500 likes | 755 Views
Salvage Sale Accuracy Project. Six Sigma Black Belt Training Project. GENCO OVERVIEW. FOUNDED 1898 THE OLD BLIND HORSE AND A CART STORY 90+ FACILITIES – 21,000,000 SQ FT OF WAREHOUSE SPACE INNOVATORS OF THE RETURNS PROCESSING TECHNOLOGY 2 PROPRIETARY WAREHOUSE MANAGEMENT SYSTEMS
E N D
Salvage Sale Accuracy Project Six Sigma Black Belt Training Project
GENCO OVERVIEW • FOUNDED 1898 • THE OLD BLIND HORSE AND A CART STORY • 90+ FACILITIES – 21,000,000 SQ FT OF WAREHOUSE SPACE • INNOVATORS OF THE RETURNS PROCESSING TECHNOLOGY • 2 PROPRIETARY WAREHOUSE MANAGEMENT SYSTEMS • 5 BUSINESS UNIT STRUCTURE ASSET RECOVERY • ASSIST CUSTOMERS IN LIQUIDATING SALVAGE MERCHANDISE • 3 MAIN CHANNELS • B2B • B2C • TRADITIONAL BULK SALVAGE • IN 2003 LIQUIDATED $ 538 MILLION DOLLARS WORTH OF MERCHANDISE
BACKGROUND SLP REWAREHOUSE SCANNING ZONE SORT
THE BREAKTHROUGH STRATEGY DEFINE MEASURE ANALYZE IMPROVE CONTROL • Select Output Characteristics • Define Performance Standards • Validate the Measurement System • Establish Process Capability • Define Performance Objectives • Identify Variation Sources • Screen Potential Causes • Discover Variable Relationships • Establish Operating Tolerances • Validate Measurement System • Determine Process Capability • Implement Process Controls
DEFINE • Select Output Characteristics • Define Performance Standards • Validate the Measurement System • Establish Process Capability • Define Performance Objectives • Identify Variation Sources • Screen Potential Causes • Discover Variable Relationships • Establish Operating Tolerances • Validate Measurement System • Determine Process Capability • Implement Process Controls
PROJECT DEFINITION Overview • PROBLEM STATEMENT • It has been perceived that there is an issue with the accuracy of the pull down and shipping functions for salvage product. This inaccuracy is preventing the Asset Recovery group from increasing their prices and capitalizing upon earlier improvements to the receiving and put away functions. • PROJECT TEAM • BLACK BELT ANN SAMLER • CHAMPION PETE RECTOR • FINANCE LARRY SHOENENBERGER • BU HEAD DAN EISENUTH • PROCESS OWNER DAVE SHIPCOTT • TEAM MELISSA CULLEN • CHRIS SCHNEIDER • NAGESH COWLAGI
PROJECT DEFINITION Background Information METRIC % of Defects vs. Sales CTSs • CTQ – Increased accuracy of shipments vs. BOL • CTC – Reduced claims • CTC – Reliability of BOL vs. Actual DEFECT DEFINITION • Defect = any discrepancy of actual shipment to BOL. Can include short products, over products, or incorrect product. OBJECTIVE • Improve accuracy of salvage shipments PROJECTED BENEFITS • Increased accuracy translates to a reduction in total claim dollars paid • Reductions in claim dollars paid results in reduction in non-value added time spent researching and rectifying defects. • Increased accuracy provides an opportunity to increase the cost per dollar paid for the shipment and allows for an increase in both recovery dollars for customers and an increase in commission dollars for GENCO.
