日本語 
Acroquest Technology Co., Ltd. is pleased to announce that it has implemented an automated visual inspection system using generative AI for the Wi-Fi router rental business provided by Vision Co., Ltd., thereby improving the efficiency and quality of inspection work.
Vision Co., Ltd. offers rental Wi-Fi services to meet a variety of needs, including “Global WiFi®” for overseas use, “NINJA WiFi®” for visitors to Japan, and “WiFiRental.com®” for domestic residents. As the number of users increased, the number of devices handled also increased rapidly, and the work hours for visual inspections at four locations nationwide exceeded 200 hours per month. Conventional visual inspections placed a heavy burden on personnel, so a system was needed to improve efficiency while maintaining quality.
Acroquest has built an appearance inspection AI agent that utilizes generative AI using Amazon Bedrock . By simply taking a photo of the device with a tablet app, it automatically detects scratches, dents, swollen batteries, and USB terminal abnormalities, and allows users to check and record the results on the spot. Prompt tuning enables automated inspection with accuracy close to that of human inspection, without the need for training data.
Also Read: WingArc1st Launches AI Assistant for SVF Chatbot
Effects of implementation
Monthly working hours were reduced by 30%, from 200 hours to 140 hours, resulting in cost reduction.
We have expanded the number of inspection locations from four to six, establishing a system for rapid inspection of returned devices.
Even staff without specialized knowledge can perform inspections of a certain quality, reducing the burden of training personnel.
As a result, Vision Co., Ltd. aims to “continuously improve testing accuracy” and “sustainable growth through the promotion of digital transformation,” and is working to provide even more reliable services to its users.
Acroquest will continue to provide strong support for its customers’ digital transformation efforts through business reform using generative AI.
SOURCE: PRTimes
English