium 200x Detail D M A I C
$2 20 ,0 19 $2 0, 49 2
$250,000
$200,000
$150,000
$100,000
$50,000
$7 ,3 3
8
0
2
$1 ,8 5
$1 ,6 1
$1 ,1 1
0
$6 59
$0
Behind Schedule
Trans / Customs
Data Processing
Engineering
Internal Quality
Deliver Miss
Req / Inventory Adj.
Tool Capacity
Data was obtained from the Premium Web Site Tollgate Review
$3 00
Plant – 200x Internal IPM D M A I C
10,000 9,000 8,000 7,000 6,000 5,000 4,000
9, 25 2 4, 89 4 7, 65 5
3, 83 4 2, 50 7 2, 04 7 2, 16 9 27 0
Jan Feb March April
3,000 2,000
0 May June July Aug Sept Oct Nov Dec YTD
0
Data was obtained from the Premium Web Site Tollgate Review
23 9
26 1
1,000
2, 89 2
Plant 200x – Internal IPM D M A I C
3, 22 5
3,500
3,000
2, 20 6
2,500
2, 08 9
2,000
1, 97 6
1,500
0 Jan Feb March April May June July YTD
Data was obtained from the Premium Web Site Tollgate Review
24 3
500
50 8
79 9
1,000
1, 59 6
SIPOC
Process Name: Process Owner: Suppliers
(Providers of the required resources)
D M A I C
Customers
Plant 11 Terminal Manufacturing Process Plant 11 Premium Reduction Team Inputs Process
(Resources required by the process)
Outputs
(Deliverables from the process)
(Stakeholders who place the requirements on the outputs)
Customer
Forecast • Pulls • Orders Production Schedules PPAP Tool Release Press Release Part Inspection (PPC) Production Tools Set - UP
Customer Requirements Forecast • Pulls (Internal) • Orders (External)
Requirements Parts Shipped on Time TERMINALS Good Quality Parts INTERNAL • Mexico • Mississippi
PC&L Quality
Manufacturing Tool Room • Tool & Die Makers
• Scheduling • Manufacuturing
Right Quantities
Correctly Labeled Tool Engineer Maintenance Support Support Change-over Matrial Handling Shipping Develop Store Goal Raw Material
SHIP from OHIO
EXTERNAL • Lear • Yazaki • PS&G • Condumex
Terminal Store
Suppliers
Tollgate Review
D M A I C
Plant Premium Reduction & Delivery Improvement
Measure
Date: 01 / 06 / 0x
Tollgate Review
Cause & Effect Diagram
PC&L (x) Mtl Movement Delays (x) Lack - Raw Material Manufacturing (x) Uptime 7 Information Systems (x) Label
D M A I C
2
(x) Absenteeism (x) Line Insp. Coverage (x) Care Insp. Coverage (x) Labor Allocation/Job (x) Operator Training
1
(x) Real Time Data (x) Inventory Visibility
1
(x) Partial Reels (x) Terminal Store Picks (x) Lack - Order Fillers 1
1
4
3 (x) Shipping Patterns 1 (x) Lack - Real Time Info
2 (x) Press Availability 4
(x) Critical Parts List
(x) Scrap Mtce. Response Time 6 Spare Equip. Parts Spare Tooling Parts 4
6 Tool Capability
(Goal Changes)
Why do We have missed/late shipments?
(x) Unforseen Orders/Pulls
6
Tool 3 Availability 6 Setup Times Labor Resourcing/Job Ass Changeover Time 6
(x) Increased Pulls
2
(1 - Group w / Un-forcasted orders / pulls?) (x) Pull Variations (x) Goal Changes (x) Defective Raw Customer/Supplier 5
Low Vol. & 4 Obsolete Parts
(Note comment from Red Tags?) (x) FTQ 1
Make vs Buy Unplanned 3 Mtce. Downtime
(x) PPAP (1 - Group with FTQ?) Should add up to 84
Issues Running 2 Precious Metal
(x) Red Tags Quality
3
Eng/Tool Room/Maint.
