Data Science for Sustainability- Green Inventory Management
.
Green inventory management can be defined as managing inventory in an environmentally sustainable way.
This involves a set of processes and rules reducing the environmental impact of order preparation and delivery.
Most retailers use inventory management systems that use a rule-based approach to replenishing stores and covering demand.
The replenishment frequency is an important parameter that can be used to optimize the distribution network.
As a Data Science Manager, can you simulate the impact of the delivery frequency on the CO2 emissions?
As major retail groups should commit to reducing their CO2 emissions by 2030, this can become your strategic tool for reaching these targets.
In this article, we will explore how to use Data Science to simulate the sustainable initiative of implementing green inventory management for the distribution of a fashion retail company.
Problem Statement
Scenario: Inventory Management for Retail
You are the Data Science Manager of an international clothing group that has stores all around the world.
The distribution team is managing the replenishment of stores from local warehouses.
Your colleague, the Inventory Manager, is ** in charge of setting the store replenishment rules in the ER**P.
She has implemented a periodic review policy Order-Up-To-Level (R, S)
- The ERP is reviewing stores' inventory levels (also called inventory on hand) every R days: IOH
- For each review, the gap between the inventory level and the target inventory S is calculated: S – IOH
- A Replenishment Order is created and transmitted to the warehouse with the quantity Q = S – IOH
The idea is to deliver the missing quantity to reach this target level.
After transmission, the order is prepared at the warehouse and delivered to your store after a certain lead time LD (days).
The target stock is defined to absorb the demand variability and the replenishment lead time to avoid stock-outs (empty shelves) at the store.
I will not enter the details of how to set these differences in this article.
However, if you need more information, you can find a detailed explanation in this article.
Among the key parameters of this rule is the time between two reviews, which will drive the frequency of replenishment in your stores.
What if we change this review period?
What is Green Inventory Management?
The review period is setting the frequency of store replenishment order creation.
-
For R = 2 days: stores are replenished very frequently You can set a lower target stock level to cover the demand during the review period.
-
For R = 15 days: stores are replenished less frequently The order quantity per replenishment must be higher as your target stock level needs to absorb the demand during a longer review period.
On the left side, we have more store deliveries (with a lower quantity per shipment) for the same duration.
This will impact the efficiency of your warehouse and transportation operations.
Can we estimate the impacts of these two different approaches on the CO2 emissions?
You can simulate these and help your colleague estimate the impact of her inventory rules on carton consumption and CO2 emissions.
Impact on the carton usageItems are delivered at the stores in cartons containing units picked individually.
If the store orders 5 units of the reference XXX, the operator will
- Open a box of 20 units and take 5 units;
- Take a new box and put these 5 units;
- Complete the box with other items ordered by the store;
We must use additional carton material to create these mixed cartons containing different items. (instead of shipping full cartons)
How many additional mixed cartons do we have to prepare?
You can calculate the total number of mixed cartons using the formula below.
These boxes (or mixed cartons) will require additional packing material that will impact your footprint.
With a higher replenishment frequency, the quantity per replenishment is reduced, and this situation can occur more.
Can we estimate the impact on the Transportation efficiency?
Impact on transportation emissionsAs it is the main driver of CO2 emissions, you should also estimate the impact on the number of trucks used and their filling rate.
The review period impacts the number of deliveries during a certain period.
For instance, doubling the delivery frequency will
- Multiply the number of delivers for the same quantity replenished;
- Reduce the quantity per replenishment and potentially increase the space in trucks
How can we translate these insights in estimated CO2 emissions increase?
In the next section, we will translate these operational insights into a simulation model to select the optimal inventory rules.
Have you heard about Sustainable Supply Chain Optimization?
Simulate Green Inventory Management Scenarios
Build a Model with Assumptions
I will use the example of a logistic network that replenishes ten fashion retail stores in Shanghai (PRC).
In this simulation, I will consider
- 90 days of sales of 10 stores located around the warehouse
- 740 unique items (SKU) sold in these stores
- The number of units per full carton provided by the master data
- 12 pieces per mixed carton
- 1 day lead time between order creation and store delivery
Which metrics to use for the assessment of a specific inventory management rule?
Using these parameters and the formula of the previous section, I can estimate truck fill rates and the number of additional cartons needed.
How to translate that into environmental impacts?
Based on the mixed carton dimension and thickness, we can estimate the additional material quantity per carton: 0.3 kg/Carton.
CO2 emissions are estimated using the NTM (Network for Transport Measures) methodology using distances and emissions factors.
