On December 14, UNIST announced the release of five types of artificial intelligence (AI) datasets, specialized in manufacturing via the Korea AI Manufacturing Platform (KAMP) of the Korean Ministry of SMEs and Startups (MSS).

These datasets are particularly concentrated in root industries that are related to the infrastructure of manufacturing businesses, such as CNC machining, injection molding machines, welding machines, machine vision, and presses.

UNIST expects that these datasets will also be useful for the establishment of smart factories in small and medium-sized manufacturing enterprises, as well as for the application of AI during the process. The accompanying guidebook includes an introduction to each process, as well as the AI algorithms that can be used for the analysis of various segments and sub-segments of AI in manufacturing industries.

Images of industrial sites at a local manufacturing company that currently carries out big data analysis of manufacturing processes, in partnership with UNIST. l Image Credit: Department of Industrial Engineering, UNIST

A total of 12 datasets have been released via KAMP on December 14, 2020. The KAIST K-Industry 4.0 Promotion Headquarters, an organization under KAIST, took charge of the operation, and thus created two datasets. UNIST Consortium and Impix Consortium took charge of five datasets, respectively.

Professor Sunghoon Lim in the Department of Industrial Engineering within the College of Information-Bio Convergence Engineering at UNIST was the lead researcher for the project. It has been also participated by Professor Seungjoon Yang (Department of Electrical Engineering, UNIST) and Professor Taeseok Jeong (UNIST Institute for the 4th Industrial Revolution). In addition, InterX Co., Ltd., which partook in the UNIST Consortium as a participating organization, also carried out data collection and pre-processing for each of the processes.