Tollgate Review
Data Collection Plan
Measure Data Collection Plan Time Period (Y) Joe Bonneau Name of Factor: Operator (1) (2) Trial Part Number Time to fill back order (X1) John Salerno Uptime The uptime of the press while the back order part number is in the machine (X2) Joe Bonneau Set up / Change Over (X3)
D M A I C
(X4) Ken Dillion Press Availability Is an acceptable press available to the run part? (X5) John Salerno Spare Tooling Part Availability The time it takes to get spare toolign when parts are not available
Ken Dillion Tool Availability
Operational Definition
Time to convert good part to good part
Sitting on ready rack
Type of Data: Specific to: P/N's
Continuous
Continuous
Discrete
Dicrete
Continuous
Tollgate Review
Data Collection Plan
Measure Data Collection Plan
Time Period Operator (1) (2)
D M A I C
(X9)
Marne Cario
Low Volume
(Y)
(X6)
(X7)
John Salerno
Tool & Die Response Time
(X8)
Marne Cario
Unforeseen Orders / Pulls
(X10)
Bill Katko
Red Tag
Joe Bonneau Donnell Conner Name of Factor: Trial Part Number
Time to fill back Maintenance order Response Time
Operational Definition
Time from The difference in time Difference when a call is between when a tool between the placed to the has failed & marked on forecast & the time the work the board & when the actual pull is complete machine is running Continuous Continuous Continuous Rounded to the hour Build area checked 1118
Is this part number low Volume? Discrete
Is this part red tag or care?
Type of Data: Specific to: P/N's
Discrete First time? Quality?
Tollgate Review
Gage R & R Study
Repeatability
D M A I C
– Within Appraisers
Appraiser 21
# Inspected 30
# Matched 30
% Percent 100
95.0% CT (90.5 , 100)
2
30
29
96.7
(82.8 , 99.9)
Appraiser agrees with him / herself trials
Reproducibility
– Between Appraisers
# Inspected 30
# Matched 29
% Percent 96.7
95.0% CT (82.8 , 99.9)
All appraisers assessments agree with each other
Tollgate Review
D M A I C
Plant Premium Reduction & Delivery Improvement
Analyze
Date: 03 / 24 / xx
Tollgate Review
Descriptive Statistics
D M A I C
USL = 2 Days
Variable: (Y)
Time a Part is on Backorder (days) Anderson-Darling Normality Test A-Squared: P-Value: 5.269 0.000 3.18609 2.89036 8.35419 1.32074 1.59636 115 0.3000 1.0000 2.3000 5.4000 14.0000
DPMO = 1.4 sigma process
1
4
7
10
13
Mean StDev Variance Skewness Kurtosis N Minimum 1st Quartile Median 3rd Quartile Maximum
95% Confidence Interval for Mu
95% Confidence Interval for Mu 2.6522
1.5 2.5 3.5
3.7200 3.3212
95% Confidence Interval for Sigma 2.5589
95% Confidence Interval for Median
95% Confidence Interval for Median 1.6000 2.7000
The median time for parts on backorder is 2.3 days. With a target of 2 days our process defect rate is 53%
Tollgate Review
Descriptive Statistics
D M A I C
Run Chart for (Y)
14 12 10
(Y)
8 6 4 2 0 10 60 110
Observation
Number of runs about median: Expected number of runs: Longest run about median: Approx P-Value for Clustering: Approx P-Value for Mixtures: 58.0000 58.4957 10.0000 0.4630 0.5370 Number of runs up or down: Expected number of runs: Longest run up or down: Approx P-Value for Trends: Approx P-Value for Oscillation: 76.0000 76.3333 3.0000 0.4704 0.5296
Our process is stable with no clustering, mixtures, trends, and oscillations
Tollgate Review
