Harvesting Pearls with Supply Chain Analytics Track 3 Session 2
Therese Costich - Managing Partner - The TCM Group - tcostich@t-c-m-group.com - 585 314 1579 - www.t-c-m-group.com
Agenda Abstract Definition What is Supply Chain Analytics? Uses Why Use Supply Chain Analytics (Case Studies)? Common Tools Used in Supply Chain Analytics Recommendations Key Takeaways Additional Resources 3
Abstract There's no shortage of big data in our world. What's missing is a manageable way to boil that ocean down to meaningful, relevant observations that help us make great decisions. This hour is about Supply Chain Analytics; a dimension of Business Intelligence. It involves the tools and processes that extract transactional, behavioral and operational data and report concise trends and forecasts. Modern tools allow the owner to draw correlations, plan and test scenarios before they happen. Wonderfully unfair! 4
Definition Wouldn t it be nice to: - Know Precisely How Much Raw Material to Purchase over the next 5 Months - Know Exactly the Cost of a Given Commodity 5 Months in Advance - Know Exactly When the Next Snowstorm Will Hit in Cleveland so You Can Make a Trucking Decision 5
Definition Wouldn t it be great if we could: - Know Exactly When Our Supplier Will Have a Defect in a Lot of Raw Material? - Know Exactly How Much Product our Customers Will buy in the Next Quarter? - Know The Exact Cost of Shipping Over the Next Year? - Never Have a Rush Order, or Overnight Delivery? 6
Definition We Can t 7
Definition Not Without a Crystal Ball. And There is No Such Thing 8
Definition Supply Chain Analytics The use of Skills, Technologies, Tools, and Data to gain insight into key future variables to make better business decisions. 9
Definition Supply Chain Analytics Historical Data Guess Work I Think We Should Our Best Customer Says Excessive Buffers Predictive Statistics Modeling Sensitivity Analysis Science Monitoring 10
Uses Demand Forecasting Raw Material Purchasing Manufacturing and Operations Planning Shipping Options Shipping Costs Currency Fluctuation Outsource Manufacturing 11
Uses Case Studies An industrial manufacturing company used Supply Chain Analytics to determine not only the correct carrier, but also to smooth out the fluctuation in routes and costs A global electronics company was able to use Supply Chain Analytics to better understand demand for a board manufactured in China, thereby reducing airfreight significantly 12
Uses Case Studies A manufacturing company used Supply Chain Analytics to identify supply risk factors in its far eastern operations including (floods, typhoons, lower-tier supply disruptions, and so forth). A small assembly operation used Supply Chain Analytics to smooth out sourcing disruptions thereby decreasing backorders and overnight shipments by 40%. 13
Tools-Identifying and Modeling Identifying areas in your Supply Chain Model which are susceptible to future risks - Value Stream Maps - FMEA - Supply Chain Models - Modeling Software - Risk Analysis - Cause and Effect Matrix WT = 0 Note: 29.7% of orders are manual (Jan- May 2007) Customers can submit orders in any/all of the order methods available Note: Excludes manual orders rec d in customer specific email DB Cycle times based on travelers from 5/21 to 6/21 CSR performs daily account credit check CT = 21.1 min Fax admin assigns Order to customer name WT = 0 Assigned CSR Receives order Sum of received outside ACME Order Ops to Assigned by CSR in DB CT = 924 min (15.3 hrs) WT = 212 min (3.5 hrs) Manufacturing Est. CT = 3 min Processes included in order receipt: Order change/cancel Lotus Notes DB: ACME Order Operations CT = 5-10 min est. CT = 60 min R = 60 minutes to 1.5 days (worst case) Invoice creation in CP6 & FP9 following shipment, removal of pertinent blocks Freq every 15 min I CSR validates Credit check ISSR or Secretary receives & faxes to Lotus Notes DB Processes included in FOP: CT = 21.1 min Ship to Override Price Override 52 min. WT = 6.5 min ISSR/customer communication CMR request 10 min CT = 3 min CT = 3 min Email ISSR or CSR receives & emails to Lotus Notes DB Note: After invoice is created in FP9, invoice is delivered to customer (EDI, paper/mail). Method is dependent on customer set up. CT = 1 /day (night) CT = 11 days Note: Date of USPS stamp to date received by admin. CT = 3 min CSR checks for Firm Order Policy (FOP) Real Time Customer SAPP Freq = 15-30 min Freq = 15-30 min CSR places order in (CP6) SAPF(FP9) plus add l OE steps Scheduling team actions CT = 31.4 min WT = 0.4 min CSR confirms order CSR updates ACME CSR checks SAPF Delivery, Value added (VA) To customer, Order Ops DB Billing Block, Incomplete order time If applicable to close PO report, provides comments Weekly BA Defect = 0.3% Included in prior Included in prior F = Daily, multiple times; step step Weekly Block D = 1% (LED); 5-10% units (AG) CT = 251.9 min (4.2 hrs) WT = 158.8 min (2.6 hrs) Lotus Notes DB: ACME Order Operations Processes included post OE: Billing adjustments Tax adjustment Returns Invoice disputes There Is no One Way To Model Your Supply Chain 14