PROJECT DEFINITION Goals and Benefits
APPROVED PROJECT CHARTER ORIGINAL ON FILE IN SIX SIGMA DATABSE
Salvage defects from those facilities where GENCO handles the Returns Processing as a whole Kmart Georgia Illinois Nevada Pennsylvania Auction Salvage defects from each of these individual facilities Sears California Georgia Ohio Auction AREAS OF FOCUS
MEASURE • Select Output Characteristics • Define Performance Standards • Validate the Measurement System • Establish Process Capability • Define Performance Objectives • Identify Variation Sources • Screen Potential Causes • Discover Variable Relationships • Establish Operating Tolerances • Validate Measurement System • Determine Process Capability • Implement Process Controls
DATA COLLECTION • EXCEL SPREADSHEET • ORACLE BOL DATA • CATEGORICAL • TRACKING REASONS • MISSING PIECES • OVERAGES • ALTERNATIVE PRODUCT • MISSHIPPED • SYSTEMS • MISSCANNED • LOSS PREVENTION
MEASUREMENT SYSTEM ANALYSIS Further Review • NO GAGES OR MEASUREMENT “TOOLS” ARE UTILIZED IN THE PROCESS (IE: TAPE MEASURES, CALIBERS, ETC) • NO OPERATOR DECISION IS NECESSARY TO CREATE THE MEASUREMENT (IE: INSPECTOR, AUDITOR, ETC) • WHERE IS THE DATA COMING FROM? • DEFECT DATA: COMPARISON OF ACTUAL TO BILL OF LADING • SALES DATA: QUANTIFICATION OF BILL OF LADING • WHERE DOES THE BILL OF LADING COME FROM? • GENERATED ELECTRONICALLY FROM SCAN STATION DATA • CLOSED ELECTRONICALLY VIA RF
MEASUREMENT SYSTEM ANALYSIS HOW GOOD IS THE DATA? DO I TRUST MY DATA? • YES WHY? • DATA IS ELECTRONICALLY UPDATED AT TIME OF SCAN • ACCURACY OF ELECTRONIC DATA TRANSFER IS CHECKED ON A DAILY BASIS THROUGH QUALITY AUDIT PROCESS ANY DATA CONCERNS? • DEFECTS ARE REPORTED FROM CUSTOMER • MAY NOT BE ENTIRELY HONEST • MAY NOT UNDERSTAND THAT PRODUCT IS SALVAGE
MEASUREMENT SYSTEM ANALYSIS Further Review • NUMBERS ARE A PERCENTAGE OF SALES AND SOME NUMBERS ARE > 100% • SOME CUSTOMER ISSUES
MEASUREMENT SYSTEM ANALYSIS WHAT IS REALLY GOING ON? • BRUSHED DATA POINTS OVER 50% TO GET A CLEARER PICTURE OF THE BOTTOM • REALIZATION THAT MAY NEED TO LOOK AT INDIVIDUAL CUSTOMER REQUIREMENTS AND CLAIMS HISTORY • THE DATA IS THE BEST AVAILABLE THEREFORE, USE WHAT WE HAVE AND THEN ADDRESS CUSTOMER ISSUES
Y-METRIC GRAPH The Big Y - Baseline
FISHBONE DIAGRAM ORIGINAL
FISHBONE DIAGRAM REVISED WITH TEAM INPUT
FMEA Pareto Chart
