Deploying Predictive Maintenance for the Maximum Benefit in Industry 4.0 Facilities

By Jeff Shepard

Contributed By DigiKey's North American Editors

2025-08-14

Predictive maintenance (PdM) that leverages data analytics and machine learning (ML) is crucial in Industry 4.0 because it enables proactive equipment management, optimizing efficiency, plus maintenance scheduling, and minimizing downtime to support increased sustainability. Timely and accurate data collection is a key to the successful deployment of PdM.

The data also needs to be complete. Power supplies can monitor DC voltage (V_DC) and current (I_DC), peak current (I_PEAK), run time, and replacement time. Motor condition monitors are needed for vibration, temperature, current, and insulation resistance (ground fault).

Thermal condition monitors are needed for an array of devices like high voltage control panels, power transformers, hydraulic equipment, motors and bearings, gearboxes, and more. All those power supplies, motors, and thermal monitors need Ethernet/IP or Modbus TCP connectivity to send the data for real-time analysis.

This article begins with a brief overview of PdM, its numerous benefits, and how it can fit into Industry 4.0 system architectures. It then takes a deep dive into the numerous PdM devices and software offered by Omron. It concludes by looking at how to use artificial intelligence (AI) to optimize PdM performance.

PdM is one of three approaches to equipment and facility maintenance. In terms of balancing environmental and business costs, it lies between reactive and preventative maintenance (Figure 1). One factor in choosing between the three approaches is the relative importance placed on environmental and business costs.

Diagram of PdM lies between reactive and preventative maintenance Figure 1: PdM lies between reactive and preventative maintenance approaches and provides a balance between business and environmental considerations. (Image source: Omron)

Reactive maintenance waits for failures to occur, then deals with them, increasing both the environmental and business costs. Preventative maintenance prioritizes minimal environmental costs by using regular manual inspections to identify impending failures, but can result in excessive equipment downtime and high business expenses. It’s considered another form of reactive maintenance where the driving factor is a predetermined, and somewhat arbitrary, schedule, not outright equipment failure.

The availability of advanced sensors and the emergence of AI and ML tools has enabled the development of PdM that deploys technology to balance environmental and business costs.

Scalable and flexible

PdM is not a “one size fits all” option. It’s scalable and flexible and can be deployed for a single enterprise-critical piece of equipment, for multiple pieces of equipment, or across an entire facility using centralized monitoring. That enables organizations to begin small and expand the deployment of PdM over time, minimizing disruption when retrofitting existing facilities.

Scalable solutions are supported with a variety of compatible components like sensors, monitoring units, and controllers that can be added as needed. The use of industrial communication protocols like Ethernet/IP and Modbus TCP simplifies integration with existing systems and supports enhanced functions like remote monitoring of multiple devices simultaneously.

Scalable software solutions are available for analyzing data and managing devices from a centralized office control center or various locations across the facility floor.

These solutions can be integrated with existing equipment without major overhauls, enhancing flexibility. They can be optimized for almost every industry, including food and beverage, automotive, medical device manufacturing, semiconductors and electronics, military and aerospace, logistics and warehousing, and more.

That flexibility is supported by a wide range of PdM devices, including solutions for monitoring power supply, motor condition (current, vibration, temperature, insulation resistance), and thermal conditions. In addition, standard software functional blocks (FBs) are available for data acquisition, communications, data processing and analysis, setting alarms and sending notifications, data logging and reporting, and for implementing customized AI- and ML-based PdM analysis.

Condition monitors replace simple sensors

A key differentiator between PdM and other approaches is the use of condition monitors instead of simple sensors to track equipment performance and enable proactive maintenance. Like sensors, condition monitors are collocated with the piece of equipment they are monitoring.

However, while sensors can be deployed with relatively simple protocols like IO-Link, condition monitors require more sophisticated connectivity like EtherNet/IP or Modbus TCP. The condition monitors can perform local data processing and often include status displays not typically associated with sensors.

The condition monitors can be linked through one or more communication hubs to higher-level devices like human machine interfaces (HMIs) that can provide a centralized location for data visualization or to programmable logic controllers (PLCs) or centralized monitoring systems with more comprehensive data analysis tools, including AI and ML (Figure 2).

Figure 2: Omron’s PdM suite of solutions can be deployed individually to monitor critical assets, starting small, then grow incrementally to reach across entire manufacturing or logistics sites for comprehensive solutions. (Image source: Omron)

Diving deeper

Omron offers a selection of PdM devices and software. For example, the S8VK-X Ethernet Connected Smart Power Supplies measure numerous aspects of performance, including Vout and Iout for monitoring energy consumption, and I_PEAK to identify potential overload conditions.

These power supplies measure actual operating hours. They also estimate the remaining lifespan of the electrolytic capacitors using the Arrhenius equation, which states the lifetime of a capacitor is roughly halved for every 10°C increase in temperature, combined with actual operating temperatures, and displays the results as years or as a percentage of lifetime remaining.

S8VK-X power supplies are available rated from 30 W to 480 W and output voltages of 5 V_DC, 12 V_DC, and 24 V_DC. They are also offered with an integrated monitor display like the model S8VK-X48024A-EIP rated for 24 V_DC and 480 W, or without an integrated display like model S8VK-X03005-EIP rated for 5 V_DC and 30 W.

