MilkRun Simulation Members Advisor Daniel Dawson Jesus Jimenez

“Milk-Run” Simulation Members: Advisor: Daniel Dawson Jesus Jimenez Reid Pierson Richard Mc. Evoy-Kemp Industry Advisor: Eleazar Zavala Sarah Chowdhury

Background • Philips Lighting in San Marcos currently lacks a steady manufacturing line to assemble products § Largely due to custom orders • Philip’s hopes that implementing an optimized assembly line will allow their facility to be the top manufacturing facility worldwide • Implementation of a Kanban Supermarket will provide an increase in production and efficiency within their custom facility § § Reduced bottlenecks Increased throughput Increased inventory control Increased profits

“Milk Run” “The milkman would not only deliver fresh milk in the morning, but would collect empty bottles also to eliminate unnecessary trips” • A round trip delivery method which facilitates collection and distribution • In this project, we will determine the frequency at which components should be replenished and the “milk-runner” will fetch the components from the warehouse • This idea will have an effect on cycle time, throughput, and overall efficiency

2 -Bin System / Kanban System • • Will help reduce waste due to under and over production Use of Safety stock Card replacing Work inside out to control inventory

Deliverables/Goals • Increase efficiency and throughput in the Philips model line 741 • Configure and optimize Kanban Supermarket variables such as bin capacities, reorder points, and frequency of milk runs • Develop detailed model of the ordering, picking, supermarket and workstation assembly process given the yearly demand

Methodology • Microsoft Excel § Data collection and analysis • WITNESS Simulation § Modeling the assembly line and the throughput of the parts § 2 -bin Kanban Supermarket § “Milk-Run”

Data collected • • Parts to use within the supermarket Designated workstation for each of the parts Part usage rate Lead time Classification of parts based on lighting models Bill of materials for each product made within the assembly line Finished Goods Pure. Form

Demand forecast • Use Holt’s Method to produce forecast of Pure. Form product • The parameters alpha and beta were set to 0. 2

Model Methodology • Our model simulates the effectiveness and efficiency of a “Milk. Run” process to achieve the forecasted monthly demand • Simulated 6 workstations and 7 workstations within the modelline; each simulated for 140 hours (4 working weeks) • Inputs needed to construct the model • • • Milk Run Reorder points Bin sizes Component inter-arrival times Each scenario evaluated with “Experimenter” within WITNESS

Model Methodology • 6 workstations utilized a triangular distribution for the stations process times: • Min = 7 minutes • Mean = 10 minutes • Max = 13 minutes • 7 workstations utilized a triangular distribution for the stations process times • Min = 5 minutes • Mean = 8. 5 minutes • Max = 10 minutes

The Model The model will be running at our booth for those interested in seeing it work!

6 stations o Six Stations-Baseline a. Inter-arrival time: 0. 185 hours b. Max bin capacity: 50 c. Reorder point: 25 o Six Stations-Experimenter

7 Stations • Seven Stations a. Inter-arrival time: 0. 159 hours b. Max bin capacity: 50 c. Reorder point: 25 • Seven Stations-Experimenter

Outcomes/Conclusions • Preferred number of workstations • 7 stations produced the highest throughput, although the demand was still not able to be met • A total of 8 stations is ultimately needed to meet the forecasted demand without having backorders

Acknowledgments • Philips Lighting-San Marcos • Dr. Jesus Jimenez (Advisor) • Sarah Chowdhury (Industry Advisor) • Psyche Bryant (Industry Supervisor)