The rise of smart devices ushered in by Industry 4.0 has led to an increase in the practice of predictive maintenance (PdM). PdM is a form of preventive maintenance (PvM) that incorporates methods such as sensors and smart devices to constantly monitor performance parameters.

Changes in conditions such as temperature, flow rate, loading, position or vibration can signal a maintenance issue that, if unaddressed, can lead to an unexpected and costly failure. For example, a bearing that is starting to fail could easily be detected with an accelerometer to measure vibration levels. Lubrication issues may present themselves as an increase in temperature due to excess friction. Changes in flow rates can indicate a blockage. Any of these issues can lead to serious problems.

Traditional PvM programs could miss such signs until it is too late. To prevent premature failures, many manufacturers have developed PdM plans to routinely inspect equipment at predetermined machine points. Maintenance personnel check and log these points to ensure equipment is operating within desired specifications and look for changes that could signal a developing issue. Data loggers record this information, which can be evaluated for changes or trends that might indicate a problem. While this can be an effective maintenance plan, it also requires dedicated additional manpower and adherence to specified maintenance schedules. It also requires human interpretation of the data, which can be subjective.

Figure 1: Smart devices and sensors could revolutionize predictive maintenance. Source: U.S. Air Force photo by Senior Airman Kelly O’ConnorFigure 1: Smart devices and sensors could revolutionize predictive maintenance. Source: U.S. Air Force photo by Senior Airman Kelly O’Connor

Today, OEMs have a new PdM solution by incorporating smart devices into their equipment. Although sensors have been used for PdM applications for years, smart devices embed these sensors in machine elements and communicate the data in real time through the industrial internet of things (IIoT) to a central data storage location. Machine learning and predictive analytics software can interpret this data and notify maintenance and production personnel of potential issues. This enables remote monitoring of equipment located anywhere in the plant. The addition of mobile apps allows personnel to monitor conditions even if they are in a remote location.

Smart devices can also be retrofitted to existing equipment. Modern industrial equipment is manufactured for years of operation. While retrofitting does come with a cost, it may be far less expensive than replacing equipment that is otherwise in good working condition. Many smart devices are designed to replace existing components with little to no required modifications, minimizing the implementation costs. Others, such as ABB’s Ability Smart Sensor can be incorporated into traditional motors, pumps and mounted bearings to convert them to wireless smart devices.

Any retrofitting program should be planned carefully to select when and what equipment to address. Other factors to consider are how much downtime is required and how feasible and expensive the retrofit is to execute.

When it comes to cloud storage systems that integrate with the IIoT, users have many options to choose from, including:

· Amazon AWS IoT

· Bosch IoT Cloud

· Cisco IoT Cloud Connect

· GE Predix Platform

· Google Cloud IoT

· IBM Watson IoT

· Microsoft Azure IoT Hub

Some smart products send information to proprietary cloud services, such as SKF’s Insight bearings. SKF offers users remote diagnostic services for condition monitoring, data analysis and report generation.

Cost savings

The primary benefit of integrating smart devices to a PdM program is cost savings realized through increased machine uptime and an increase in machine efficiency and productivity.

IBM states that savings up to 15% in unplanned downtime can be obtained by capturing equipment data, and Agosto claims that costs can be reduced by up to 25% by implementing an effective PdM program. A 2015 McKinsey report puts the global economic impact for factories using IIoT for operations, management and PdM at $1.2 to $3.7 trillion by 2025.

The savings do not stop with incorporation of smart devices. Machine learning and data analytics can continue to improve the return on investment as data is collected over time.

Smart devices and the IIoT are set to revolutionize industrial PdM programs for years to come.