Solving Reverse Logistics Problem | Walmart

This was work done by me in a team of 3 members for a competition called Walmart Sparkathon 2024.

 PROBLEM STATEMENT: 

The reverse logistics process in large retail organizations like Walmart is complex and often inefficient. When customers return products, these items undergo a multi-layered inspection before being reintegrated into the supply chain. Currently, two primary inspection layers are involved:

1. A primary inspection is performed at the store or customer’s location.
2. A secondary inspection is conducted at a distribution centre or warehouse.

This process, while thorough, has several drawbacks:

● Time-consuming: Each product must pass through two levels of inspection, causing delays in processing returns.
● Costly: The multi-layered inspection adds extra labour and logistical costs.
● Manual Errors: The traditional method relies heavily on manual assessment, which increases the chances of human error and inconsistencies.

In light of these challenges, our solution aims to streamline the reverse logistics process using a tech-driven approach that leverages machine learning to minimize the inspection layers while maintaining high accuracy and efficiency.

Our Solution: Reverse Logistics Application: Our app transforms the traditional reverse logistics process by integrating machine learning (ML) to assist Walmart employees in making quick, accurate decisions about returned products. Here’s how the app works: 

UI and Workflow:

1. Sign Up Page: The Walmart employee starts by creating an account within the app. They provide basic information like name, employee ID (WIN Number), and email to set up their profile. 

2. Login Page: Once the account is created, the employee logs in using their Employee ID (WIN No.) and password. The authentication screen ensures that only authorized personnel can access the system.

3. Mode Selection:
   ●After login, the employee is presented with a dashboard displaying two options: 1. Inspection Mode 2. Resale Mode
   ● For this example, the employee selects Inspection Mode.

4. Product Info Entry:
   ●The employee is asked to scan the product's barcode or manually enter details such as:
   ○ Product Name  ○ Model Number ○ Company
   ●The app then cross-references the entered information with Walmart's internal product database and displays a suggested image of the product.
   ●The employee confirms whether the displayed image matches the physical product by selecting "Yes" or "No."

5. Primary Visual Inspection: Once the product is confirmed, the employee proceeds with a primary visual inspection. The app prompts the user to answer questions regarding the product’s condition, such as:

○ Is the product scratched? ○ Are there broken parts? ○ Has the colour faded?
○ Are screws/vertices loose?

6. Upload Photo and 360 deg video:
● To supplement the visual inspection, the employee uploads photos of the product from different angles (e.g., front, side, top) as per the guidelines.

● Additionally, the employee records and uploads a 360-degree video of the product to provide the ML model with a complete view of the product’s condition.

7. ML Model Discussion: 
● The uploaded images and videos are then processed by the app’s machine learning model, which evaluates the product’s condition based on predefined parameters.

● The model then categorizes the product into one of three options:
1. Send for Repair – The product is damaged but repairable.
2. Reuse – The product is in good condition and can be resold.
3. Recycle – The product is beyond repair and should be recycled. 

This instant categorization helps Walmart save time and reduce costs associated with the secondary inspection.

8. Additional Features:

● Profile Picture Options: ○ When the employee taps on their profile picture, a drop-down list of options appears. These options include:
■ Dashboard: Provides an overview of employee activity and app statistics.
■ Logistics Cart: A section that displays detailed information about all the returned products the employee has processed. This includes:
a.Transportation Status: Whether the product is in transit to a repair center, warehouse, or store. b. Return Destination: Indicates if the product is being restocked in an online or offline store after inspection.
■ Change Mode: Allows the employee to switch between Inspection Mode and Resale Mode.
■ Logout: Signs the employee out of the application.

● Logistics Cart Feature:
○ The Logistics Cart ensures easy tracking of returned products and helps Walmart with better inventory management by providing real-time updates on product locations and their status in the reverse logistics cycle.

This report outlines both the problem of traditional reverse logistics and the solution offered by our app, showcasing how integrating technology can simplify complex retail processes.