Introduction

The pressure to stay ahead in today's competitive business world demands efficient supply chains. One area ripe for improvement is streamlining the goods receipt process.

This case study showcases how computer vision technology can automate the validation of deliveries against purchase orders, boosting efficiency and eliminating human error.

Challenge

The manual verification of goods delivery, specifically here being tinplate’s or steel rims receipts against purchase orders, was a time-consuming and error-prone process for our client. Each delivery receipt contained a QR code encoding essential information such as order number, product details, and quantity. Manually matching this information against corresponding purchase orders in the ERP system not only consumed valuable time but also led to occasional discrepancies and delays in inventory management.

Implementation

Data Collection and Training

We collected a large dataset of goods delivery receipts containing QR codes along with corresponding purchase order data. This dataset was used to train a deep learning model capable of accurately recognizing and decoding QR codes under various lighting conditions and orientations.

QR Code Recognition

We developed a computer vision pipeline using state-of-the-art libraries and frameworks to preprocess images, detect QR codes, and extract information encoded within them. The trained model was deployed to efficiently recognize and decode QR codes from incoming delivery receipts.

Data Validation

Upon successful QR code recognition, the extracted information was compared against the corresponding purchase order data stored SharePoint online. Any discrepancies or mismatches were flagged for manual review, while accurately matched deliveries were automatically validated and updated from SP online into ERP system.

Integration with ERP System

The computer vision system was seamlessly integrated with the client's existing ERP system, allowing real-time synchronization of validated delivery information. This integration ensured that inventory levels were updated promptly, enabling better inventory management and order fulfillment processes.

Result

The implementation of computer vision technology for streamlining the receipt of goods delivery yielded significant improvements in efficiency and accuracy for our client. By automating the validation process, the time required to process each delivery receipt was reduced by an average of 90%. This resulted in faster turnaround times for inventory updates and reduced the risk of manual errors, leading to enhanced operational efficiency and cost savings.