The Korea Institute of Machinery and Materials (KIMM) held an artificial intelligence (AI) competition utilizing real-world industrial data and set out to spread machine data-based physical AI technology. Participants verified the performance and applicability of AI models based on the theme of predicting bearing remaining lifespan.

The Korea Institute of Machinery and Materials announced on the 26th that it held the '2026 KSPHM-KIMM Machine Data Challenge' and conducted the final presentation evaluation and awards ceremony at the Westin Chosun Hotel in Busan on the 25th.

This competition is a data contest designed to validate AI-based Prognostics and Health Management (PHM) technology. It was conducted by evaluating the potential for industrial application using actual machine data held by the Korea Institute of Machinery and Materials (KIMM).

A total of 80 teams and 259 participants took part in this year's competition. In addition to 205 participants from 64 university teams, participation also extended to industry and research institutions. Major participating organizations included Ajou University, Sungkyunkwan University, the Electronics and Telecommunications Research Institute, and Hyundai Motor Company.

The theme of the competition is 'Predicting Remaining Life Using Bearing Degradation Data'. Participants developed an algorithm to predict the point of failure based on data collected under variable operating conditions.

Bearings are core components of rotating machinery, and failure can lead to equipment shutdown and safety issues. Particularly in industrial settings, where operating conditions are inconsistent and difficult to predict, this technology is presented as a core technology applicable to the fields of autonomous manufacturing and smart maintenance.

As a result of the final judging, the Grand Prize was awarded to the Electronics and Telecommunications Research Institute (EHEI) team. The Top Excellence Award went to the Dongguk University team (BRIDGE), while the Excellence Awards were given to the Korea Aerospace University and Seoul National University teams. Encouragement Awards were presented to teams from the State University of New York Korea, Ajou University, and Seoul National University, among others.

The winning teams cited the ability to validate AI models using actual industrial data as a key feature. They also suggested that prediction methods combining domain knowledge and AI could be effective for industrial applications.

Ryu Seok-hyun, President of the Korea Institute of Machinery and Materials, stated, “AI-based predictive diagnostic technology is key to the realization of autonomous manufacturing and physical AI,” adding, “We plan to expand the data utilization ecosystem centered on the machine data platform.”

The Korea Institute of Machinery and Materials plans to develop this competition into a demonstration case of machine data-based AI technology and continue to promote the expansion of the related industrial ecosystem.