Potential Failure Potential Failure Activity Name Key Process Input Mode Effects SEV Potential Causes OCC DET RPN # COULD BE PROCESSED PILFERAGE AT THE PRODUCT IS NOT INCORRECTLY AND STORE LEVEL, PROPERLY BE SCANNED AS AN OPERATOR ERROR, RETURNS RECEIVED C: LABEL AFFIXED DESIGNATED AS ALTERNATE LACK OF 11 @ CENTER AT STORE SALVAGE PRODUCT 9 INSPECTION 9 8 648 PRODUCT WILL SHIP WITH WRONG BOL SORT VIA SLP ZONE PRODUCT IS SORTED AND MAY NOT BE SLOT NAME INTO ZONE C: CONTAINER TO INCORRECT REPORTED AS AN CONFIGURATION, 5 CONTAINER CONFIG CONTAINER OVERAGE 7 OPERATOR ERROR 6 10 420 PALLETS LEFT SHORTAGE TO THE BEHIND, INCORRECT CUSTOMER CAUGHT INACCURATE PAPER 3 SCAN ERRORS C: SCANNER PRODUCT SHIP UPON RECEIPT 9 WORK 7 6 378 BOL POSTED FOR WILL CAUSE A SYSTEMS ISSUES, AUCTION OR FOR INCORRECT BOL DEFECT UPON INCORRECT 2 BIDDING C: BOL LISTING POSTED FOR SALE RECEIPT OF BOL 8 SCANNING 5 6 240 PRODUCT WILL BE LOCATION IS NOT UNACCOUNTED FOR OPERATOR ERROR, EMPTY AND AND WILL BECOME LACK OF PRODUCT IS PUT AN INVENTORY AUTOMATED AWAY TO DISCREPANCY AS PUTAWAY, REWAREHOUSE TO C: EMPTY INAPPROPRIATE WELL AS A DEFECT INACCURATE 9 RANDOM LOCATION LOCATIONS LIST LOCATION WHEN BOL IS SOLD 9 DOCUMENTATION 5 5 225 PRINTER NOT FAST INCORRECT LABLE ATTACH SLP TO ENOUGH TO KEEP UP CAN BE PLACED ON 18 PRODUCT C: PRINT SPEED WITH SCANNING PRODUCT 9 BACKLOG 3 5 135 PRODUCT IS MISSORTED DUE TO ISSUES WITH PRINTER UPKEEP, READING THE PRINTER QUALITY, ATTACH SLP TO LABEL IS NOT CORRECT SORT FACILITY 19 PRODUCT C: PRINTER QUALITY LEGIBLE ZONE ON LABEL 9 CLEANLINESS 3 5 135 CAN CAUSE LOSS OF SORTED PRODUCTS AND OVERFILLING OF DECREASE SORT VIA SLP ZONE CONTAINERS OR AVAILABLE INADEQUATE INTO ZONE C: REWAREHOUSE UNDERFILLING OF STORAGE SPACE IN SUPPLIES, VOLUME, 21 CONTAINER STAFFING CONTAINERS THE WAREHOUSE. 4 STAFFING 5 4 80 INCORRECT SCAN, WILL PICK WRONG INCORRECT INCORRECT PICK OR INCOMPLETE PUTAWAY 17 PICK BILL OF LADING C: PICK TICKET TICKET ORDER 9 LOCATION 3 1 27 PALLETS STAGED IN REWAREHOUSE TO UNMARKED 4 RANDOM LOCATION C: CAPACITY % DATA COLLECTION Focusing on Potential Critical X’s? • CUSTOMER ISSUE – OUTSIDE SCOPE • ADDRESSED – 2003 BB PROJECT • SCANNING ERRORS UNABLE TO LOCATE PALLETS FOR VOLUME, SLOT LOCATIONS SHIPMENT 6 LOCK 3 1 18 BOL POSTED FOR AUCTION OR FOR WILL DELAY 1 BIDDING C: AUCTION SYSTEM SYSTEM DOWN PROCESS 5 SYSTEM FAILURE 1 1 5
ANALYZE • Select Output Characteristics • Define Performance Standards • Validate the Measurement System • Establish Process Capability • Define Performance Objectives • Identify Variation Sources • Screen Potential Causes • Discover Variable Relationships • Establish Operating Tolerances • Validate Measurement System • Determine Process Capability • Implement Process Controls
HYPOTHESIS TESTING Monthly Trending?
