On the 6th, Deputy Prime Minister and Minister of Science and ICT, Lee Kwang-hoon, visited Shinseong Delta Tech in Changwon, Gyeongnam, where he checked on the performance of the physical AI pre-validation project. This project aims to enhance the practical applicability of AI by enabling it to understand the laws of the physical world. Last year, a consortium involving Gyeongnam University, Seoul National University, and eight local manufacturing firms participated in the pre-validation project to assess the application of physical AI in manufacturing processes.

The government is launching the 'Gyeongnam AI Transformation (AX)' initiative this year, which seeks to move beyond simple automation of processes to incorporate the physical characteristics of the workplace and the expertise of skilled workers directly into AI models. Unlike traditional AI that focuses on analyzing and making decisions on screens, the goal of physical AI is to enable systems to understand and actually operate the manufacturing processes. Ultimately, the initiative aims to drive technological innovation within South Korea's manufacturing sector through the application of AI.

The validation results indicate significant improvements among participating firms. Shinseong Delta Tech, a home appliance parts manufacturer, reduced its defect rate by 15% and improved equipment utilization by about 20%. Similarly, Huaseung R&A, an automotive parts company, was able to predict material deformation in its rubber extrusion process, leading to over a 5% improvement in overall equipment efficiency. CTR, another automotive parts manufacturer, reduced defect rates and processing time by over 17% by predicting machine chatter during aluminum processing. These outcomes highlight the potential of physical AI to transform manufacturing processes in South Korea.