Intelligent maintenance with servo systems

In 1987, Mitsubishi Electric introduced its first fully digital servo amplifier. Since then, the company has been collecting operational data with and about its servos and their peripherals. This information forms the basis for an in-depth understanding of all the mechanical systems driven by its motors. 

Using this comprehensive data store, Mitsubishi Electric developed an advanced diagnostic tool for predictive maintenance of its latest MELSERVO MR-J5 series servo motors and amplifiers. This solution uses the company's proprietary deep-learning AI technology Maisart® (Mitsubishi Electric's AI creates the State-of-the-ART in Technology), for example, to detect wear on mechanical components before maintenance is required.

This predictive maintenance concept is based on 'Deep Reinforcement Learning'. This is an AI application that automatically processes data and thus learns to recognise patterns and anomalies on its own. Users can thus implement an intelligent setup in a very short time without having to have advanced knowledge in programming or machine learning.

Above all, however, the knowledge is oriented towards the individual system in which the servos are used. Different applications all have their own optimal system condition. The AI determines the ideal operating parameters and  conditions, as well as any behaviour that should be considered an anomaly. 

These capabilities are complemented by CC-Link IE TSN network technology with gigabit bandwidth and Time-Sensitive Networking (TSN) functionalities. This enables the servo drives to transmit large amounts of data for time-critical control tasks as well as less volatile AI analysis information promptly and without delays.