Electric motor condition monitoring is an important aspect of PdM, and Omron’s K6CM Motor Maintenance Monitors are suited for all types of water pumps, plus motors in heating ventilation and air conditioning (HVAC), agriculture, escalators, and most other electric motor applications.

Motor maintenance monitors are available for vibration and temperature monitoring, insulation resistance monitoring, and motor current monitoring. Models are available for 100 to 240 three-phase V_AC or 24 V_AC/V_DC input power.

Vibration and temperature can be monitored using the K6CM-VBMD-EIP that operates from 24 V_AC/V_DC. All the temperature monitors work with the K6CM-VBS1 vibration and temperature sensor, which consists of a sensor head located on the motor and a pre-amplifier that connects between the sensor and the monitor.

Insulation resistance health can be monitored using the K6CM-ISMD-EIP that operates from 24 V_AC/V_DC together with the K6CM-ISZBI52 zero current transformer (ZCT) and insulation resistance transfer (IRT) sensor. The ZCT function measures the leakage current in a three-phase motor circuit, while the IRT function measures the insulation resistance between the motor windings and ground.

Motor condition of three-phase induction motors can also be monitored using the K6CM-CIMA-EIP that operates from 100 V_AC to 240 V_AC, together with the K6CM-CICB400 current sensor rated for 400 A. Other models of current sensors are available from 5 A to 600 A.

These monitors use Omron’s complete current diagnosis technology. They can detect abnormalities like cavitation or air contamination by quantifying the deviation between an ideal sine wave and the measured current waveform. Conditions like misalignment, load imbalance, or foreign matter adhesion are quantified by analyzing the frequency components of the measured current waveform.

The K6PM Thermal Condition Monitor system can be used to implement PdM for a variety of industrial equipment, such as high voltage control panels, transformers, hydraulic equipment, data centers, bearings, gearboxes, and so on. It includes the K6PM-THS3232 thermal image controller and the K6PM-THMD-EIP thermal imaging infrared (IR) sensor that can monitor temperatures from 0°C to +200°C.

A single K6PM thermal condition controller can monitor up to 31 IR sensors. The free PC monitoring software includes abnormal temperature detection algorithms and three-level temperature alarms. The software also supports user-defined alarm thresholds.

Figure 3: Core functions of Omron’s PdM offering include smart power supplies, plus motor and thermal condition monitors and related sensors. (Image source: Omron)

Adding PdM AI on the edge

The AI PdM Library from Omron is one of the Sysmac Library software function components for using the AI capabilities of the NX/NY-series artificial intelligence machine automation controllers (AI controllers). The AI PdM Library includes FBs for each mechanism (devices and components like a cylinder, ball screw, or belt pulley).

Users can create and integrate custom FBs in the form of reusable code blocks to implement customized PdM functions. Custom FBs can be used for:

Development of application-specific algorithms
Interfacing with a wider range of sensors or other equipment for data acquisition
Customized data processing that aligns with specific PdM implementation strategies

FBs generate variables that are used as inputs for the AI engine in the NX/NY controllers. The AI engine analyzes the data collected in a time series database for abnormal patterns or behaviors in the equipment. The complete AI engine is designed to operate autonomously within an AI controller.

AI Controllers for PdM

Omron’s NX/NY-series AI controllers include the NX701-Z700, which can implement PdM AI without connecting to the cloud. The NX701 can control up to 256 axes of motion in a single program, lowering installation costs and simplifying program design, verification, and revision control.

It includes integrated Ethernet/IP and EtherCAT communication ports. That gives industrial control network designers the option of leveraging EtherNet/IP for larger packet data size and EtherCAT for guaranteed packet delivery to support deterministic motion. The 80 MB program memory can support numerous FBs as well as Omron’s AI PdM Library (Figure 4).

Image of Omron NX701-Z700 can implement PdM AI Figure 4: The NX701-Z700 can implement PdM AI without connecting to the cloud, and it can control up to 256 axes of motion in a single program. (Image source: Omron)

Additional features of the NX701-Z700 AI controller include:

Plug and play integration with 120+ NX input/output (I/O) units
Comprehensive software environment including Sysmac Studio, Omron Vision, Omron Motion, Omron Robotics, and Omron Safety Components
Support for multiple industrial communication protocols, including EtherNet/IP, EtherCAT, Fail Safe Over EtherCAT, IO-Link, and Factory Scale Motion
Guaranteed EtherCAT cycle times from 0.125 ms to 0.250 ms in 0.125 ms increments and from 0.250 ms to 8 ms in 0.250 ms increments

Conclusion

PdM is a new paradigm for optimizing the balance between environmental and business considerations in Industry 4.0 factories and logistics operations. It leverages advanced sensors, time series databases, and PLCs with edge AI to continuously monitor the performance and health of equipment, predict potential failures, and proactively schedule maintenance.

Regular monitoring and maintenance based on PdM insights help identify and address potential problems early on, preventing premature wear and tear and extending the lifespan of assets. PdM helps maximize equipment uptime, reduce waste, and improve operational efficiency. Finally, PdM is highly flexible and scalable, and an organization can roll out PdM across a facility as quickly or gradually as desired and still realize significant performance improvements and cost savings.