Simulation Model - Arena
Zuo Tao LIU Chih‐Yuan KANG Yang TANG
Agenda y Introduction & Objectives y Assumptions y Model Construction y Results of Simulation Models y Conclusion & Recommendation
Introduction y Background of JPR Corporation 1. Provide the shipping service to 3 western states: California, Nevada, and Arizona 2. Downsize its fleet size to 50 trucks during economic crisis
Problems y Uncertainty of Future Demand y Volatility of Diesel Price y The management noticed that overcapacity and stock‐out costs climbed
Objectives For the following month, y Slightly adjust its capacity to a appropriate utilization ‐ 41, 44, 47, 50, 53, 56, 59 y Maximize its profit
Business model y Business Days: 5 days y All collected orders are shipped at 05:30pm on each day, and no late order is accepted. y The leftover half truck‐load orders will be shipped on the following day. y JPR loses the order and incurs stock‐out cost if the capacity is full when an order arrives. y There is only one‐way shipping and trucks return with no loading.
Daily Demand y Average Daily order number:
y Price of Different Destinations
Based on our assumption
Based on our assumption
# of Pallets of Each Order y The Maximum Truck Loading: 20 Pallets y The truck load of each order has a Discrete Distribution.
Based on our assumption
Estimated Distance of Destinations y Travel distance is determined by the total distance of two destinations on Google Map.
Destination
CA
NV
AZ
Distance
554 Miles
580 Miles
650 Miles
Shipping Time Destination Time Per Round Trip (day)
CA Norm (1,0.05)
NV Norm (1,0.18)
AZ Norm (1,0.12) Based on our assumption
Priorities of Destinations y Policy: Highest Revenue, Highest Priority
Based on our assumption
y Two of a half truck‐load (6‐10 pallets)
will be batched to one shipment
Jarrett I’ve died because of
You!!!
Simulation Models 1. Diesel Price ‐ Excel 2. MPG ‐ Excel 3. Order accumulation and Shipment ‐ Arena
Diesel Price y The historical data of 2007‐2009 from U.S. Energy Information Administration y Sample size : 1,000
Result
MPG y The average mileage per gallon diesel of California Transportation Bureau for shipping trucks in California y Sample size: 1,000
3 key components Total : 9 Steps 1.
Assign Orders Truck Loading –Step 2
2. Separate Balking Orders –Step 5 3. Batch the Half Truck‐Loading Orders –Step 6
Step 1 y The demand of order is generated from the parameter we defined.
Schedule Module :
Step 2 y Assign number of pallets to each order based on our assumption of order loading
Coding by Discrete Distribution :
Code
1
2
3
Step 3 y Use 1st Decision module to separate each order by the destination
Step 4 y 2nd Decision module to determine whether the shipment achieve the minimum shipment requirement.
1‐5 pallets
Step 5 y 3rd Decision module to determine whether the shipment need to batch or not
11‐20 pallets
Step 6 y Batch the half truck‐loading orders Wait another half‐loading order
Directly process
Step 7 y Setup shipping time and priority as assumptions.
Step 8 y Record the balking rate for order to different destinations
Step 9 y Repeat Step 4 – 7 for the other two destinations
Run Setup y 5 Days a Week = 5 Days Replication Length y 4 Weeks a Month * 5 Observations = 20 Replications
Run Simulation
Richard
Cost Assumptions Destination Revenue Per Order Operation Cost Per Order
CA $750
NV $800
AZ $825
$200
$208
$217
Inventory Cost Per Truck
$225
Stock-out Cost Per Order
$150
$240
$412.5
Average Distance
554 Miles
580 Miles
650 Miles Based on our assumption
Monthly Profit Analysis Revenue Operation cost Inventory cost Gas cost Stock-out cost Total cost Estimated profit
Price * Shipping times Operation cost pre shipment * Shipping times Total truck fleet * $225 (Distance/MPG)* Gas Price Balking times * Stock‐Out Cost Operation + Inventory + Gas + Stock‐out cost Revenue – Total Cost
Sensitivity Analysis
Findings‐Utilization vs. Profit Chart of Utilization by Different Capacity
Chart of Estimated Profit by Different Capacity
Over‐ capacity
Highest Profit
Findings‐Revenue vs. Cost Chart of Total Cost by Different Capacity
Higher Inventory Cost
Chart of Revenue Cost by Different Capacity
More Revenue
Conclusion y It’s not JPR’s best interest to downsize capacity to 50.
Dad’s company is losing money!
Recommendation‐1 y Fleet size increase
50
53
Recommendation‐2 y Adding smaller‐size trucks Don’t even think I am small!
Possible Improvement‐1 y Increase sample size
Possible Improvement‐2 y Considering seasonality
It will rain or not tomorrow?
