GlobalSourcingandVendorRiskManagementinSupplyChains.pptVIP

Nov. 23, 2010 A. Ravi Ravindran (Penn State) * Objective Functions Minimize the purchasing and fixed cost. Minimize the average lead-time. Minimize loss due to rejects (modeled as MtT risk) Minimize loss due to hurricanes (modeled as VaR risk) Nov. 23, 2010 A. Ravi Ravindran (Penn State) * Model Constraints Capacity Constraints Demand Constraints Maximum number of suppliers Linearizing Constraints for quantity discounts Non-negativity constraints. Binary Constraints. Nov. 23, 2010 A. Ravi Ravindran (Penn State) * Solution Method Goal Programming (GP) is used to solve the multi-objective supplier selection problem. Get from the decision maker goals/target. All the goals may not be achievable. Get decision maker’s preference on achieving the goals. Find a solution that will come as close as possible to the stated goals in the specified order. Nov. 23, 2010 A. Ravi Ravindran (Penn State) * GP Methods Preemptive GP Target set at 110% of the Ideal values. Preemptive priorities, Price, MtT risk of quality, lead-time, VaR risk (from Phase 1 results) Non-preemptive GP Weights obtained from Phase 1. Tchebycheff (Min-Max) GP Minimize the maximum weighted deviation from the targets. Weights obtained from Phase 1. Fuzzy GP Minimize the maximum weighted deviation from the ideal values. Weights obtained from Phase 1. Nov. 23, 2010 A. Ravi Ravindran (Penn State) * Problem Size For a problem with 2 products, 2 buyers, 5 suppliers with each supplier having 2 price breaks, the problem size is as follows: Total number of continuous variables: 40. Total number of binary variables: 45. Total number of constraints: Capacity constraints: 10. Demand constraints: 4. Number of supplier constraint: 1. Linearizing constraints: 60. Nov. 23, 2010 A. Ravi Ravindran (Penn State) * Value Path Approach Efficient way to visualize different solutions and their trade-offs Display contains set of parallel lines; one for each objective. Value of each solution on the axis is that sol