Digital stores are a collection of digital technologies and applications from a variety of sectors, in which shelf management plays a key role. The importance of shelf management to the retail business means that improving on-shelf availability (OSA) has always been a top priority. A study by the IHL Group found that the out-of-stock (OOS) ratio of the retail industry reached as high as 8%, accounting for 4.1% of lost revenue for the average retailer. Another study found that more than 50% of OOS was caused by poor shelf operation, namely inaccurate inventory, misplaced SKUs, and related shelf inventory issues.
The Importance of Shelf Management in Digital Stores
The conventional method of strengthening shelf management has been manual observation, recording, and review. There are however many problems with this method. Supermarkets are typically made up of a large number of shelves holding a wide variety of products. Manual shelf management carries a high manpower cost and can only be carried out at a defined frequency, making it difficult to realize timely insights.
To solve these problems, shelf management solutions based on AI, computer vision, and other technologies have emerged. Through the identification, classification, and inventory of shelf items with algorithms such as price tag recognition, product recognition, OOS detection, activity recognition, and optical character recognition (OCR), these solutions are capable of automatically determining OOS items, incorrect placements, and other issues. Relevant data can also be transmitted to backend systems for processing to support further data processing and decision-making.
Hanshow AI Digital Store Solution Features and Benefits
The Hanshow AI digital store solution provides a variety of shelf management options, including fixed camera and robot inspection. In specific, the fixed camera solution is suited for scenarios in which displays change frequently and must be checked for abnormalities in real-time. The solution allows retailers to conduct real-time display monitoring and product recognition while obtaining data related to human-product interaction. The robot inspection solution is suited for scenarios that have a large number of products, but no high-frequency requirements.
The solution realizes the collection, monitoring, and verification of display data through digital price tags and product recognition, ensuring that retailers are notified of any OOS or other abnormal events promptly. While ensuring compliance with consumer privacy and data protection laws and regulations, the solution also realizes the recognition and tracking of consumer attributes, record data such as consumer browsing time and location, generate in-store heat maps and customer flows, and assist in targeted marketing.
With Microsoft Azure Cognitive Services, our solution can deliver a low-friction, state-of-the-art algorithm model, with a seamless and highly secured user experience. No machine-learning expertise is required. The terminal device can accurately execute assigned tasks even in an intensive space, which greatly improves data accuracy and integrity. Meanwhile, Microsoft Azure Machine Learning is adopted with the aim of accelerating time to value with industry-leading MLOps and integrated machine learning tools. Azure Machine Learning also allows you to access timely and accurate data analysis and forecasting with its high transparency, reliability, and compliance.