Identify variation Sources
Factor
X1
D M A I C
Name
Press Uptime
Data Type
Continuous
Analysis Tools
Simple Linear Regression Mood's Median Test Test for Equal Variances Simple Linear Regression Mood's Median Test Test for Equal Variances Mood's Median Test Test for Equal Variances Mood's Median Test Test for Equal Variances Mood's Median Test Test for Equal Variances
Statistically Significant
No
X2
Set Up / Change Over
Continuous
No
X3 X4 X5
Tool Availability Press Availability Spare Tooling Part Availability Mantenance Response Time
Discrete Discrete Discrete
No No No
X6
Continuous
Simple Linear Regression Mood's Median Test Yes - Difference in Center Test for Equal Variances Simple Linear Regression Yes - Difference in Center and Mood's Median Test variation Test for Equal Variances Mood's Median Test Test for Equal Variances Mood's Median Test Test for Equal Variances Mood's Median Test Test for Equal Variances Mood's Median Test Test for Equal Variances No No No No
X7
Tool & Die Response Time Orders vs. Pulls Low Volume Red Tag or Care Start-Stop Part
Continuous
X8 X9 X10 X11
Discrete Discrete Discrete Discrete
Tollgate Review
Identify variation Sources X1 - Uptime
Fitted Line Plot
D M A I C
Y1 = 4.632 - 0.04227 X1 30 25 20
Y1
S R-Sq R-Sq(adj) 3.69418 3.7% 2.9%
Regression Analysis: Y1 versus X1 (Uptime)
The regression equation is Y1 = 4.632 - 0.04227 X1 S = 3.69418 R-Sq = 3.7% R-Sq(adj) = 2.9%
15 10 5 0 0 10 20 30 40 50 X1 60 70 80 90
Analysis of Variance Source Regression Error Total DF 1 120 121 SS 62.52 1637.63 1700.16 MS 62.5237 13.6470 F 4.58 P 0.034
Data analysis shows that machine uptime and time to fill a backorder do not have a strong correlation – not statistical root cause
Tollgate Review
Identify variation Sources X2 – Set Up / Change Over
Y1 = 3.108 + 0.08494 X2 14 12 10 8 6 4 2 0 0 2 4 6 X2 8 10 12 14
S R-Sq R-Sq(adj) 2.95590 0.7% 0.0%
D M A I C
Fitted Line Plot
Mood Median Test: Y speed versus x2 speed
Mood median test for Y speed Chi-Square = 0.02 DF = 1 x2 speed Long Short N<= 10 49 N> 9 47 Median 2.30 2.30 P = 0.899 Individual 95.0% CIs ----+---------+---------+---------+-(--------*------------------------) (-----*-) ----+---------+---------+---------+-1.5 3.0 4.5 6.0
Y1
Q3-Q1 5.60 3.90
Overall median = 2.30
Mood Median Test: Y versus X2
Mood median test for Y Chi-Square = 0.09 DF = 1 X2 No Yes N<= 34 24 N> 35 22 Median 2.40 2.15 Q3-Q1 4.40 4.10 P = 0.761 Individual 95.0% CIs --+---------+---------+---------+---(-------------------*------------) (--------------*-----------------) --+---------+---------+---------+---1.50 2.00 2.50 3.00
Overall median = 2.30
Data analysis shows that Set Up / Change Over Time and time to fill a backorder have no correlation – not statistical root cause
Tollgate Review
Identify variation Sources X3 – Tool Availability
Mood Median Test: Y1 versus X3
Mood median test for Y1 Chi-Square = 0.00 DF = 1 P = 0.948
D M A I C
Boxplot of Y1 vs X3
30 25 20 15
X3 N Y
N<= 7 55
N> 7 53
Median 2.35 2.30
Q3-Q1 5.63 4.30
Individual 95.0% CIs ---+---------+---------+---------+--(--------*--------------------------) (---*--) ---+---------+---------+---------+--1.5 3.0 4.5 6.0
Overall median = 2.30
Y1
Test for Equal Variances: Y1 versus X3
10 5 0 N X3 Y
