Restaurant safety check? New AI platform watches, reports violators
One of China's largest state-owned enterprises is paying seven figures to data intelligence platform KanKan to provide facial- and object-recognition tech that helps to ensure ensure restaurants comply with local food safety laws. A news release said a Shanghai municipal health agency is having the technology put in 200 restaurants initially, a news release said, but plans to eventually install it in 10 times that number.
The platform, developed by a subsidiary of Remark Holdings, uses AI to determine whether workers are wearing hats and masks as required. The mostly automated and scalable system is designed to eliminate human error.
The tool instantly analyzes the images obtained from KanKan's camera-mounted AI devices in restaurant kitchens. Software detects specified food safety law violations and pulls out images with violation details that the agency can review.
The video below demonstrates how the system works; red borders on the facial-tracking boxes identify workers not wearing masks, a violation of food safety standards.
"This contract shows that we are able to monetize our artificial intelligence technology, including facial and object recognition," Remark Holdings Chairman and CEO Kai-Shing Tao said in the release. "It sets the foundation for us to potentially triple our business with the city of Shanghai in 2018 and expand to most of the 8 million restaurants located in cities throughout China."
The company uses a type of algorithm-based machine learning called "deep learning" to model high-level abstractions in data and train its AI products. KanKan's facial- and object-recognition technology has shown a 96 percent accuracy rate, the release said.
The platform is designed as a one-stop shop item for all sizes of entities and can be customized and trained to each business's AI models.