Your company is a manufacturer of wave-runner personal watercraft. Currently, you have committed to serving the U.S. customer market by manufacturing the products at two pro- duction facilities. One is in Monterrey, Mexico (MON) and the other is located in Newnan, Georgia (NEW). Production capacity is larger in MON and is currently 160 units per week. Production capacity at NEW is 110 units per week. Your demand planning team has grouped retail customers into eight (8) major customer markets across the U.S. and you refer to them as C1, C2, ..., C8. Weekly demand in each customer market is given as follows: C1 C2 C3 C4 C5 C6 C7 C8 Weekly Demand 18 30 40 12 20 65 15 50 After production, units are first shipped to one of three distribution centers. The distri- bution centers are located in Harrisburg, PA (HAR); Atlanta (ATL); and San Bernardino, CA (SB). Orders from partner retailers from each of the 8 markets are filled from the distri- bution center(s). This study will focus on determining good plans for distributing units from produc- tion facilities to distribution centers and then to the customer markets. Suppose that the value of an average wave-runner unit is $6,000 and that the company uses a 22% inventory carrying-cost rate per year. All freight transportation is conducted by truckload trucking, and each truckload can carry Q = 25 finished wave-runner units. Each of your truckload carriers charges a minimum of $525 per shipment regardless of the distance, but a higher rate per truckload is charged for longer distances by multiplying a cost per mile by the mileage traveled. The attached spreadsheet specifies the average mileage between the production facilities and the distribution centers, and then from the distribution centers to each cus- tomer market. Note that HAR is never used for customers in markets C7 or C8, and SB is never used for customers in C1, C2, or C3. The spreadsheet also gives the order lead time in days between all pairs of locations. When computing the average distribution cost per unit between any two locations, you decide to assume that units are transported in full loads and to also include a pipeline inventory charge based on the lead time. For the truckload per mile rates, your preferred carrier provides individual rates for all of the plant-to-DC lanes but they charge the same rate per mile for all outbound lanes from each individual DC. The spreadsheet shows correct cost computations from the production facilities to the DCs, but leaves it up to you to compute the DC-to-customer market costs per unit shipped.
optimization models to make flow decisions in logistics networks:
Question 3
As a logistics analyst, you are wondering whether or not you could decrease your total
distribution costs if you no longer are constrained to move the same unit volume through
your distribution centers given your changed demand. For the analysis in this part, remove
the constraint that each plant sends at least one truckload a week to each distribution center.
Using the same distribution costs per item from the plants to the DCs and then from
the DCs to the customer markets from the earlier two questions, you decide to build a single
MCNF model that captures flow from plants and flows outbound from DCs to customer
markets simultaneously.
1. Using a minimum cost network flow model, formulate a combined minimum cost dis-
tribution problem. Combining your earlier models, define plant nodes, DC nodes, and
customer nodes in a single model. What are the b values at the plant nodes? What are
the b values at the customer nodes? How about the DC nodes? Draw the arcs needed
for this problem; you need not write down all the arc unit costs again on your picture
if you are re-using the costs from the earlier two questions.
2. Solve your combined MCNF model and report the solution. What is the total cost per
week to both resupply the plants and serve the customers? What is the total cost per
unit of demand? What are the total number of units that flow through each of the
distribution centers each week?
3. Looking at your solution, you decide that it may be time to re-size your San Bernardino
and Atlanta distribution centers. Your real estate team has already found a new,
smaller SB facility that can manage 70 units per week. However, in the tight Atlanta
commercial real estate market, the best available new ATL site for lease can only handle
140 units a week throughput. Propose an approach to add these weekly throughput
capacities (70 for SB, 140 for ATL, and 60 for HAR) to your combined minimum cost
network flow model.
4. Solve your modified model with the throughput upper bounds. By how much does the total distribution cost per week and cost per unit increase given this constraint on the Atlanta facility?
Step by step
Solved in 2 steps with 8 images