Tools-Sensitivity Analysis Sensitivity Analysis Using varying input data to understand the effects on the output - Supports Better Decision Making - Supports Models Through Varying Scenarios - Tools Excel Spreadsheets XY Scatter Plots Correlation Regression Main Effects Plots 15
Tools-Response Surface Optimization A predictive design of experiment which allows the user to predict the optimal output given a set of input data - Provides a Predictive Model - Uses Statistical Methodology - Provides Optimum Setting - Determines Optimum Operating Conditions 16
Tools-Predictive Statistics Using historical and real time data to give statistical insight to future conditions - Hypothesis Testing - Control Charting - Scatter Plots - Box Plots - Analysis of Variance Use the Tool Based on the Predictive Outcome Desired 17
Tools Summary One Tool Does Not Satisfy Every Scenario Tools are Not Mutually Exclusive - Used together provides Better Results Not Every Tool Has To Be Statistically Challenging - Brainstorming Works Too!! Understand Your Needs Before Applying Tools - Garbage In/ Garbage Out Back Order 18
M e a n Case Example Objective: Optimize freight transportation revenue given various variables and factors Step 1 Main Effect Plots to understand Key Input Variables (KPI) 50. 0 47. 5 45. 0 42. 5 Main Effects Plot for Freight Revenue M a i n E f f e c t s P l o t f o r R e s i d u e D a t a M e a n s A Weight T e m p e r a t u r e Fuel Efficiency B T i m e KPIs for Revenue Optimization: Weight Fuel Efficiency Mileage 40. 0 50. 0 200 300 C C Mileage o n c e n t r a t i o n 1 5 4 5 4 7. 5 4 5. 0 4 2. 5 4 0. 0 0. 5 0 0. 7 5 19
Case Example 2) Optimize 3 variables using a 3 factor DOE 3) Determine the Optimum Spec Limit for Weight Factorial Fit, Weight, Efficiency, Mileage Weight Fuel Milage Concentration Weight*Fuel Weight Concentration Fuel Concentration 4) Collect data on new optimized process 5) Monitor and Control Process Using SPC 20
Recommendations Recommendation 1: Decide on a Goal - Determine what you want to accomplish Smooth out Purchasing Spikes Reduce Raw Material Inventory More Accurate Forecasting More Level Production - Understand Gaps - Where is the Variation - Make your Goal Meaningful - Understand When you Have Accomplished Your Objective 21
Recommendations Recommendation 2: Use Data - Most Organizations Have Files Upon Files Of Data - Make the Data Meaningful - Don t Always Look in the Rear View Mirror Not Just Historical Data - Benchmark Other Organizations - Brainstorming Will Not Get You There - Don t Rely on I Think In God We Trust Everyone Else Bring Data 22
Recommendations Recommendation 3: Make It Project Based 1. Gather a Team 2. Decide on an Objective 3. Determine What Data You Need 4. Gather the Data 5. Analyze the Data 6. Make Recommendations Based on the Data 7. Decide How to Modify the Process to Accomplish the Objective 8. Gather Data to Monitor Results 23
Recommendations Recommendation 4: Get Help - Consultants - Experts within Your Organization - MBA Students - Graduate Students in Statistics - Local Colleges and Universities - Co-Ops 24
Recommendations Recommendation 5: Start Small - Don t Boil the Ocean - Look at Areas of Fluctuation or Variation - Scope Projects to Manageable Levels - Build Upon Successes - Start with Small Problems and Work Up Look for Low Hanging Fruit 25
Key Takeaways Supply Chain Analytics Help Make Better Business Decisions Supply Chain Analytics and Predictive Statistics Can Save Organizations Thousands of Dollars There are a Variety of Tools that Can be Used - Tools are Not Mutually Exclusive Recommendations: * Use Data * Project Based * Get Help * Start Small * Decide on a Goal 26
Additional Resources The Promise of Advanced Supply Chain Analytics By Jerry O Dwyer and Ryan Renner http://www.scmr.com/article/th e_promise_of_advanced_supp ly_chain_analytics/ Driving Six Sigma By Ron Sicker TCM Publishing 2002 Supply Chain Analytics: How Hard Should You Squeeze?- Deloitte Debates http://www.deloitte.com/view/e n_us/us/insights/browse-by- Content-Type/deloittedebates/cefe46d054aaa210Vgn VCM3000001c56f00aRCRD.htm The Black Belt Memory Jogger Six Sigma Academy Goal/QPC Publishing 2002 27
Questions? 28