HYPOTHESIS TESTING Monthly Trending? H0: mMONTH1= mMONTH2= ETC. Ha:AT LEAST ONE IS DIFFERENT • THE RESULTS OF THIS TEST SHOW A P-VALUE OF > .05 THEREFORE, THE NULL HYPOTHESIS IS ACCEPTED CONCLUDING THAT THERE IS NO STATISTICAL DIFFERENCE BASED ON MONTH One-way ANOVA: TOTAL DEFECTS versus DATE Source DF SS MS F P DATE 15 16468980 1097932 1.48 0.118 Error 167 124036131 742731 Total 182 140505111 S = 861.8 R-Sq = 11.72% R-Sq(adj) = 3.79% Individual 95% CIs For Mean Based on Pooled StDev Level N Mean StDev -+---------+---------+---------+-------- 200301 3 95.3 163.4 (-----------*-----------) 200302 7 617.1 831.3 (-------*-------) 200303 3 1209.7 1785.5 (-----------*-----------) 200304 9 1080.8 1836.9 (-------*------) 200305 14 70.1 152.3 (-----*-----) 200306 8 263.0 716.4 (------*-------) 200307 13 207.0 428.4 (-----*----) 200308 13 439.8 811.5 (----*-----) 200309 19 99.9 199.2 (----*----) 200310 21 13.2 23.1 (---*----) 200311 6 263.7 459.1 (-------*--------) 200312 9 223.4 416.2 (------*------) 200401 17 129.0 482.3 (-----*----) 200402 4 63.5 104.1 (----------*---------) 200403 20 676.1 1795.6 (---*----) 200404 17 134.5 517.9 (----*----) -+---------+---------+---------+-------- -800 0 800 1600
HYPOTHESIS TESTING Facility? • THE SEARS OH FACILITY APPEARS TO HAVE A SIGNFICANTLY HIGHER AMOUNT OF DEFECTS THAN THE OTHER FACILITIES • THIS ANALYSIS DOES NOT TAKE INTO ACCOUNT DIFFERING VOLUMES OF THE FACILITIES. • ADDITIONAL ANALYSIS NEEDS TO BE DONE TO DETERMINE IF THIS VARIABLE WILL BE USED FOR FURTHER ANALYSIS
HYPOTHESIS TESTING Facility? • THE RESULTS OF THIS GRAPHICAL ANALYSIS ARE THAT EVEN AFTER THE DATA WAS NORMALIZED TO ACCOUNT FOR VOLUME DIFFERENCES THE SEARS OH FACILITY STILL SHOWS A SIGNIFICANTLY HIGHER RATE OF DEFECTS. • THIS VARIABLE WILL BE USED FOR FURTHER ANALYSIS
HYPOTHESIS TESTING Scanning Issues? ANOVA OF SCANNING ISSUES H0: mMISSING PIECES = mALT PROD IN BOX= mMULTI SCANNED/MIS SCANNED Ha:AT LEAST ONE MEAN IS DIFFERENT One-way ANOVA: MISSING PIECES, ALT PROD IN BOX, MULTI SCANNED/ MIS SCANNED SING Source DF SS MS F P Factor 2 28029330 14014665 24.70 0.000 Error 174 98713494 567319 Total 176 126742824 S = 753.2 R-Sq = 22.12% R-Sq(adj) = 21.22% Individual 95% CIs For Mean Based on Pooled StDev Level N Mean StDev -----+---------+---------+---------+---- MISSING PIECES 133 147.8 379.0 (-*--) ALT PROD IN BOX 27 11.1 42.1 (----*----) MULTI SCANNED/ M 17 1465.1 2232.0 (-----*-----) -----+---------+---------+---------+---- 0 600 1200 1800 • MULTI SCANNED PRODUCT IS A KEY INPUT VARIABLE TO THE ACCURACY OF A BILL OF LADING
HYPOTHESIS TESTING BOL Size? H0: mSALES 2003 (YTD) = mSALES 2004(YTD) Ha: mSALES 2004(YTD) < mSALES 2003(YTD) • SHIFT IN TRADITIONAL BUSINESS SPECS TO SMALLER PURCHASES • IN LINE WITH SHIFT TO B2B LTL AUCTIONS • IMPLICATIONS IN TERMS OF OVERHEAD COSTS AND CLAIMS One-Sample T: A Test of mu = 5722.93 vs < 5722.93 95% Upper Variable N Mean StDev SE Mean Bound T P ACTUAL 1336 5199.09 5769.25 157.84 5458.89 -3.32 0.00
