THE INTERNATIONAL UNIVERSITY (IU)

School of Industrial Engineering & Management

----------------------------------------------------------------------------------------------------------------------

MIDTERM EXAMINATION RUBRIC- ASSESSMENT ANALYSIS

Course name: Logistics Engineering and Supply Chain Design

Academic year: 2020-2021
ASSESSMENT ANALYSIS
Goals Descriptions Program Learning Level of Question
Outcomes Competence
(Gx) (Description & points)
ABET CDIO

G1 Understanding of Q1 (10 pts):

concepts and key
points of Supply Understanding ways to
Chain Management. face with demand
variability with inventory
4,7 Understand management
4.1
Q2 (20 points):

Understanding about
distribution networks and
their applications

G2 Understanding of Q3 (40 points):

Logistics and Supply
Chain structure, and Apply Cost Analysis and
how to design an 1,2 1.1, 1.2 Apply EOQ Model to
effective supply recommend the better
chain option

G3 Formulate and solve Q4 (30 points):

problems related to
logistics and supply Formulate and solve
chain with 3,5 1.3, 3.1, problem related to a
Analyze
optimization 4.1 symmetric game, with bi-
techniques. matrix, Nash equilibrium
development

MARKING SCHEME
Question 1: (10pt)
 THE INTERNATIONAL UNIVERSITY (IU)
School of Industrial Engineering & Management
----------------------------------------------------------------------------------------------------------------------

In terms of inventory management, list 3 ways firms can do to cope with huge variability in
customer demand.
• SOLUTION:
Student can list 3 points out of the following:
− Hold inventory to protect against uncertainty in demand
− Hold safety stock during lead time to protect against deviations from average demand during
lead time
− Move from a decentralized system to a centralized system to aggregate demand, which helps
reduce demand variability
− Develop various demand forecasting techniques to have an effective inventory management
If student has another point that’s irrelevant, give 0.5 – 1 for extra points. If another point
makes sense, give full credit.
Question 2: (20 points)
Gateway, a PC manufacturer, opened their retailing stores throughout the US in the late 1990s. Its
stores carried no finished-goods inventory, and were primarily focused on helping customers select
the right configuration to purchase. All PCs were manufactured to order and shipped to the
customer from one of the assembly plants.
2.1. (10 points) List 2 advantages that Gateway’s chain offers.
2.2. (10 points) What is Gateway’s current distribution network? Is there any other distribution
network that might fit Gateway’s products?
• SOLUTION:
2.1. (10pt) Several advantages:
− High product variety
− Low finished-goods inventory
− Increasing customer experience with store presence
If student has another point that’s irrelevant, give 0.5 – 1 for extra points. If another point
makes sense, give full credit.
2.2. Gateway’s current distribution network is Manufacturer storage with direct shipping.
(3 pt)
Other distribution network might fit Gateway’s products must involve Manufacturer storage for
high product variety. Since PCs are usually sent in one shipment, there is no need for in-transit
merge. Therefore, Manufacturer storage with customer pickup is another fit. (7pt)
Question 3: (40 points)
The following table provides historical data for two typical products of ABC. The table includes
weekly demand information of the two products in thousands of units for the last 6 weeks in each
market area.
 THE INTERNATIONAL UNIVERSITY (IU)
School of Industrial Engineering & Management
----------------------------------------------------------------------------------------------------------------------

Week Week Week Week Week Week

Product A
1 2 3 4 5 6
Binh
16 18 23 25 17 22
Duong
Bac Ninh 11 18 15 23 20 19

Week Week Week Week Week Week

Product B
1 2 3 4 5 6
Binh
4 4 9 6 7 8
Duong
Bac Ninh 9 7 7 13 10 11

Table 1: Historical data for two products (unit: thousands of units)

The weekly demand is assumed to be normally distributed. It costs $200 whenever a warehouse
places an order to the factory. The unit annual holding costs is $7.8 per hundred of units. It usually
takes 2 weeks for the factory to fulfil an order. An increasing competition has pushed ABC to find
ways to reduce costs. In order to do so, ABC is considering an alternative distribution strategy in
which the two regional warehouses are replaced with one single, central warehouse to fulfil all
customer orders. The CEO insists to maintain the service level of 98 percent.
In the current system, the unit inbound transportation cost to each warehouse is $0.4 per hundred
of units, while the unit outbound transportation cost is $0.6 per hundred of units. Table 2 provides
information about unit transportation costs from each existing warehouse to other market areas.

