As more factories become smart, the importance of preventive maintenance becomes even more important. Preventive maintenance realizes “production without stoppage” by detecting causes of failures ahead of time. This is done by monitoring the conditions of production equipment. Many companies, including distributors of factory automation (FA) equipment and electronics components, are proposing various preventive maintenance methods. They include analysis of sensing data, such as vibrations and sound of the motor; analysis of vibrations of machine element components, such as linear motion guides; and application of acoustic emissions (AE).

Fuji Electric Co., Ltd. has employed a programmable logic controller (PLC) in its solution, as preventive maintenance among many other methods. The company has developed and started market offering of an abnormality diagnostic solution that detects and analyzes abnormalities in the processing of products in the production process.

PLC for Preventive Maintenance

A PLC incorporates signals from input equipment, and controls connected output equipment through sequence control performed by various types of processing in accordance with the program. PLCs are used primarily for control of equipment in the manufacturing industry. 

It is crucial to discover abnormality early in the production process and take measures before defective products are shipped. In this way, customers’ production lines can maintain and improve quality, and defective products are prevented from entering the market.

To this end, infrared sensors and pressure sensors are generally used, and further improvement in the abnormality detection accuracy is required to prevent an outflow of defective products.

Abnormality Diagnostic Solution

Fuji Electric has developed a diagnostic module that can be incorporated in the PLC to detect abnormality in the product processing and analyze causes. The company provides the diagnostic module as an abnormality diagnostic solution combined with a servo system and a programable display unit.

The abnormality diagnostic solution uses the servo system (servo motor), which controls the operation of production equipment, as a sensor. Abnormality diagnosis is performed by analytic artificial intelligence (AI) multivariate statistical process control (MSPC), which compares operation data at normal processing of equipment (diagnostic model) and data during operation and detects discrepancy as an abnormality. MSPC features higher accuracy in abnormality detection compared with conventional sensors and contributes to the reduction of equipment cost as it eliminates sensors.

Diagnostic results accumulated in the diagnostic module are displayed on the programmable display unit. Results or waveforms to be checked can be selected from the history, thus supporting clarification of the cause of abnormality.

For example, in the automatic packaging process of foods, the bite of waste food at the adhesive part of the enclosing port can inhibit tight sealing and cause adverse effects, such as taste deterioration of the food and contamination by foreign substance. In this process, the enclosing port is adhered, and products are cut into individual packages. Here, a servo system is used to control adhering and cutting operations. When waste food is bitten at the adhering section, MSPC finds discrepancies in load torque on the servo motor and other parameters compared with the normal processing and detects it as an abnormality. Main applications of the solution include food packaging machines, semiconductor manufacturing equipment, and metal processing machines.