"AI Predicts Equipment Failures in Advance" LGU+ Launches 'U+ Smart Factory' Solution

[Asia Economy Reporter Eunmo Koo] LG Uplus announced on the 29th that it has launched a new smart factory solution that diagnoses and inspects factory equipment failures and malfunctions in advance, in collaboration with OnePredict, an industrial artificial intelligence (AI) specialist company.

The newly established solution is a service that performs predictive maintenance of various equipment using AI. It analyzes diverse data such as vibration, current, and speed generated by the equipment through AI technology to diagnose the condition and predict the timing of failure and remaining lifespan. For example, in the case of pumps or motors, appropriate Internet of Things (IoT) sensors are first attached to the equipment. These sensors collect and transmit vibration data from the rotating component bearings inside. The server compares and analyzes this data with previously collected vibration values and data from other bearings to predict the current condition and failure timing. Through this, operators can inspect the bearing’s inner ring, outer ring, balls, etc., in detail and prevent accidents.

In this process, OnePredict provides industrial AI-based predictive maintenance algorithms and products, while LG Uplus is responsible for data transmission via wireless communication and customer-facing monitoring platform services. Especially at the monitoring stage, operators can easily diagnose equipment status through smartphones or PCs, making it efficient for large-scale factories with many monitoring targets.

Another advantage is the ability to introduce various measurement methods depending on the equipment’s condition and type. For instance, motors measure current values, transformers analyze dissolved gases in insulating oil, and speed reduction rotating equipment reads vibration and speed data. This data is accumulated to build big data, and as the amount of measured data increases, predictive maintenance evolves more precisely through deep learning. As a result, the range of industries where the solution can be applied is gradually expanding.

A representative from a company that piloted the AI predictive maintenance solution explained, "According to company policy, we used to perform an overhaul by completely disassembling, inspecting, repairing, and adjusting machinery every 20,000 hours of compressor operation, but now we perform overhauls based on equipment condition rather than operating hours, which has greatly improved productivity."

Jae-yong Seo, Executive Director of Smart Infrastructure Business at LG Uplus, emphasized, "This solution is expected to have high applicability in various industrial fields including bearings, pumps, robot fans, extruders, turbines, wind power generators, and transformers. We will continue to expand the scope of predictive maintenance solutions going forward."