Warehouse Binh Duong Bac Ninh

Binh Duong 0.2 0.8

Bac Ninh 0.75 0.30

Table 2: Unit transportation costs (unit: $/ hundred of units)
3.1. Which product would benefit more when ABC changes their distribution strategy? Explain.
(15pt)
3.2. Given that the decision over the distribution system is based on cost comparison over the
product that benefits less from the shift to a centralized system (as identified in the previous
question), should ABC adopt the alternative system? If yes, which warehouse should be chosen as
the centralized warehouse? (25 pt)
• SOLUTION:
3.1.
Product A (4pt) Product B (4 pt)
 THE INTERNATIONAL UNIVERSITY (IU)
School of Industrial Engineering & Management
----------------------------------------------------------------------------------------------------------------------

𝑺𝑻𝑫 𝑺𝑻𝑫
AVGA STDA CoVA=𝑨𝑽𝑮𝑨 AVGB STDB CoVB=𝑨𝑽𝑮𝑩
𝑨 𝑩

Binh
20.17 3.66 0.18 6.33 2.07 0.33
Duong
Bac Ninh 17.67 4.18 0.24 9.50 2.35 0.25
Centralized 37.83 6.85 0.18 15.83 3.25 0.21
Since Product B has Coefficient of Variation higher than Product A in Binh Duong warehouse,
Bac Ninh warehouse, and in the Centralized system, product B benefits more from the change in
the distribution network. (2pt)
3.2. We compare the total cost of the two systems for Product A (1pt)
7.8
Weekly holding cost 𝐶ℎ = = $0.15 𝑝𝑒𝑟 ℎ𝑢𝑛𝑑𝑟𝑒𝑑~$1.5 𝑝𝑒𝑟 𝑡ℎ𝑜𝑢𝑠𝑎𝑛𝑑 (1pt)
52

Safety factor 𝑧 = 2.05 (1pt)

Lead time 𝐿 = 2 𝑤𝑒𝑒𝑘𝑠 (1pt)
In the current system: unit inbound cost = $4 per thousand; unit outbound cost = $6 per thousand
(1pt)
In the centralized system: The unit transportation cost is per thousand of units as follows: (1pt)

Warehouse Binh Duong Bac Ninh

Binh Duong 2 8

Bac Ninh 7.5 3

Safety Stock 𝑆𝑆 = 𝑧 ∗ 𝑆𝑇𝐷 ∗ √𝐿

2𝐶𝑝 ∗𝐴𝑉𝐺
Order quantity 𝑄 ∗ = √ 𝐶ℎ

𝑄∗
Average inventory = + 𝑆𝑆
2
𝐴𝑉𝐺
Number of order per week 𝑁 = 𝑄∗

Product A (5pt) SS Q* Average inventory N

Binh Duong 10.6 73.33 47.27 0.275

Bac Ninh 12.12 68.64 46.44 0.26

 THE INTERNATIONAL UNIVERSITY (IU)
School of Industrial Engineering & Management
----------------------------------------------------------------------------------------------------------------------

Centralized warehouse 19.87 100.44 70.09 0.38

For the current system

Weekly holding cost =1.5 * (47.27+46.44)=140.55

Weekly ordering cost (= unit ordering cost * N) =200*(0.275+0.26)=106.48

Weekly inbound cost (= unit inbound cost * AVG) =4*(20.17+17.67)=151.33

Weekly outbound cost (=unit outbound cost =6*(20.17+17.67)=227

*AVG)

Total cost 625.36 (2pt)

For the new system:

