Maximize Equipment Reliability: Expert Guide to Predictive Maintenance Tools

Predictive Maintenance vs Preventive Maintenance

Preventive maintenance and PdM maintenance are both effective maintenance strategies, but there are key differences between the two. Understanding the differences between preventive and predictive maintenance can help your team select the most suitable type of maintenance for your organization. Similarly, understanding the benefits of predictive maintenance and preventive maintenance can help you choose the right strategy. Many successful maintenance programs use a combination of both strategies.

Preventive maintenance uses the expected life cycle of an asset to determine when to perform maintenance tasks. One common preventive maintenance example is changing a car’s oil every three months or every 3,000 miles.

A preventive maintenance schedule is straightforward and sufficient for some assets. Preventive maintenance on assets may be performed based on the calendar, a certain number of hours of use, or some other usage-based metric. It could include tasks like changing filters, performing lubrication, or replacing worn parts.

Of course, preventive maintenance presents some challenges. When the calendar dictates maintenance actions, some components are replaced before they need to be. There is also some risk incurred every time a machine is worked on. Preventive maintenance can be simpler to plan, but it uses more time, money, and parts.

Predictive maintenance uses the actual operating condition of an asset to determine what steps to take and when to take them. Instead of basing maintenance on a schedule, maintenance occurs when predictive maintenance analytics identify an irregularity in the asset’s performance. While similar steps, such as lubrication or parts replacement, may be taken, the difference is that predictive maintenance actions occur exactly at the time they are needed.

A predictive maintenance strategy can save both time and money, but it poses challenges, too: chiefly, the complexity of PdM maintenance implementation. Fortunately, with the right tools, you can overcome this. While equipment is operating normally, it can be monitored by predictive maintenance technologies and condition monitoring devices, like remote sensors. They can take measurements at regular intervals or continuously.

When paired with predictive maintenance software, these sensors can alert maintenance teams when any asset’s condition changes. Automatically generated work orders via a CMMS enable teams to act quickly, preventing equipment failures.

Maintenance teams can track and analyze asset condition data to help spot patterns and make more informed decisions for future maintenance. Ultimately, the goal of PdM maintenance is to maximize asset availability and minimize the time and cost spent repairing each asset.

Predictive Maintenance Challenges

Implementing a new maintenance strategy always includes challenges, and predictive maintenance is no exception. PdM maintenance involves high upfront costs and new techniques. If you’re accustomed to a reactive maintenance approach, then transitioning to predictive maintenance will also require a fundamental shift in your whole methodology.

The benefits of predictive maintenance undoubtedly outweigh the challenges. But it’s important to be aware of those challenges before you get started, so you can prepare to face them.

Costs

Predictive maintenance relies on sensors, analytic software, and IIoT technology, all of which have a relatively high upfront cost. Integrating new technology can also be challenging, and it can be difficult to get buy-in for investing in costly predictive maintenance solutions.

Training

Your employees will require extensive training in how to implement predictive maintenance and how to use the new predictive maintenance technology correctly. They may also need time to adjust to the new maintenance approach. Employees sometimes resist the new predictive maintenance strategy, especially if they are used to a more reactive approach, but the right training program can help overcome any reluctance to PdM adoption.

Lack of High-Quality Data

Predictive maintenance software relies on huge data sets. Analytic models need historical data about your asset performance in order to create a baseline and track deviations from the norm. If your organization hasn’t been collecting asset data, this can pose some problems.

How To Overcome Predictive Maintenance Challenges

Predictive maintenance comes with some built-in challenges. The program has a relatively high upfront cost, it requires managers to oversee complex operations, and it usually calls for training maintenance teams to use new technology. You can overcome these barriers if you implement your PdM maintenance program carefully.

Overcoming Cost and Implementation Challenges

It’s a good idea to start out with a pilot program, instead of trying to convert your whole organization to a predictive maintenance approach. Piloting the system lets you keep costs low, minimizes training, and limits the operation’s administrative requirements. It’s much more affordable to buy predictive maintenance technologies in small quantities, for example — and you’ll find that they quickly pay for themselves.

A successful pilot program will deliver a significant return on investment (ROI) that can then be invested in a larger PdM program. The pilot will also help drive understanding of predictive maintenance; maintenance crews will likely get on board with the new approach when they see results.

Overcoming Training Challenges

It’s crucial to ensure your employees are fully trained in the new predictive maintenance applications. Depending on your organization, this could be difficult: in a large organization with remote staff, for example, it’s challenging to organize team training sessions. That’s why it’s a good idea to provide asynchronous and on-demand training programs, like eMaint University, which lets users fit lessons into their unique schedules. eMaint also helps track employee training and certification.

Overcoming Data Challenges

If you’re not already collecting condition monitoring data, it’s time to start. Install IIoT sensors on your critical assets to collect vibration, temperature, and other key performance data.

Sensors stream the data to your CMMS/EAM so that you don’t have to worry about data entry errors or incomplete data sets. As you collect condition monitoring data, your predictive maintenance software will build up a baseline “normal” customized for each asset, providing enough data to create an effective predictive maintenance program.