95% Bonferroni confidence intervals for standard deviations X3 N Y N 14 108 Lower 2.12450 3.32853 StDev 3.06088 3.84031 Upper 5.31933 4.52931
F-Test (normal distribution) Test statistic = 0.64, p-value = 0.361 Levene's Test (any continuous distribution) Test statistic = 0.02, p-value = 0.894
No difference in either the center (median) or spread (variation) between tool Available yes vs. no – not statistical root cause
Tollgate Review
Identify variation Sources X4 – Press Availability
Mood Median Test: Y1 versus X4
Boxplot of Y1 vs X4
30 25 20 15 10 5 0 N X4 Y
F-Test (normal distribution) Test statistic = 3.18, p-value = 0.000 X4 N Y N<= 22 40 N> 16 44 Median 1.65 2.40 Q3-Q1 5.00 4.40 Mood median test for Y1 Chi-Square = 1.11 DF = 1 P = 0.293
D M A I C
Individual 95.0% CIs ---+---------+---------+---------+--(---------*-------------------) (-----------*------------) ---+---------+---------+---------+--1.20 1.80 2.40 3.00
Overall median = 2.30 A 95.0% CI for median(N) - median(Y): (-1.70,0.60)
Y1
Test for Equal Variances: Y1 versus X4
95% Bonferroni confidence intervals for standard deviations X4 N Y N 38 84 Lower 4.11245 2.47510 StDev 5.18848 2.90730 Upper 6.98004 3.51331
Levene's Test (any continuous distribution) Test statistic = 0.84, p-value = 0.360
The p values are greater than .05; we accept the null hypothesis ~ there is no difference in backorder time when a press is available – yes vs. no
Tollgate Review
Identify variation Sources X5 – Spare Tooling Part Availability
Mood Median Test: Y1 versus X5
D M A I C
Boxplot of Y1 vs X5
14 12 10
Mood median test for Y1 Chi-Square = 0.70 DF = 1 X5 N Y N<= 2 54 N> 4 52 Median 4.30 2.30 Q3-Q1 5.83 4.40
P = 0.401 Individual 95.0% CIs -------+---------+---------+--------(-------------*--------------------) (--*-) -------+---------+---------+--------2.5 5.0 7.5
Overall median = 2.35
8
Y1
A 95.0% CI for median(N) - median(Y): (-2.00,9.70)
6 4 2 0 No X5 Yes
F-Test (normal distribution) Test statistic = 1.97, p-value = 0.179 Levene's Test (any continuous distribution) Test statistic = 0.52, p-value = 0.472
Test for Equal Variances: Y1 versus X5
95% Bonferroni confidence intervals for standard deviations X5 N Y N 6 106 Lower 2.34663 2.46978 StDev 4.00221 2.85313 Upper 11.4501 3.3707
Although the medians appear different, there is not enough data to statistically conclude there is a difference – there were only six samples when spare tooling was not available
Tollgate Review
Identify variation Sources X6 – Maintenance Response Time
Mood Median Test: Y1 versus X6 > 4
D M A I C
Boxplot of Y1 vs X6 > 4
14 12 10 8
Y1
Mood median test for Y1 Chi-Square = 8.35 DF = 1 X6 > 4 No Yes N<= 53 5 N> 40 17 Median 1.70 5.70 Q3-Q1 3.00 4.73
P = 0.004 Individual 95.0% CIs -----+---------+---------+---------+(---*--) (---------------*-----) -----+---------+---------+---------+2.0 4.0 6.0 8.0
Overall median = 2.30
6 4 2 0 No X6 > 4 Yes
X6 > 4
Test for Equal Variances for Y1
F-Test No Test Statistic P-Value Test Statistic P-Value 0.67 0.197 2.41 0.124
Lev ene's Test
Yes
2.0
2.5 3.0 3.5 4.0 4.5 95% Bonferroni Confidence Intervals for StDevs
5.0
No
X6 > 4
Yes
Statistical Root Cause – with a p-value of .004 we reject the null hypothesis; there is a difference in times. When the maintenance call takes over 4 hours then the backorder time goes up significantly!!