Weekly holding cost= 1.5*70.09=105.14
Weekly ordering cost = 200 *0.38=75.33
Weekly inbound cost = 151.33
Weekly outbound cost:

Warehouse Weekly outbound cost Total cost

Binh Duong 2*20.17 + 8*17.67=181.67 513.47 (2pt)

Bac Ninh 7.5*20.17 + 3*17.67=204.25 536.05 (2pt)

Since the new system save costs, comparing with the current system (513.47 or 536.05 vs. 625.36)
→ should change. (2pt)
Binh Duong should be chosen as the location for the centralized warehouse. (1pt)
Question 4: (30 points)
Two retailers compete in a symmetric game on choosing a right order quantity among 8000, and
10000 units of a product. The wholesale price is $6 per unit. Assuming each customer would
possess a loyalty card from his favourite retailer, and prefers to buy there. Each retailer charges
customers with loyalty card $11 per unit, while charging customers without loyalty card $11.5 per
unit.
If customers cannot buy from his favourite retailer, they will seek to buy in the other retailer. Once
one retailer doesn’t have enough inventory to fulfil the demand, the seeking customers are willing
to pay $13 per unit at the other retailer. Each retailer has the probabilistic demand forecast as
follows: (C6)
 THE INTERNATIONAL UNIVERSITY (IU)
School of Industrial Engineering & Management
----------------------------------------------------------------------------------------------------------------------

Demand 8000 9000

Probability 40% 60%

4.1. What is the seeking customer’s surplus if he finds alternative? (5pt)

4.2. Develop a bimatrix for this game, and find the Nash equilibrium with maximin strategy (20pt).
4.3. Find the Nash equilibrium with best response strategy (5pt).
• SOLUTION:
4.1. CS = WTP – P = 13-11.5=1.5
4.2. 2 players (Retailer 1 & 2), 2 moves for each player (order quantity of 8000, and 10000)
Here is an example, the rest follows the same logic:

R2

8000 9000

8000 16% 24%

R1

9000 24% 36%

If R1 chooses 8000, R2 chooses 10000, then

Demand scenario R1 R2

R1 = 8000, R2=8000 (11-6)*8=40 11*8-6*10=28

R1 = 8000, R2=9000 (11-6)*8=40 11*9-6*10=39

R1= 9000, R2 = 8000 (11-6)*8=40 11*8+11.5*1-

6*10=39.5

R1= 9000, R2 = 9000 (11-6)*8=40 11*9+11.5*1-

6*10=50.5

Expected profit 40 41.5

If R1 chooses 8000, R2 chooses 8000, then

Demand scenario R1 R2

R1 = 8000, R2=8000 (11-6)*8=40 (11-6)*8=40

R1 = 8000, R2=9000 (11-6)*8=40 (11-6)*8=40

 THE INTERNATIONAL UNIVERSITY (IU)
School of Industrial Engineering & Management
----------------------------------------------------------------------------------------------------------------------

R1= 9000, R2 = 8000 (11-6)*8=40 (11-6)*8=40

R1= 9000, R2 = 9000 (11-6)*8=40 (11-6)*8=40

Expected profit 40 40

If R1 chooses 10000, R2 chooses 10000, then

Demand scenario R1 R2

R1 = 8000, R2=8000 11*8-6*10=28 11*8-6*10=28

R1 = 8000, R2=9000 11*8-6*10=28 11*9-6*10=39

R1= 9000, R2 = 8000 11*9-6*10=39 11*8-6*10=28

R1= 9000, R2 = 9000 11*9-6*10=39 11*9-6*10=39

Expected profit 34.6 34.6

Bimatrix: (units in thousands USD)

R2

8,000 10,000

8,000 40, 40 41.5, 40

R1

10,000 41.5, 40 34.6, 34.6

2 Nash equilibria: (8K, 8K) and (10K, 10K)

4.3.

R2

8000 9000

8000 40, 40 41.5, 40

R1

9000 41.5, 40 34.6, 34.6

2 Nash equilibria: (8K, 9K) and (9K, 8K)