IMPROVE • Select Output Characteristics • Define Performance Standards • Validate the Measurement System • Establish Process Capability • Define Performance Objectives • Identify Variation Sources • Screen Potential Causes • Discover Variable Relationships • Establish Operating Tolerances • Validate Measurement System • Determine Process Capability • Implement Process Controls
IMPROVE OPPORTUNITES Identified • SYSTEMIC • NO STANDARDIZED THRESHOLDS SET UP FOR QUANTITY OVERRIDE FUNCTIONS • NO USER CONTROLS IN PLACE TO LIMIT QUANTITY OVERRIDE USAGE • ABILITY TO NEGATIVE SCAN PRODUCT AFTER BILLS HAVE BEEN POSTED FOR BID • DISCREPANCIES BETWEEN THE SOP’S AND ACTUAL PRACTICE • FAILING TO INSPECT PRODUCT THAT IS MEDIUM SIZE OR GREATER • OVERQUANTIFYING PRODUCT AND QUANTITY SCANNING IT • SCANNING AS THE WRONG EACH UNIT (IE: ONE BAG VS. ONE PIECE, 12 PCS VS. ONE CARTON) • PILFERAGE • LACK OF EFFECTIVE TRACKING METHOD • MEASUREMENT STANDARD DOES NOT REPRESENT SALVAGE OPERATION – OVERSTATING ACCURACY • CUSTOMERS ABUSING THE CLAIMS PROCESS • TEAMMATES ABUSING THE CLAIMS PROCESS
IMPROVE OPPORTUNITES Implemented • SYSTEMIC • MANUALLY SHUT DOWN ABILITY TO QUANTITY OVERRIDE IN ALL SEARS FACILITIES • CONFIGURED THE SYSTEM WITH CERTAIN LEVEL TEAMMATE PERMISSIONS FOR QUANTITY OVERRIDE CAPABILITIES • FIXED NOTIFICATION PROCESS TO ELIMINATE LARGE AMOUNT OF EMAILS RECEIVED REGARDING CHANGES TO BOL’S. RECEIVING NOTIFICATION ONLY REGARDING THOSE BILLS WHICH MEET A CERTAIN CRITERIA • DISCREPANCIES BETWEEN THE SOP’S AND ACTUAL PRACTICE • DETERMINED MANUAL MEANS OF IDENTIFYING PRODUCT THAT IS INCORRECTLY SCANNED • ELIMINATED 6 KEY POSITIONS WITHIN THE WAREHOUSE MANAGEMENT TEAM WHERE PERSONS WERE BELIEVED TO HAVE BEEN FALSIFYING DATA BY INTENTIONALLY INSTRUCTING TEAMMATES TO “SHIRK” THE SYSTEM TO INCREASE FACILITY’S PRODUCTIVITY NUMBERS • IN THE PROCESS OF ELIMINATING MORE FRONT LINE POSTIONS THAT HAVE BEEN ASSOCIATED WITH PILFERAGE. • LACK OF EFFECTIVE TRACKING METHOD • IMPLEMENTED A TRACKING SYSTEM SPECIFIC TO THE AR TEAM AND THEIR KEY INDICATORS • IDENTIFIED A LIST OF SUSPECTIVE CUSTOMER ABUSERS. ADDRESSING THESE ISSUES AS A SEPARATE SIX SIGMA PROJECT • PURSUING APPROPRIATE RECOVERY PROCESS (IE: FINANCIAL, LEGAL, CRIMINAL) FROM THOSE BUYERS WHO ARE IN VIOLATION OF THEIR PURCHASING AGREEMENTS • DETERMINING WAYS OF PROACTIVELY IDENTIFYING SALVAGE CLAIMS BEFORE THEY REACH THE CUSTOMER. DETERMINING WAYS OF AUTOMATING THIS IDENTIFICATION PROCESS.
Y-METRIC GRAPH The Big Y AUGUST, 2003 AUGUST, 2004
THE IMPROVEMENT A Graphical Representation
THE IMPROVEMENT A Statistical Representation H0:p(1) - p(2) <= D(0) Ha:p(1) - p(2) > D(0) Test and CI for Two Proportions Sample X N Sample p 1 51212 22343070 0.002292 2 8700 7614820 0.001143 Difference = p (1) - p (2) Estimate for difference: 0.00114957 95% lower bound for difference: 0.00112344 Test for difference = 0 (vs > 0): Z = 72.38 P-Value = 0.000 • THE RESULTS OF THIS TEST SHOW A P-VALUE OF < .05 THEREFORE, THE NULL HYPOTHESIS IS REJECTED CONCLUDING THAT THERE IS IN FACT A STATISTICAL DIFFERENCE TO THE ERROR RATES SINCE THE IMPLEMENTATION OF THE IMPROVEMENTS.
