AI safety diagnosis widely used in national defense industry and public sectors
KIMM AI technology diagnosing defects in machinery ensures safety various fields
(Left) AI equipment installed in ventilation of Daejeon city subway station. (Right) Safety of Panam station ventilation system is monitored based on data collected by AI analysis equipment [Photo provided = KIMM]
Korea Institute of Machinery & Materials (KIMM President: Park Sang-jin) has invented AI diagnosis technology that is used in national defense industry and public sectors.
A Navy vessel enters the drydock and water is drained for overhaul. At the moment the AI detects a different vibration in the drainage pump and gives an early warning. It prevents possible delays in overhaul session and assures efficient management to promptly clear the deck. KIMM provided the navy logistics command a software that diagnoses their four 700kW drainage pumps.
Circulating pumps are essential equipment in water purification and sewage facilities. The technology detects abnormal vibrations as signs of defection. It makes an early warning to insure stable water supply and sewage. 6th of the last month KIMM transferred the technology to ATG and supported the application into various industries including water purification and sewage. Henceforth increase of efficiency in water management is expected.
Ventilation system in stations is one of the essential facilities in subway stations. Defects in motors a may cause harm to public health. KIMM installed IoT sensors that records vibrations and voltage and made AI learn and diagnose with the big data accumulated. Last February they applied the technology in Daejeon City hall and Gapcheon stations by collaborating with Daejon City and Daejeon Metropolitan Express Transit Corporation.
Diversifying and complexification in equipment somehow developed interest in safety. The prognostic and health (PHM) technology that KIMM has developed is based on IoT big data and AI. It prevents defects by surveilling states of equipment and giving an early warning.
AI based diagnosis technology is not easy to develop due to complexity in machinery. Facility experts must be involved from the development stage to accumulate defection data analyze machine ringing and formulate AI algorithm. KIMM developed its technology not in the labs but in the fields.
Sun Kyung-ho senior researcher of KIMM Department of System Dynamics was in charge in development of the AI based PHM technology and made efforts to industrialize the technology.
The technology learns vibration signals in pumps and motors to prevent its defections. Vibration signals are important for specifying sings of defections and with AI and deep learning technology the specification process may become faster.
The research team first began its development in 2018. They then progressively applied the technology beginning in February 2021 to Panam station in collaboration with Daejeon City and Daejeon Metropolitan Express Transit Corporation under the AI learning data construction project then in April to Navy drainage pumps and finally in June they transferred it to an equipment analysis company.
Sun expects that big data an AI technology will bring innovation in speed in safety analysis in machinery. He wishes that the technology is spread to ensure safety in buildings and plants.
The research was conducted under AI based machine system prognostics and incident response technology supported by the Ministry of Science and ICT.
Website: : http://www.wip-news.com/news/articleView.html?idxno=7759