caseijrsdjprsvjp[m'viorjvspodmjpoane.pptx

good

1. Team Diwap IIM Mumbai(NITIE) Vaibhav Awasthi Vivek Sharma Vinod Kumar

2. Problem Statement To design network for clustered Segments/SKUs basis logistical synergy and provide replenishment strategy while optimizing service level and logistics cost. Secondary Research Proposed Clusters Primary Research SD-OFR: 70-80 percent in Mumbai Reason of Delay: Traffic, congestion Order placed by dealer via Myawaz App WhatsApp and by call No visibility of POS data Issues identified from Mumbai dealers Demand Forecasting: AI/ML-based forecasting (Regression, Decision Tree, SVR, ANN or Demand Sensing) Inventory Positioning: J&J decide inventory positioning based on average monthly demand and Monthly COV (variability) POS data Visibility: Lenskart Pos for seamless inventory and payment Management D: Wood Finishes E: Water-Proofing & Construction Chem. G: New Products Volume Variability Medium & High Low Low Medium High IV B C I J V VI I II III G A H F D E G B : Emulsion Paint – Economy  C: Primers  I: Bulk Volume Products – Putty J: Seasonal Products      (Distemper &     Enamels) A: Emulsion Paint – Premium & Luxury F: Colorants H: Project Sales Assumptions G: New Products Volume ∝ 1/(Relative WASP) COV (Variability) ∝ Forecast Error

3. Chapter 1 Proposed Network for Clusters & Replenishment Strategy Multi-Echelon Inventory Optimization Chapter 2 Cluster Network Replenishment Method Initial Positioning for Optimization I DC → TL → Sales Unit (R, s, S) Policy 55% Upstream + 45% Downstream II DC →TL → Sales Unit (R, s, S) Policy with less R 80% Upstream + 20% Downstream III DC →TL → Sales Unit (R, s, S) Policy with lesser R 80% Upstream + 20% Downstream IV Combination of DC → Dealer & DC → Sales Unit Inventory Pushed Based on forecast replenishment after fixed period V Combination of Lead Time Pooling + DC → Sales Unit (S, s) policy 70% Upstream + 30% Downstream VI Lead Time Pooling (S, s) policy 80% Upstream + 20% Downstream There are multiple Plants, TLs and Sales Unit, so we will perform Multi -Echelon Inventory Optimization to predict the amount of inventory at each node. KPI : SD OFR & Cost to serve Chapter 3 Advantage : Optimum ROP, base stock of inventory @ each echelon for each SKU Multi Echelon Optimization Initial Guess of Inventory level & ROP at each node using scattered plot Global Optimization to minimize Inventory level & meet SL of 95% Average demand, its variability, Lead time Variability Dealer Supplier Relationship: VMI Fill-rates increased, better control on shipment and inventory & On-shelf availability Dealer level forecasting with integration of promotions & improved Vehicle fill rate Efficient & Reliable Order Fulfilment Chapter 4 Suggested Method becomes more feasible as Variability and Lead time from DC increases DC Depots Lead Time Pooling Depots TL DC Implementation SKUs Clustering Network Design Inv. Positioning - Chapter 2 POS Int. (1-1.5 Years) Region-wise then Pan India

4. Orders will be initiated based on a review period (R) and inventory levels (s and S), ensuring replenishment occurs when the inventory drops below the reorder point (s). Here, S = order-up-to level s = reorder point Orders will be initiated as inventory depletes to or falls below the reorder point (s) to reach the order-up-to level (S), maintaining inventory levels as needed. Appendix Integrating POS data will reduce Volatility Ratio, improve fill rate, on-shelf availability, inventory optimization. 1. (R,s,S) Policy 2 . (S,s) Policy Here, S = order up to level s = reorder point R = review period LT = lead time from TL(Transhipment Location) to Depot(Sales Units) 3 . POS Integration 4. Benchmarking with J&J Average Volume per month per SKU Monthly Demand Variability (CoV) For high volume and low CoV keep 80% at Sales Unit For high volatility put 20% at Sales Unit Note: SKU falling under these quadrants may change from Region to region References Simchi-Levi. (n.d.). Designing & Managing The Supply Chain (4th ed.). Rofin TM, Asst. Professor, IIM Mumbai 3.    Massachusetts Institute of Technology. (n.d.). Improving Forecast Accuracy through Demand Sensing, June, 2018 Why we Need DSP ? Demand based on sales out (POS) Dealer Sales Unit Transhipment Locations Distribution Centre Order Placed Order Demand Dealer Var(Sells in) Var(Sells Out) Inventory Ownership Volatility Ratio Reduce