Department Store Micromarket Innovates with AI Self-Checkout

March 13, 2024

OasisTime, a micro-market operator, has been setting up micromarkets within factories, logistics centers, and office spaces across Japan. The goal is to create a cozy break area and serve as an oasis for employees. Without needing to leave the work premises, employees can easily buy snacks, grab a cup of coffee, and even purchase their breakfast or lunch.

Recently, OasisTime integrated AI image recognition technology into its self-checkout system to enhance the user experience and deployed it in a large, well-known department store in Sendai.

▲ The micromarket that OasisTime creates for department store employees

No Need to Locate Barcodes! Self-Checkout with AI
An employee at the department store shared, “This is my first experience with AI-powered self-checkout, and it’s incredibly user-friendly! No need to search for and scan product barcodes. The process is fast and straightforward!”

The current usage of OasisTime’s AI-powered self-checkout system ranges between 300 to 400 transactions per day. All transactions are completed using AI recognition. With over 1,000 SKU items available in-store, AI effortlessly identifies each item, eliminating the need to scan each barcode.

AI Allows for Better Product Selection: barcode-less beverage cups and small items
Empty beverage paper cups can now be identified with AI, saving users the trouble of searching through an order menu.

AI also enables the selling of individual items typically bought in bulk. Since these non-bulk individual items do not have barcodes, they cannot be sold with a barcode self-checkout system. With AI, these and other barcode-less items can now be recognized and sold easily.

Even miniature chocolates as small as 3 centimeters in length can be identified by AI. The AI’s recognition capability applies to all product items, enhancing the flexibility and variety of items sold in a micromarket.

Minimizing Loss with Image Records
AI not only helps to identify checkout items but also records the result of each transaction in an image, thereby assisting OasisTime in reducing shrinkage and losses.

Coupled with OasisTime’s on-premises security cameras, the recognition records can further help OasisTime analyze customer behavior to understand whether mis-scanned items or lack of scanning were inadvertent or deliberate. This has not been easy to determine in the past with a barcode scanning self-checkout system. With AI recognition, it can now be clearly determined whether a customer intends to pay for all items, as all items need to be first placed down in the checkout area.

Transitioning Users to AI Technology
OasisTime believes that encouraging users to utilize AI technology is a process requiring familiarity and education, similar to the introduction of self-checkout systems. However, OasisTime President Hideaki Harada stated, “We were surprised by the use case of the department store micromarket; we found that getting users accustomed to using AI wasn’t as difficult as we imagined, and usage quickly ramped up.”

For other consumers who are accustomed to scanning barcodes for self-checkout, there may be initial concerns about the implementation of AI, such as not knowing how to use it, being afraid to use it, or fearing misuse.

However, Mr. Harada confidently asserts, “Ultimately, consumers will become familiar with it. Just like how they embraced other technologies like the Internet, smartphones, and various ride-hailing, food delivery, and social apps, people will gradually get used to AI through exposure, education, and repeated use. Moreover, using AI-powered self-checkout is straightforward: you just place the items under the camera, scan, and complete the checkout. It’s simpler than using any app.

The key advantage of Viscovery’s AI image recognition system is its remarkably low capital expenditures. The setup requires only a camera and application software. There is no need to invest in and maintain complex machines. As the solution is cloud-based, rolling out to new locations is quick and simple.