0 2 4 6 8 10 12 14 Y1
Tollgate Review
Identify variation Sources X7 – Tool & Die Repair Time
Mood Median Test: Y1 versus X7 > 4
Mood median test for Y1 Chi-Square = 11.72 DF = 1
D M A I C
P = 0.001 Individual 95.0% CIs -----+---------+---------+---------+(----*---) (---*--------) -----+---------+---------+---------+1.0 2.0 3.0 4.0
Boxplot of Y1 vs X7 > 4
14 12 10 8
Y1
X7 > 4 No Yes
N<= 22 36
N> 6 51
Median 1.00 2.70
Q3-Q1 1.53 5.00
Overall median = 2.30
Test for Equal Variances for Y1
F-Test No
X7 > 4
6 4 2
Test Statistic P-Value Test Statistic P-Value
1.02 0.897 3.63 0.059
Lev ene's Test
Yes
2.0
0 No X7 > 4
No
X7 > 4
2.5 3.0 3.5 95% Bonferroni Confidence Intervals for StDevs
4.0
Yes
Yes
Statistical Root Cause – with a p-value of .001 we reject the null hypothesis; there is a difference in times. When the maintenance call takes over 4 hours then the backorder time goes up
Y1
0
2
4
6
8
10
12
14
Tollgate Review
Identify variation Sources X8 – Orders vs. Pulls
Mood Median Test: Y1 versus X8
D M A I C
Boxplot of Y1 vs X8
14 12 10 8
Y1
Mood median test for Y1 Chi-Square = 0.01 DF = 1 X8 No Yes N<= 28 30 N> 28 29 Median 2.35 2.30 Q3-Q1 3.55 4.40
P = 0.928 Individual 95.0% CIs ---+---------+---------+---------+--(----------*--) (---------------*-----------------) ---+---------+---------+---------+--1.40 2.10 2.80 3.50
Overall median = 2.30
6 4 2 0 No X8 Yes
No
Test for Equal Variances for Y1
F-Test Test Statistic P-Value Test Statistic P-Value 0.76 0.299 1.21 0.273
Lev ene's Test
X8
Yes
2.0
2.5 3.0 3.5 95% Bonferroni Confidence Intervals for StDevs
4.0
No
X8
Yes
0
2
4
6 Y1
8
10
12
14
No difference in either the center (median) or spread (variation) when we ask If the pull was greater than the order, yes vs. no – not statistical root cause
Tollgate Review
Identify variation Sources X9 – Low Volume
Mood Median Test: Y1 versus X9
D M A I C
Boxplot of (Y) vs (X9)
14 12 10 8 6 4 2 0 N (X9) Y
Mood median test for Y1 Chi-Square = 1.01 DF = 1 X9 N Y N<= 53 9 N> 53 5 Median 2.35 1.85 Q3-Q1 4.10 6.98
P = 0.315 Individual 95.0% CIs -------+---------+---------+--------(---*-) (-----*--------------------------) -------+---------+---------+--------2.0 4.0 6.0
Overall median = 2.30
(Y)
Test for Equal Variances for (Y)
F-Test N Test Statistic P-Value Test Statistic P-Value 0.35 0.007 1.17 0.282
(X9)
Lev ene's Test
Y
2
3
4 5 6 7 95% Bonferroni Confidence Intervals for StDevs
8
9
N
(X9)
Y
0
2
4
6 (Y)
8
10
12
14
No difference in either the center (median) or spread (variation) when we ask If the part is low volume, yes vs. no – not statistical root cause
Tollgate Review
Identify variation Sources X10 – Red Tag or Care
Mood Median Test: Y1 versus X10
D M A I C
Boxplot of Y1 vs X10
14 12 10 8
Y1
Mood median test for Y1 Chi-Square = 1.31 DF = 1 X10 No Yes N<= 21 37 N> 15 42 Median 1.85 2.40 Q3-Q1 4.30 4.40
P = 0.253 Individual 95.0% CIs ------+---------+---------+---------+ (-----------*-------------) (----------*------------) ------+---------+---------+---------+ 1.40 2.10 2.80 3.50
Overall median = 2.30
6
Test for Equal Variances for Y1
4 2
X10
F-Test No Test Statistic P-Value Test Statistic P-Value 1.17 0.567 0.01 0.941
Lev ene's Test
0 No X10 Yes
Yes
2.50
2.75 3.00 3.25 3.50 3.75 4.00 95% Bonferroni Confidence Intervals for StDevs
4.25
No
X10
Yes
0
2
4
6 Y1
8
10
12
14
No difference in either the center (median) or spread (variation) when we ask If the material had any pieces in redtag, yes vs. no – not statistical root cause
Tollgate Review
Identify variation Sources X11 – Start/Stop Part
Mood Median Test: (Y) versus X12
D M A I C
Boxplot of (Y) vs X12
14 12 10
Mood median test for (Y) Chi-Square = 3.28 DF = 1 X12 No Yes N<= 29 29 N> 19 38 Median 1.50 2.40 Q3-Q1 4.60 4.40
P = 0.070 Individual 95.0% CIs ---+---------+---------+---------+--(-------*--------------------------) (-----------*---------) ---+---------+---------+---------+--1.20 1.80 2.40 3.00
