[Edaily Reporter Shin Yeong-bin ] SHINSEGAE INFORMATION & COMMUNICATION Co., Ltd. (#SHINSEGAE INFORMATION & COMMUNICATION Co., Ltd.) is enhancing its AI vision-based retail solutions and accelerating its push into the domestic and international AI Transformation (AX) market for brick-and-mortar stores.

Shinsegae Co.,Ltd. announced on the 28th that it has enhanced its AI checkout solution—which recognizes multiple items at once without the need to scan each product’s barcode individually at self-checkout counters—and has rebranded it as “Sparos Scan-Free.”

Sparos Scan-Free is a solution where AI vision simultaneously recognizes multiple products when they are placed on the checkout counter. While the company’s previous solution recognized products with 99.5% accuracy in one second, Sparos Scan-Free recognizes products with 99.9% accuracy in less than 0.4 seconds, according to the company. Shinsegae Co.,Ltd. I&C Technology Co., Ltd. stated that this can reduce customer wait times by up to nine times compared to standard self-checkout counters.

The range of recognizable products has also expanded to over 5,000 types. In addition to manufactured goods, the system can recognize fresh foods such as fruits and vegetables, as well as barcode-less items—like bread, where even the same product may vary in shape.

A feature has also been added that presents customers with a list of options to choose from when the AI has difficulty identifying a product. The system is designed so that the AI does not arbitrarily provide answers but instead learns from customer feedback in real time, improving recognition accuracy the more it is used.

Currently, Sparos ScanFree has been implemented at the smart store within the SK hynix Icheon Campus. Shinsegae I&C Technology Co., Ltd. is in discussions to roll out the system to various retail stores, including bakeries, ready-to-eat food specialty shops, and household goods stores.

“Sparos ScanCare,” launched in 2024, is also showing positive results. ScanCare is a solution that uses AI vision to analyze self-checkout counter usage in real time and guide customers to scan products correctly. It detects situations where items are not scanned due to customer errors or barcode issues, prompts customers to rescan, and notifies staff if necessary.

According to Shinsegae Co.,Ltd. I&C Technology Co., Ltd., at one retail store that adopted ScanCare, the frequency with which customers called for staff assistance dropped to one-fifth of the rate seen in standard self-checkout environments where pre- and post-checkout weights are compared. Losses resulting from missed scans were also reduced by up to several billion won annually. However, these results are based on a specific store case and may vary depending on the country, industry, and store environment.

Shinsegae I&C Technology Co., Ltd.’s AI vision solution is designed with an edge AI architecture that processes image and video data within the store without requiring a separate server or external graphics processing unit (GPU). The company explains that since data is processed on the device itself, this enhances data security and allows for implementation without significantly altering existing store systems. Implementation costs have been reduced to about one-fourth of those of competing solutions.

Building on its success in the domestic market, Shinsegae Co.,Ltd. plans to expand into Asian markets, including Japan. The company believes there is growing demand in the Japanese retail market—where cashless payments and self-checkout counters are becoming widespread—for technology that enhances checkout speed, accuracy, and operational stability.

The company plans to identify franchise clients—including supermarkets, convenience stores, cafes, and bakeries—and expand the supply of AI vision-based retail solutions.

Yang Yoon-ji, CEO of Shinsegae I&C Technology Co., Ltd., stated, “Successful offline retail AX depends not only on the adoption of AI solutions but also on how accurately and stably stores can be operated,” adding, “We plan to accelerate our expansion into both domestic and overseas markets.”

She added, “In the future, we will develop this into an expanded solution that supports customer service, product guidance, and store operations by integrating agent functions based on large language models (LLMs).”