Possible Improvement‐3 y Use number of pallets instead of coding to reflect the real world
Let’s make some money! 赚钱啦～
Agenda y Introduction & Objectives y Assumptions y Model Construction y Results of Simulation Models y Conclusion & Recommendation
Introduction y Background of JPR Corporation 1. Provide the shipping service to 3 western states: California, Nevada, and Arizona 2. Downsize its fleet size to 50 trucks during economic crisis
Problems y Uncertainty of Future Demand y Volatility of Diesel Price y The management noticed that overcapacity and stock‐out costs climbed
Objectives For the following month, y Slightly adjust its capacity to a appropriate utilization ‐ 41, 44, 47, 50, 53, 56, 59 y Maximize its profit
Business model y Business Days: 5 days y All collected orders are shipped at 05:30pm on each day, and no late order is accepted. y The leftover half truck‐load orders will be shipped on the following day. y JPR loses the order and incurs stock‐out cost if the capacity is full when an order arrives. y There is only one‐way shipping and trucks return with no loading.
Daily Demand y Average Daily order number:
y Price of Different Destinations
Based on our assumption
Based on our assumption
# of Pallets of Each Order y The Maximum Truck Loading: 20 Pallets y The truck load of each order has a Discrete Distribution.
Based on our assumption
Estimated Distance of Destinations y Travel distance is determined by the total distance of two destinations on Google Map.
Destination
CA
NV
AZ
Distance
554 Miles
580 Miles
650 Miles
Shipping Time Destination Time Per Round Trip (day)
CA Norm (1,0.05)
NV Norm (1,0.18)
AZ Norm (1,0.12) Based on our assumption
Priorities of Destinations y Policy: Highest Revenue, Highest Priority
Based on our assumption
y Two of a half truck‐load (6‐10 pallets)
will be batched to one shipment
Jarrett I’ve died because of
You!!!
Simulation Models 1. Diesel Price ‐ Excel 2. MPG ‐ Excel 3. Order accumulation and Shipment ‐ Arena
Diesel Price y The historical data of 2007‐2009 from U.S. Energy Information Administration y Sample size : 1,000
Result
MPG y The average mileage per gallon diesel of California Transportation Bureau for shipping trucks in California y Sample size: 1,000
3 key components Total : 9 Steps 1.
Assign Orders Truck Loading –Step 2
2. Separate Balking Orders –Step 5 3. Batch the Half Truck‐Loading Orders –Step 6
Step 1 y The demand of order is generated from the parameter we defined.
Schedule Module :
Step 2 y Assign number of pallets to each order based on our assumption of order loading
Coding by Discrete Distribution :
Code
1
2
3
Step 3 y Use 1st Decision module to separate each order by the destination
Step 4 y 2nd Decision module to determine whether the shipment achieve the minimum shipment requirement.
1‐5 pallets
Step 5 y 3rd Decision module to determine whether the shipment need to batch or not
11‐20 pallets
Step 6 y Batch the half truck‐loading orders Wait another half‐loading order
Directly process
Step 7 y Setup shipping time and priority as assumptions.
Step 8 y Record the balking rate for order to different destinations
Step 9 y Repeat Step 4 – 7 for the other two destinations
Run Setup y 5 Days a Week = 5 Days Replication Length y 4 Weeks a Month * 5 Observations = 20 Replications
Run Simulation
Richard
Cost Assumptions Destination Revenue Per Order Operation Cost Per Order
CA $750
NV $800
AZ $825
$200
$208
$217
Inventory Cost Per Truck
$225
Stock-out Cost Per Order
$150
$240
$412.5
Average Distance
554 Miles
580 Miles
650 Miles Based on our assumption
Monthly Profit Analysis Revenue Operation cost Inventory cost Gas cost Stock-out cost Total cost Estimated profit
Price * Shipping times Operation cost pre shipment * Shipping times Total truck fleet * $225 (Distance/MPG)* Gas Price Balking times * Stock‐Out Cost Operation + Inventory + Gas + Stock‐out cost Revenue – Total Cost
Sensitivity Analysis
Findings‐Utilization vs. Profit Chart of Utilization by Different Capacity
Chart of Estimated Profit by Different Capacity
Over‐ capacity
Highest Profit
Findings‐Revenue vs. Cost Chart of Total Cost by Different Capacity
Higher Inventory Cost
Chart of Revenue Cost by Different Capacity
More Revenue
Conclusion y It’s not JPR’s best interest to downsize capacity to 50.
Dad’s company is losing money!
Recommendation‐1 y Fleet size increase
50
53
Recommendation‐2 y Adding smaller‐size trucks Don’t even think I am small!
Possible Improvement‐1 y Increase sample size
Possible Improvement‐2 y Considering seasonality
It will rain or not tomorrow?
Possible Improvement‐3 y Use number of pallets instead of coding to reflect the real world
Let’s make some money! 赚钱啦～