THE IMPROVEMENT A Financial Representation
CONTROL • Select Output Characteristics • Define Performance Standards • Validate the Measurement System • Establish Process Capability • Define Performance Objectives • Identify Variation Sources • Screen Potential Causes • Discover Variable Relationships • Establish Operating Tolerances • Validate Measurement System • Determine Process Capability • Implement Process Controls
ONGOING CONTROL Tracing and Preventing • “RED FLAGS REPORT” IN DEVELOPMENT • WILL PROACTIVELY IDENTIFY SALVAGE BL’S IN THE WAREHOUSE THAT MAY HAVE A CLAIMS ISSUE BASED ON THE CRITERIA DEVELOPED IN THE AFOREMENTIONED MANUAL PROCESS • WILL PREVENT THAT BILL FROM BEING SOLD UNTIL SOMEONE WITH APPROVAL PERMISSIONS HAS AUTHORIZED ITS DISPOSITION • AR WILL CONTROL TEAMMATES WITH PERMISSIONS TO THE SYSTEM • IMPLEMENTED COLORED TAPE IN THE INSPECTION PROCESS • TECHNICALLY NOT THE TEAM’S IDEA • RECEIVED AN IMPROVEMENT IDEA FROM THE FRONT LINE • WILL EASILY POKE A YOKE THE ISSUES WITH BOXES NOT BEING OPENED AT THE INSPECTION STATION • ADDITION OF ASSET RECOVERY TRACKING MECHANISM • DEVELOPING AN ADJUSTMENTS REPORT THAT WILL CONTINUE TO TRACK BILL OF LADINGS WHICH HAVE CLAIMS PAID ON THEM EACH MONTH • CONTINUE TO REPORT MONTHLY METRICS SPECIFIC TO THE AR GROUP EACH MONTH INCLUDING INFORMATION REGARDING TOTAL SALES AND ACCURACY
CONTROL Reaction Plan • “RED FLAGS” REPORT HAS A SYSTEMIC REACTION PLAN THAT NOTIFIES THE PROCESS OWNER OF A POTENTIAL PROBLEM AND THEN WILL NOT ALLOW THE BOL TO GO ANY FURTHER UNTIL RESOLUTION HAS BEEN ACHIEVED • IF PRODUCT IS SEEN ON THE LINE WITHOUT COLORED TAPE TO SEAL THE INSPECTED BOX BACK TOGETHER, THAT BOX IS REMOVED FROM THE LINE AND REPROCESSED TO PREVENT A DEFECT FROM REACHING THE END CUSTOMER • WHEN THE MONTHLY METRICS SHOW A FACILITY THAT IS EXCEEDING THE EXPECTED DEFECT VALUE FOR THE MONTH, AR TEAMMATES ARE SIGNALED TO PULL THE CLAIMS INFORMATION FOR THAT FACILITY FOR THE MONTH AND DETERMINE THE CAUSE OF THE INCREASE. THE AR TEAMMATE SHOULD SOLICIT ASSISTANCE FROM THE APPROPRIATE DEPARTMENT (IE: LOSS PREVENTION, SIX SIGMA, ENGINEERING, ETC)
NEXT STEPS • CONTINUE TO DRIVE DEVELOPMENT AND IMPLEMENTATION OF RED FLAGS REPORT • DOCUMENT SOP FOR TRACKING AR MEASURES AND PASS OFF TO THE PROCESS OWNER • TRACK FINANCIAL SAVINGS • DOCUMENT “SPIN OFF PROJECTS” IDENTIFIED DURING DMAIC PROCESS AND DEVELOP A PLAN FOR ADDRESSING • RESOURCE MANAGEMENT • STORE LEVEL LABELING • SHRINK WRAP AUTOMATION • GENERIC SALVAGE SKU VALUATION • CUSTOMER SERVICE ISSUES • SALVAGER CLAIM ABUSE • ROLL OUT RED FLAG REPORT TO OTHER RETURN CENTERS AND ASSET RECOVERY CUSTOMERS
QUESTIONS? THANK YOU