Go to Contents
We connect all devices in your plant to provide process optimization
with manufacturing-related information in real time.
air-CMS is a tracking system jointly developed with KAIST as an industry-university collaboration project based on an existing, human-dependent prediction technology that analyzes and concludes mechanical failures. It operates on intelligent sensor modules and AI algorithms that reduce the system’s dependency on human and promise effective prediction technology for mechanical-failures.
ex) On-set date of defective parts (2018.07.09) Detection by existing system (2018.07.03) → detection by air-CMS system (2018.06.12) : Result of trial implementation at Company H: defects detected 15 days earlier than the existing system
ex) Defect detection by man → AI detection : Test result: 96.2% accuracy
Application of multivariable control-chart algorithms (MEWMA and MCUSUM, and early detection through management statistics of various clustered sensors (analysis of correction between multisensors.
As an intelligent prediction (frequency time-series) model through dynamic values from sensor data, it increases the accuracy of automatic prediction and detection of facilities.