This preview shows page 1. Sign up to view the full content.
Unformatted text preview: Case Study Professors given tour of large automotive g g parts processing plant (PP) Integrating the Threads
Peter L. Jackson Professor School of O.R. and I.E.
2/17/2011 Supplier Parts Processing Plant (PP) Distribution Center (DC) Sh Shown charts of production history h f d i hi
weeks of relative inactivity Dealer Repair Shop High variability: weeks of overtime follow
1 2/17/2011 Industrial Data and Systems Analysis Industrial Data and Systems Analysis 2 Data Analysis We suspect the problem is lot sizes p p The so-called "bullwhip effect" Lot sizes increase as you go up the supply chain and Data Analysis (cont'd) (cont d) Using SQL (wish we had a student to do all this): amplify the variability of demand Collect data ('000s of part numbers) By part number, sales rate from DC to dealers By part number, DC standard order quantity By part number, rack size (minimum supplier ship Simulate randomized DC sales Derive DC order stream on PP Round DC order stream to next largest rack quantity) piece By p y part number, standard production order q , p quantity y By part number, by workcenter, production hours per quantity Derive PP production orders A Aggregate b workcenter ( t by k t (convert t t to production hours) Import to MS Access
2/17/2011 Display results as graph
3 2/17/2011 Industrial Data and Systems Analysis Industrial Data and Systems Analysis 4 Source of Production Variability V i bilit
Simulated Pipeline Ordering Activity
120 Production Hours Orde H ered 100 80 60 40 20 0 0 10 20 30 40 50 60 70 Simulated D Si l t d Day
DC Sales DC Orders DC Orders by Rack Production Orders Obvious Conclusion Conclusion: "Oh, we do it to ourselves!" Production order quantities are a major source of variability Parts plant (PP) has control of production order quantities "We have seen the enemy and he is us." Pogo 2/17/2011 Industrial Data and Systems Analysis 5 2/17/2011 Industrial Data and Systems Analysis 6 Consequences of Large Lot Sizes Si Cost reduction due to batching g Avoidance of setup time and cost (This is a first order effect) Strategy to Reduce Production V i bilit P d ti Variability Cost increase due to batching Hidden costs Ex. Overtime to cope with unpredictable production loads (This is a second order effect) "Economics is the science of studying secondary and tertiary effects" Ludwig von effects Mises First order effects are usually obvious to everyone Economist should look beyond the obvious 2/17/2011 Industrial Data and Systems Analysis 7 2/17/2011 Industrial Data and Systems Analysis 8 The Strategy in Words Need to reduce production order quantities in The Value of Inventory Models Observe the role of inventory models in our y order t reduce th variability that is causing d to d the i bilit th t i i overtime But, must attack root cause (setup costs) for this to t work k Also, must be prepared to increase safety stock to buffer against demand variability since cycle stock has acted as buffer in the past In the long run, work to reduce supplier lead times so that safety stock can also be reduced Goal: leaner, smoother system with G l al th t ith dramatically lower overtime costs and inventory requirements
Industrial Data and Systems Analysis
9 thinking Queueing theory ("decoupling stock") helps us to understand connection between lot sizes and variability Lot sizes exist for a reason; need to pursue root cause Large lots hide the need for safety stock; need to be prepared to i d t increase safety stock as l t sizes go d f t t k lot i down Safety stock exists for a reason: demand uncertainty and long lead times. Need to reduce lead times. It' not so much th t th f It's t h that the formulas will save us: l ill they simply guide us to be working on the right problems: setup costs and lead times.
Industrial Data and Systems Analysis
10 2/17/2011 2/17/2011 Integrating the Threads What tools were useful? MS Access, SQL (joins and aggregations) MS Access macros (for simulation, see later in ( course) MS Excel (for charts) I Inventory f t formulas (f i t iti ) l (for intuition) IDEF0 (for visualizing strategy) 2/17/2011 Industrial Data and Systems Analysis 11 ...
View Full Document
This note was uploaded on 03/18/2012 for the course ORIE 3120 taught by Professor Jackson during the Spring '09 term at Cornell University (Engineering School).
- Spring '09