Simulate the Challenges of a Circular Economy for Fashion Retail
The concept of a circular economy includes models aiming to reduce waste and improve resource efficiency.
Some fashion retailers have implemented a subscription model with customers paying a regular fee to rent a product for a specific period.
Have you ever considered renting your clothes?
In a previous article, I used data analytics to simulate this rental model using an example of a Fashion retailer with ten stores.
The objective was to estimate the environmental performance, i.e., reducing CO2 emissions and water usage.
However, logistic operations will face additional challenges in collecting and processing rented items to support this transition.
As a data scientist, can you assess the operational challenges of implementing this model?
This article will use data analytics to estimate these challenges and understand which metrics are crucial to redesigning the logistic network.
The aim is to help sustainability and logistics teams build a solid business case (with ROI and risk assessment) to get the green light from top management for the transition to a circular economy.
Summary
I. Implement a Circular Model for Fashion Retail
1. Problem Statement: Sustainability Roadmap of a Fashion Retailer
A fashion retailer would like to implement a rental model in 10 stores
2. Introduction of the Rental Model
Simulation of the impact of this model on CO2 emission and water usage
3. How do we implement these processes?
Use analytics to derive the metrics needed to design logistics solutions
II. Data Analytics to Monitor a Rental Circular Model
Focus on each leg of the distribution network to collect KPIs
1. Focus on the store operations
Manage daily transactions per type (linear, circular)
2. Focus on warehouse operations
Impact on the upstream and downstream flows
3. Focus on transportation management
Organisation routing for returned items collection
III. Support the transition to a circular model with data analytics
1. Business Intelligence for operational monitoring
Collect, process and harmonize data from multiple systems
2. Advanced Workforce Planning for Store Operations with Python
Optimize the number of staff recruited at stores
3. Return Flow Optimization with Python
Allocation of the collection centers for returns
4. Measure Scope 3 Emissions of your Distribution Network
Measure the CO2 emissions impact of your collection routes
IV. ConclusionImplement a Circular Model for Fashion Retail
Problem Statement: Sustainability Roadmap of a Fashion Retailer
You are the Data Science Manager in the Supply Chain department of an international fashion retail group.
Your CEO publicly announced last year the company's commitment to supporting the United Nations Sustainable Development Goals (SDGs).
As part of this commitment, the company aims to reduce its environmental footprint along its entire value chain.
As a data scientist, how can you support this transformation?
You focus on assisting the sustainability and Logistics teams in assessing and designing solutions to implement a circular rental model.
This initiative began with a study involving ten stores for a total of 400 unique items.
What is the conclusion?
The results demonstrated that shorter rental periods maximize the efficiency of the circular model.
However, this study only focused on the environmental benefits without considering the operational challenges that stores and logistics teams face.
What are the impact on the logistic budget?
The teams must justify the return on investment (ROI) and estimate the budget before receiving approval from the top management.
Therefore, we will shift the attention to the operational side and use data analytics to estimate the additional workload at each stage of the distribution network.
Let me first briefly introduce the rental model.
Introduction of the Rental Model
To reduce the environmental impact of its Supply Chain, your company experimented with a circular rental model in 10 stores.
These locations would propose a rental subscription model to their customers for a limited scope of 400 items.
Before implementing this service, logistics and Sustainability teams requested support in simulating the processes to handle these additional flows.
As input data, we used actual sales transactions, as shown below.
The simulations are covering ten stores for a period of 365 days.
We assumed these "sales transactions" were "rental transactions", meaning the customer goes to the store to rent a specific item for n days.
After the rental period ends, customers return the items to the store
- It takes two days to have these items collected from the stores.
- They go through inspection and cleaning at the warehouse for one day.
- It takes another day to have the cleaned items delivered to stores.
How do we manage the flows rented items?
We apply the First-In, First-out (FIFO) principle to manage these flows
- After return and sorting, items are available at the warehouse
- The first store to order will be replenished with the first items returned
When a customer requests a specific item, we have two scenarios:
- Circular Transaction: the store will rent out returned clothes if there is sufficient inventory.
- Linear Transaction: The store will rent a new item if no returned items are available.
How does it look like for a specific store?
In the chart below, you visualize the volume of items rented daily with the split between linear (new items) and circular (returned items) transactions.
As you can see, the circular model starts on day 13 after the first rented items are returned (and cleaned).
Which metric do we want to estimate?
The focus was on measuring the model's environmental performance using several scenarios with different rental periods (2, 7, 14, and 28 days).
For each scenario, we calculate the percentage of circular transactions (items reused) and the impact on CO2 emissions reductions.
Sustainability Team: We can reach the highest footprint reduction with short rental periods.
For more details, have a look at this article
Data Science for Sustainability – Simulate a Circular Economy
From a sustainability point of view, this is a very insightful study.
However, it does not cover the impact on the distribution chain.
What about the operational challenges?
As a supply chain professional, I want to know how to include this operation in the current distribution model.
- Implementing short rental periods means having a high frequency of item rotation.
- Can we organise the collection with the same fleet that delivers stores?
- What about the additional workload at the warehouse for sorting and cleaning?
Let's use data analytics to answer these questions.
How do we implement these processes?
As a former Supply Chain Solution Designer, my approach starts by collecting the metrics needed to design the transportation and warehouse solutions for these reverse flows.
These solutions should cover
- The collection of returned items from stores with electric trucks. Question: How many trucks are needed to collect items every day?
-
Receive and process items in the warehouse. Question: How many operators are needed to receive and process items in the warehouse?
-
Deliver the cleaned (returned) items back to the stores. Question: How many trucks are needed to manage these deliveries?
In the following section, we will cover the main metrics needed to assess the additional workload on your distribution network.
Data Analytics to Monitor the Logistics Network for a Rental Circular Model
Let us now explore the simulation model from a logistic point of view focusing on the goods flows.
We will assume that the rental period is seven days.
First, we can monitor the rental transactions at the stores and the impact on returned items.
What happens at the stores?
Focus on the store operations: daily transactions per type
The circular model starts on day 1
- 100% of the items rented are new ;
- These items will be returned on day 8 to start a process of collection, processing and delivery, ending on day 13 ;
From day 13, the store will receive returned items and be ready to start the second rental cycle.
What are the impacts for the store teams?
