Vinasoy 用 AWS 生成式 AI 监测货架陈列，缺货率降 20%

Vinasoy says an AWS-based generative AI system for retail display monitoring reduced its out-of-stock rate by 20 percent within two months of deployment.

The Vietnamese soy milk manufacturer developed the system with Renova Cloud. It uses photographs taken during store visits to assess product displays and return compliance scores to sales teams within one to two days.

Le Ba Be, National Sales Director at Vinasoy, said: “Our sales teams used to wait weeks for display compliance data, and by the time they got it, the shelf problem had already cost us. With AWS’ Generative AI solutions, they now get answers in days and can act before a gap becomes a lost sale.”

Vinasoy manages product placement across a retail network spanning 34 provinces. Before the deployment, staff manually assessed store product-layout photographs through the company’s Distribution Management System (DMS). Each scoring cycle required nearly 2,000 work hours and took more than 20 days to complete.

That process covered approximately 17 percent of outlets each month, with individual stores monitored once a month. Vinasoy now reports monthly coverage of more than 70 percent of outlets and a weekly monitoring frequency.

“This has freed our people to do what they do best: build relationships with retailers and keep our products within reach of consumers,” explained Le Ba Be. “We’re just getting started, and we see this becoming the foundation for how we run our entire supply chain.”

AWS SageMaker identification and Bedrock scoring

The workflow begins with salespeople photographing product displays during store visits and uploading the images to the DMS. Amazon SageMaker processes those images using Vinasoy’s own image recognition technology.

The company says the recognition system identifies products, including 24 versions of Vinasoy milk with different packaging and sizes, at 22 images per second. It describes that processing rate as 1,300 times faster than manual inspection. This comparison concerns image recognition, distinct from the time needed to complete the display assessment and deliver scores.

Amazon Bedrock powers the generative AI stage, which evaluates each store’s display against approved product display standards. According to the announcement, the system can assess imperfect photographs and unusual shelf arrangements while retaining accurate results. However, no numerical accuracy measure is supplied.

Sales teams receive the resulting compliance scores within one to two days, which Vinasoy describes as up to 20 times faster than the previous process. Those scores help teams identify stock issues and displays that do not meet the company’s standards.

The shorter reporting cycle gives sales staff earlier access to store-level findings. Vinasoy attributes the 20 percent reduction in its out-of-stock rate across the retail network to the deployment, alongside the increase in monitoring coverage and frequency.

Vinasoy plans wider AWS AI deployment

Retail display monitoring forms part of Vinasoy’s wider use of AI. The company says it is applying AI-driven data analytics to sales management, supporting demand forecasting and distribution optimisation.

Vinasoy reports that preparing a sales report for management meetings now takes approximately five minutes, compared with around three days previously. That is a separate reported workflow result from the retail image-processing figures.

Eric Yeo, Country General Manager for AWS Vietnam, commented: “Vinasoy’s journey is a strong example of how FMCG companies can move from AI experimentation to meaningful business impact.

“What started as a single use case in retail display monitoring has already delivered measurable results, and we’re proud to support Vinasoy as they continue to explore new ways AI can enhance their operations.”

Vinasoy plans to expand display monitoring to all its outlets in the coming months. The company is now working with Renova Cloud on an AI-based packaging inspection solution on AWS for evaluating product packaging.

See also: Oracle Fusion drives supply chain planning for Combe

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events including IoT Tech Expo and the Cyber Security & Cloud Expo. Click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.