Arrival Time Provided to the Minute... Automatic Classification of Fake Photo Reviews

Delivery robot tested on Konkuk University campus

Delivery robot tested on Konkuk University campus

View original image


[Asia Economy Reporter Kim Cheol-hyun] Delivery application (hereinafter app) companies are actively incorporating artificial intelligence (AI) technology into their services. The application areas range widely from platforms to delivery sites. AI technology, emerging as a new national competitiveness, is being refined through actual use in the food delivery sector, which is closely connected to everyday life.


According to Woowa Brothers on the 7th, 'Baedal Minjok' recently applied AI technology to its estimated arrival time service. By analyzing delivery riders' satellite positioning system (GPS) information and accumulated customer order data with AI technology, the service provides precise arrival times down to the minute. Woowa Brothers explained that this aims to minimize customer dissatisfaction caused by delivery delays by presenting accurate arrival times.


AI technology is also concentrated in the outdoor autonomous delivery robot that Woowa Brothers recently tested for about a month on a university campus. Highly advanced AI technology is required for this robot to be commercialized since it operates outdoors. The robot must autonomously comply with road traffic laws and detect moving objects such as cars and bicycles to operate safely without accidents. To function stably under conditions like road construction, heavy rain, or heatwaves, the sensor technology and AI technology levels must also be high.


Woowa Brothers plans follow-up tests for this delivery robot. The test conducted last month at Konkuk University Seoul campus involved delivery up to the building entrance. The next phase will involve the delivery robot entering the building, taking elevators, and moving between floors. Kim Yo-seop, Director of the Robot Delivery Cell at Woowa Brothers, said, "Outdoor driving robot tests require a completely different level of technology compared to indoor tests and are quite challenging. Based on accumulated data, we plan to develop more convenient and efficient robot services."


Delivery Hero Korea (hereinafter DH), which operates services such as 'Yogiyo', 'Baedaltong', and 'Foodfly', has developed an AI process that can automatically classify fake photo reviews. This technology was completed through an in-house research project and applies 'deep learning' technology that learns from vast amounts of data to make judgments. DH explained that they enhanced the data recognition level beyond simple object recognition to improve accuracy to about 96% in classifying fake photo reviews.



DH plans to increase the accuracy of this technology and optimize it for delivery apps, applying it to 'Yogiyo'. Once applied, it is expected to filter out fake reviews from the more than 30,000 photo reviews updated monthly. This will allow consumers to refer only to genuine reviews, excluding fake photo reviews when ordering. Son Jin-kwang, a data scientist at DH, said, "This technology is not limited to delivery apps like Yogiyo but can be widely used in various e-commerce services. We expect it to detect and effectively respond to fake reviews that harm consumer experience both commercially and ethically."


This content was produced with the assistance of AI translation services.

© The Asia Business Daily(www.asiae.co.kr). All rights reserved.

Today’s Briefing