Miniterm, a Novel Virtual Sensor for Predictive Maintenance for the Industry 4.0 Era.

Area of Science:

• Industrial Engineering
• Manufacturing Systems
• Data Science

Background:

• Industry 4.0 faces challenges with Big Data and the high cost of implementing traditional sensors for predictive maintenance.
• Deterioration in machine components affects mini-term (a sub-cycle time), making it a potential indicator for failures and production measurement.

Purpose of the Study:

• To introduce a novel virtual sensor, the mini-term, for cost-effective predictive maintenance.
• To demonstrate the Miniterm 4.0 system's capability to generate industrial Big Data for predictive maintenance.
• To validate the improvements in factory operations through the implementation of Miniterm 4.0.

Main Methods:

• Implementation of a timer-based virtual sensor (mini-term) in existing PLC or PC systems.
• Collection and analysis of data related to Technical Availability (TAV), Mean Time To Repair (MTTR), Number of Work Orders (EM), and Labour Hours in EM.
• Comparison of operational metrics before and after Miniterm 4.0 implementation.

Main Results:

• Technical Availability (TAV) increased by 15%.
• Labour Hours in Emergency Orders (OM) reduced by 5000 orders.
• Mean Time To Repair (MTTR) decreased by 2 hours.
• 3000 fewer emergency orders were generated post-implementation.

Conclusions:

• The Miniterm 4.0 system provides a scalable and economical approach to predictive maintenance.
• The mini-term serves as a multifunctional indicator for machine health and production efficiency.
• Significant operational improvements were achieved, highlighting the system's practical benefits.