Overall median = 2.30
(Y)
8 6 4
X12
No
Test for Equal Variances for (Y)
F-Test Test Statistic P-Value Test Statistic P-Value 1.49 0.131 0.51 0.476
Lev ene's Test
2 0 No X12 Yes
Yes
2.0
2.5 3.0 3.5 95% Bonferroni Confidence Intervals for StDevs
4.0
No
X12
Yes
0
2
4
6 (Y)
8
10
12
14
No difference in either the center (median) or spread (variation) when we ask If this PN was a start/stop part, yes vs. no – not statistical root cause
Tollgate Review
Analyze Take - Aways
-
D M A I C
Maintenance Response Time is a significant contributor to Time to Fill Back Order (Y) Tool & Die repair time is a significant contributor to Time to Fill Back Order (Y) Both Maintenance Response Time and Tool & Die Repair Time impact Uptime but Uptime alone is not a significant factor Other potential causes (X’s) did not show significant impact to Time to Fill Back Order This statistical data indicates that in order to reduce the Time to fill a Back Order, which potentially leads to premium, we must reduce the Maintenance Response Time and the Tool & Die Repair Time
-
-
-
-
Tollgate Review
Next Steps
-
D M A I C
Continue in the Improve Phase Develop solutions to improve Maintenance Response Time and Tool & Die Repair Time Collect Data from the Recommended improvement plan Evaluate if corrective actions reduce or eliminate the impact of Maintenance Response Time and Tool & Die Repair Time on the Time to Fill a back Order / Premium
-
Tollgate Review
D M A I C
Plant Premium Reduction & Delivery Improvement
Improve / Control
Date: 05 / 05 / xx
Tollgate Review
Develop Solutions
-
D M A I C
Tool & Die Repair Time Solutions
Dedicated Tool & Die support on all Shifts Use the Andon System to improve call priorities Die Maker inspects and releases parts in place of a traditional release person Document all Die changes in the log book Dedicated tool makers for different die families
-
Maintenance Response Time Solutions
Dedicated Machine and Maintenance Repair support and on all shifts Use the Andon System to improve call priorities PC&L notifies maintenance when a part goes on back order Maintenance Supervisor copied on all critical parts
Tollgate Review
Test Solutions - Pilot
Results for: 1108 vs. 1119 Mood Median Test: Y
Mood median test for Y Chi-Square = 60.60 DF = 1 Subscripts 1119 1108 N<= 62 789 N> 150 574 Median 2.40 1.40 P = 0.000 Q3-Q1 2.68 1.80 Individual 95.0% CIs +---------+---------+---------+-----(-------*-----) (-------* +---------+---------+---------+-----1.00 1.50 2.00 2.50
D M A I C
Overall median = 1.40
Test for Equal Variances for Y
A 95.0% CI for median(1119) - median(1108): (0.60,1.30)
Subscripts
F-Test Test Statistic P-Value Test Statistic P-Value 1.38 0.001 16.49 0.000
1119
Test for Equal Variances: Y
95% Bonferroni confidence intervals Subscripts 1119 1108 N 212 1363 Lower 2.01797 1.82901 StDev 2.23877 1.90766 Upper 2.51152 1.99312
Lev ene's Test
1108
1.8
1.9
2.0 2.1 2.2 2.3 2.4 95% Bonferroni Confidence Intervals for StDevs
2.5
Subscripts
F-Test (normal distribution) Test statistic = 1.38, p-value = 0.001 Levene's Test (any continuous distribution) Test statistic = 16.49, p-value = 0.000
1119
1108
0
5
10 Y
15
20
25
The p values are less than .05; we reject the null hypothesis ~ there is a difference in backorder. The solutions have impacted both variation and median
Tollgate Review
New Capability Analysis
D M A I C
Capability Analysis
Characteristic Defs Units Opps TotOpps DPU DPO PPM ZShift ZBench
1108
414
1363
1
1363
0.304
0.303742
303742
1.500
2.014
1119
117
212
1
212
0.552
0.551887
551887
1.500
1.370
Total
531
1575
0.337143
337143
1.500
1.920
Applying the goal of less than 2 days on backorder the solutions implemented improvement our process from 1.4 sigma to a 2 sigma process.
