BUSM227: Transportation and Logistics Analytics
Coursework
DeliverEase Ltd. is an online webshop specializing in the sale and delivery of home essentials. As
customer demand continues to grow, the management team aims to ensure that the distribution
network has sufficient shipping capacity to meet this increasing demand. Additionally, they seek to
redesign the network to reduce operational costs. Figure 1 provides an overview of DeliverEase’s
current distribution network.
This coursework project focuses on evaluating the performance of the current network design and
using models to identify key factors affecting supply chain performance, as well as to develop
strategies to improve the network's robustness to uncertainty.
Figure 1. Network design Figure 2. Geographical locations of the
customers
Distribution Network
The company currently operates with two suppliers (S1 and S2), three warehouses (W1, W2, and W3),
five distribution centers (DC1 to DC5), and 50 customers (C1 to C50). As shown in Figure 1, products
are first shipped from suppliers to warehouses, then from warehouses to distribution centers, and
finally from distribution centers to customers. Some transportation routes are unavailable: S2 cannot
ship to W1; W1 cannot ship to DC4 or DC5; W2 cannot ship to DC2; and W3 cannot ship to DC1 or
DC2. Each distribution center is responsible for delivering to ten nearby customers. Figure 2 shows
the geographical locations of all customers.
Product Information
The company has selected five products (P1 to P5) for analysis. The full dataset is available in the
file “data.xlsx”. Product weights can be found on the “Product weights” sheet.
1
Shipping operations (Suppliers à Warehouses à Distribution Centres)
Products are transported by barge from suppliers to warehouses, and then by train from warehouses to
distribution centres. Each transportation link has a limited capacity. To use a link, the company must
first pay a fixed cost upfront to reserve its capacity. Only after this reservation can products be shipped
along that link. The “Shipping information” sheet in the dataset provides details on the capacity of
each link, as well as the associated fixed and variable costs. Variable costs are charged per kilogram of
product, regardless of the product type. The total weight shipped on any link must not exceed its
capacity. Note that the shipping capacity and variable costs are based on product weight, rather than
the number of units of product.
Last mile delivery (Distribution Centres à Customers)
Last-mile delivery covers shipments from distribution centres to customers. The company uses its own
vehicle fleet and estimates transport costs based on the distance between each distribution centre and
customer location. The sheet “Last-mile-cost per product per kg” provides unit costs (in euros per kg
per kilometre), which vary by product type. Note that the distribution costs are based on product
weight, rather than the number of units of product. Coordinates for all distribution centres and
customers are listed in “Last mile coordinates”, and distances should be calculated using the Euclidean
formula (assumed in miles). A visual map of the locations is shown in Figure 2. Customer demands
for each product type are provided in the sheet “Demands per product type”, representing the number
of product units ordered by each customer. Note that only whole units of the product can be delivered;
fractional quantities are not allowed.
Tasks
We will follow these tasks to formulate an optimisation model that evaluates the performance of the
current network design and use the results to develop managerial recommendations.
a) Formulate the mathematical model to determine the minimum-cost distribution plan for the
current network: clearly define all sets, parameters, decision variables, the objective function, and
constraints in algebraic form. Briefly explain the purpose of each group of constraints. [25 marks]
b) Implement and solve the mathematical model developed in part (a) using Python and PuLP.
Report the optimal solution, including the total cost and the total weight transported along each
link between suppliers, warehouses, and distribution centers. Present your results clearly and
concisely. [25 marks]
c) Use appropriate visualizations to illustrate the distribution and magnitude of flows across the
network. [10 marks]
d) Provide TWO managerial recommendations, each supported by the results of your optimisation
model and relevant experimental findings. You may also, if appropriate, include results from
alternative formulations to enhance your analysis. [30 marks]
e) Reflect briefly on any limitations or assumptions in your model. Suggest possible improvements
or extensions for future analysis. [10 marks]
2
Reporting and submission
Write clearly and concisely, and ensure your final report remains within the 2,000-word limit (±10%).
Submit both the report and the supplementary files separately via the correct submission links on
QM+ by 30 June at 5pm.
• A PDF report (maximum 2,000 words ±10%, excluding appendices)
• A ZIP folder containing all supplementary files (e.g., Python scripts and datasets)
This coursework assignment encourages you to apply the concepts, methods, and tools covered in
class to a real-world-style problem. The more effectively you integrate modelling, optimisation,
simulation, and analysis techniques, the stronger your work will be. Marks will reflect your
understanding, clarity, and how well you apply ideas from the module.
Feedback will be provided during lecture hours, giving you the opportunity to ask questions, receive
guidance, and strengthen your work before submission.
Important Notes:
• Organise your report clearly according to each of the tasks. Each section should correspond to
a specific task (a–e), with clear headings.
• If the report exceeds the word limit, only the first 2,000 words will be graded.
• Detailed code explanations are NOT required in the report. Instead, use clear graphical
visualisations to demonstrate your methods, data instances, and solutions.
• All Python scripts and data must be placed in a single folder. Do not create subfolders for
individual questions.
• Your code should be well-organised, clearly documented, and enable easy replication of the
results presented in the report.
• Late submissions will incur a penalty of 20% per day.