What Are the 3 Types of Predictive Maintenance?

There are several different types of predictive maintenance. The most widely used types of predictive maintenance include vibration analysis, infrared thermography, and acoustic monitoring.

Vibration Analysis

Every rotating asset vibrates while in use. However, changes to an asset’s baseline vibration pattern usually indicate a new fault. Vibration analysis monitors an asset’s vibration levels in real-time, looking for anomalies.

Changes in vibration level can indicate premature wear and corrosion; they can also point to looseness, misalignment, and bearing faults.

Today, vibration analysis is highly sophisticated. Done right, the technique lets you spot machine faults months before they grow serious enough to cause a breakdown.

Acoustic Monitoring

Acoustic monitoring lets you — or rather, your condition monitoring tools — “hear” the early indicators of friction or wear and tear. Rotating equipment emits characteristic sounds as it deteriorates. Sometimes, those sounds are loud enough to hear with your naked ear, but acoustic monitoring catches much fainter sounds you can’t pick up, making it an excellent predictive tool.

Acoustic monitoring is widely used as a leak prevention tool, especially in systems with extensive pipelines for gas, oil, or liquids.

Infrared Cameras

Infrared cameras can detect subtle changes in temperature that may point to emerging machine faults.

Increases in temperature often result from high levels of friction, premature wear, or deterioration. Faulty wiring or other electrical issues are another possible root cause. Infrared thermography can also assist with locating gas or liquid leaks; it can spot changes in temperature caused by moisture or gas.

Of course, there are many other approaches to predictive maintenance. If you use a CMMS to anchor your predictive maintenance program, you’ll be able to integrate all of these different types of insights into one highly effective PdM model. 

Predictive Maintenance Techniques

There are many ways to implement a predictive maintenance strategy, and many available predictive maintenance technologies. The following predictive maintenance tools and techniques give each organization the power to gather as much or as little information as they need to implement and maintain their predictive maintenance program.

• Vibration monitoring: Sensors installed on equipment can monitor in-depth vibration readings. Once the baseline for the asset is established, these sensors can be continuously monitored to detect deviations that could indicate faults like imbalances, misalignments, or bearing faults.
• Temperature monitoring: Similar to vibration monitoring, sensors can detect when temperatures rise above the asset’s normal temperatures. When a temperature increase is detected, technicians can find and address the root cause before failure occurs.
• Condition monitoring: Using a cloud-based CMMS stores sensor data in the cloud, where it can be monitored and analyzed from anywhere. Even if equipment is in a remote location or monitoring needs to occur off-site, users can access current or historical data and use it to make decisions about maintenance and replacement.
• Artificial intelligence (AI) analysis and recommendations: Learning how to read the signatures provided by vibration sensors takes years of education and experience. Now, even if your organization doesn’t have an expert on-site, advanced AI-powered analytics can assess machine vibration patterns and identify changes. It can even recognize different patterns of common issues, giving your team the insight to find and fix the problem even faster.
• Alarms: When vibration levels indicate faults, predictive maintenance software can send alerts to the appropriate personnel so they can take immediate action.
• Automated work orders: If the vibration monitoring software is integrated with a computerized maintenance management system, the CMMS can automatically trigger a work order when a fault is detected, saving time and reducing the amount of human intervention needed to fix the problem.

Predictive Maintenance Examples

There are important predictive maintenance applications in almost every industry. Here are just a few typical predictive maintenance examples.

Predictive Maintenance Examples in Automotive

Predictive maintenance tools can identify impending failures, such as a slowing conveyor belt or abnormalities in vibrations from stamping or press machines. It can also be used on other assets, like forklifts and painting equipment.

Predictive Maintenance Examples in Food and Beverage

In the food and beverage industry, predictive maintenance technologies can play a role in not only ensuring maximum uptime but also ensuring all products are created in compliance with strict food regulations. Predictive maintenance can be used on equipment like mixers and blenders, dust collection systems, extrusion equipment, pumps, and conveyor belts.

Predictive Maintenance Examples in Manufacturing

Manufacturers of all types can use predictive maintenance technology to improve the consistency and quality of their product output, reduce labor costs, and prolong the lifespan of assets. Predictive maintenance in manufacturing can help predict and reduce failures for assets like fans, pumps, and motors.

Predictive Maintenance Examples in Life Sciences

Many manufacturers in the life sciences industry are subject to audits from local, state, and federal authorities. Predictive maintenance technology can ensure equipment continues running within required parameters and provide organizations with audit-proof records of asset history. And in cases where products need to be refrigerated or frozen, sensors help ensure that the equipment used to keep them at the proper temperature is always working as intended.

Predictive Maintenance Examples in Oil and Gas

Reliability is incredibly important in the oil and gas industry, where equipment failures could have environmental consequences and pose safety threats to employees. Predictive maintenance on assets like pumps, boilers, and compressors can help reduce the risks of unplanned failure and its consequences.

