This image provided by Samsung Welstory Inc. shows the catering company’s leftover analysis system.
SEOUL, Feb. 10 (Korea Bizwire) — Samsung Welstory Inc., a catering service affiliate of Samsung Group, said Thursday it has developed an analysis system to analyze the type and amount of leftovers by scanning the plates returned by customers.
With cameras installed at the dish return window of a cafeteria, the system automatically scans the plates returned by customers and analyzes the type and amount of leftover food.
To ensure that the artificial intelligence (AI) can recognize and analyze numerous types of food using a camera, the company carried out machine learning methods by taking photos and classifying 22,000 plates and 100,000 food data records.
Samsung Welstory plans to use this system to comparatively analyze changes in the quantity of leftovers depending on main dishes, with the goal of finding the optimal composition of dishes favored by customers.
By analyzing the collected data, it also plans to reduce the generation of food waste by calculating the adequate amount of food offerings.
Starting this month, the company introduced the new system at some workplaces, collecting and analyzing 2,500 plates and 10,000 pieces of food data on a daily basis.
Based on the analyzed data, it provides information on the amount of food, food composition and customer preferences to nutritionists and cooks at each workplace twice a month.
Ashley Song (ashley@koreabizwire.com)