Tollgate Review
Timeline for Full Solution Implementation
-
D M A I C
Department 11xx was the first module installed. Implementation of the remaining modules that will utilize these solutions are as follows:
- Department 11xx will be completely installed by 5/15/xx - Department 11xx will be completely installed by 6/1/xx - Department 11xx will be completely installed by 8/1/xx - Department 11xx (Low Volume) will be completely installed by 10/xx - Department 11xx will be replaced by 11xx (8/1/xx) - Department 11xx will be replaced by 11xx (6/1/xx)
Tollgate Review
Financial Impact
-
D M A I C
January – April of 200x compared to January – April 200x indicates a reduction in premium of 16%, 80%, 84.5% and 31.4% respectively.
- This equates to a 61.9 % reduction in premium over the same four month period.
-
Over the last two months departments without the improvements have accounted for significantly more premium than department 110x.
- March - April Dept 110x 18.0 % 27.8 % Non-Improved Depts 52.0 % (1119) 58.9 % (1116)
-
Our implemented solutions improved our process by 30 %
- From 1.4 Sigma to a 2.0 Sigma
-
Our implemented solutions reduced our median time to fill a back order by 41.6 %
- From a median of 2.4 to a median of 1.4
-
Based on these improvements we conclude that the plant 11 premium will be reduced by approximately 34.0 %
-
Based on 200x premium of $308,940, the financial impact will be a reduction in premium of $105,040.00 per year.
Tollgate Review
Control Plan
-
D M A I C
Plant xx team should monitor Maintenance Response Time and tool & Die Repair time.
- This should be monitored on a monthly basis.
-
If Maintenance Response Time or Tool & Die Repair Time is greater than 4 hours, plant should investigate root cause Need to determine responsible person (s) to monitor these issues System generated reports for Maintenance Response Time and Tool & Die Repair Time will support in this effort
-
-
Tollgate Review
Intangibles
-
D M A I C
Breaking down “old” barriers and avoiding finger-pointing by using real data collected by a true cross-functional team
- Excellent cross functional team that was represented by manufacturing, PC&L, tool engineer, Quality, IT, Terminal Store Supervisor and Manufacturing & PC&L Master black belts
-
Dispelled many theories about causes for premium and allowed us to focus on real root cause Provided additional data for the manufacturing team to focus on the real causes for premium and other machine down time / Uptime issues Processes across all functions are interrelated but not all controlled by the plant manager
- Good cross functional support is required for the plant to be successful
-
-
-
Excellent job by IT support to provide accurate real time data.
- Every black belt project should have an IT person on the team
-
Excellent data collection and support by all team members.
Tollgate Review
D M A I C
Back-up
Tollgate Review
Plant xx- Dept. xxxx Premium in 200x D M A I C
$16,000
$14,000
$12,000
$10,000
7
$1 1, 42 0
$6 ,8 3
$8,000
$6,000
$6 ,8 3
7
$1 4, 81 4
$2 ,5 8
$2 ,3 8
$4,000
$2 ,6 9
2
$3 ,6 0
0
4
2
2
$1 ,5 5
$6 78
be r em D ec
$0
A pr il ay Ju ne ar ch ra ry Ju ly ry t A ug us be r er Fe bu te m em be r nu a ct ob M M
Ja
Data was obtained from the Premium Web Site Tollgate Review
Se p
N ov
O
$5 12
$2,000
$1 ,5 7
6
Plant xx- Dept. xxxx Premium in 200x D M A I C
$2,500
$2 ,1 6
8
$2,000
$1,500
$1,000
$500
$0
January Feburary March April
$4 56
$1 ,0 5
6
Data was obtained from the Premium Web Site Tollgate Review
$1 ,8 4
7