How To Create a PdM Maintenance Program

Making the switch from reactive to predictive maintenance doesn’t happen overnight. But advances in predictive maintenance technologies, such as CMMS software and wireless vibration sensors, have made predictive maintenance a more attainable strategy than ever before. There are a few questions to keep in mind for each asset when considering creating a predictive maintenance plan:

• If this asset fails, how does it impact production?
• How much does it cost to repair this asset?
• How much does it cost to replace this asset?

Answering these questions for each piece of equipment can help teams narrow down which assets to maintain on a predictive basis.

Predictive maintenance is not necessarily the most effective strategy for every asset. Some assets can be run to failure with little to no impact on production or the bottom line. Others benefit from simple and straightforward preventive maintenance. But for some assets, predictive maintenance is the best strategy.

Even if you plan to use predictive maintenance tools on just a handful of assets, it helps to plan ahead and build a program that your maintenance team can stick to. Here are six key steps for setting up your predictive maintenance program:

1. Identify which assets should be targeted for predictive maintenance solutions
2. Choose the predictive maintenance tools and methods you will use to monitor asset condition (such as sensors and a CMMS)
3. Select and train an implementation team to learn and carry out predictive maintenance technologies
4. Perform system integrations to get a complete picture of asset health
5. Coordinate your overall maintenance strategy, identifying which approach will be used and where
6. Determine how asset health data will be shared among team members, stakeholders, and auditors

Ultimately, implementing a successful predictive maintenance program requires taking a long-term view of your organization’s goals and needs. No two predictive maintenance plans will look the same.

How Can You Control Predictive Maintenance?

Predictive maintenance solutions, by definition, involve collecting and analyzing a lot of data. The best way to control predictive maintenance is by using a computerized maintenance management system (CMMS) to connect and manage data coming in from work orders, real-time predictive maintenance analytics, and maintenance history, making it accessible to the appropriate personnel no matter where or when they’re working.

Without a CMMS, maintenance teams are often left guessing about the historical maintenance of an asset. Work orders are often on paper, and paper work orders take time to find, complete, and file away. Paper work orders also make it difficult to track what’s completed or still outstanding. It’s nearly impossible to compare the full range of requests, in-progress tasks, and priority jobs when they’re all on separate sheets of paper.

A CMMS makes work orders so much easier to schedule, assign, and complete. Work orders can also be prioritized based on asset criticality, ensuring the most important tasks get assigned to the right technicians. Managers can see which tasks are outstanding and assign jobs to staff already working on a specific asset or those with the expertise needed for the task.

Technicians and decision-makers will also have access to historical maintenance records. When an asset has a history of multiple failures in a short time frame, experts can use the data and predictive maintenance analytics to get to the root cause of the issue or decide if it’s time to replace the asset.

Key Features in eMaint Predictive Maintenance Software

eMaint CMMS gives organizations a full suite of predictive maintenance tools. With it, organizations can:

• Define monitoring classes for each asset
• Monitor noise, vibration, temperature, lubricants, wear, corrosion, pressure, and flow independently
• Enter manually or import meter readings
• Define upper and lower boundaries of acceptable operation for each asset
• Display readings as a report with color-coded exceptions
• Auto-trigger emails when a boundary is exceeded
• Auto-generate work orders when a reading falls outside of predefined boundaries
• Perform data analysis to identify failures early, prevent breakdowns, and optimize maintenance resources
• View condition monitoring diagram

Case Study: Using eMaint CMMS Condition Monitoring for Predictive Maintenance

Cleveland Tubing, Inc. is a manufacturer of flexible, collapsible tubing products, including FLEX-Drain and PumpFlex. The company set up eMaint so that meter readings on key indicators (temperature, pressure, fluid levels, suction) are imported and used to trigger priority work orders when work or inspection is needed based on predefined ranges.
Gary Payne, maintenance manager for Cleveland Tubing, noted that eMaint has become their maintenance decision support system, informing them of the tasks that need to be performed each day, based on elapsed time, equipment utilization, and condition-based indicators. They also experienced:

• Automated reports for replenishing inventory on stocked and non-stocked parts
• Streamlined time tracking of labor for a department of five maintenance employees
• Improved ROI calculations with better allocation of labor and material costs to assets
• The ability to evolve from reactive maintenance to planned maintenance to predictive maintenance via condition monitoring and automated alerts of potential problems on critical equipment
• Easily measure and track KPIs against world-class standards (90% planned maintenance)

What Is the Future of Predictive Maintenance?

The future of predictive maintenance draws on artificial intelligence tools to deliver insights at a greater scale than ever before. AI tools scan vast data sets at a high speed, which is invaluable for large organizations with fleets of critical assets. The best AI tools can diagnose machine faults and determine fault severity levels, helping set clear maintenance priorities.

That doesn’t mean that maintenance will be automated, though. Human technicians and operators need to oversee AI output and build on its insights. AI is a useful tool, rather than a complete solution. It works best when it shares workflows with human employees.