Preventive vs. predictive maintenance: Which is right for you?
Ken Thayer | May 24, 2019Source: U.S. Air Force Medical ServiceReplacing smoke detector batteries twice a year is a common practice for homeowners.
This sound advice is an example of preventive maintenance (PvM). But is this necessarily the best practice today?
When the first “fire detectors” for home use were introduced in 1955, zinc-carbon batteries were still popular. Plagued by poor battery life, zinc-carbon batteries were replaced by alkaline batteries, which remained the standard until the 1980s. Although alkaline batteries are still widely available and used, lithium batteries have gained in popularity due to their significant increase in shelf and usable life. Testing by The Gadgeteer displayed that lithium batteries had a usable life up to twice that of alkaline, while MyBroadband testing yielded life spans six to seven times longer for lithium.
Changing smoke detector batteries twice a year may have been necessary with zinc-carbon, or even alkaline batteries, but some property owners may be able to rethink that routine if they have adopted the latest battery technology. This is illustrative of how a poorly structured preventive maintenance plan can cost more money and yield higher productivity losses than necessary.
Now consider this alternative: Modern smoke detectors come equipped with a low battery warning. Rather than change the batteries at a regular interval, residents could wait for the smoke detector to chirp, telling them it’s time for replacement. This is an example of predictive maintenance (PdM). The smoke detector is communicating that, although it is still functioning normally, it is time to replace the batteries before they fail.
(The author does not recommend risking the life of your family or your home to save a few dollars on a battery — this is merely an example or preventive vs. predictive maintenance.)
At a larger scale, industrial preventive and predictive maintenance programs offer different strategies to keep equipment in good operating condition and prevent unexpected failures. The overarching goals of each approach are the same: keep equipment uptime as high as possible and prevent costly shutdowns.
A third maintenance strategy is referred to as breakdown or reactive maintenance. As it suggests, this approach waits until something fails or breaks before performing maintenance on it. Although it is possible to run significantly longer before a breakdown, the costs associated with repairs and downtime can also be much greater. While both preventive and predictive maintenance programs are almost always superior to waiting and reacting to failure, there are pros and cons to each.
[Learn more about Condition Monitoring and Machine Maintenance Services name="_Hlk2183126"> on IEEE GlobalSpec.]
Advantages of a good maintenance program
Source: CC BY-SA 3.0Preventive and predictive maintenance share many of the same advantages.
Cost savings. Perhaps the biggest benefit of running a maintenance program is the savings due to unexpected shutdowns caused by equipment breakdowns. Many times, a significant failure can be prevented by routine maintenance. Failure to replace a bearing or seal can lead to much more significant problems, such as damage to a shaft that results in extensive downtime as well as maintenance and replacement part costs.
Equipment life and efficiency. A good maintenance program will increase equipment life by ensuring that all critical operating components are kept in optimal working condition. Machines that run like new also tend to be more efficient, produce better quality and consume less energy.
Safety. Safety to personnel is an often overlooked benefit to preventive maintenance. Catastrophic machine failures due to lack of maintenance raises the risk of employees getting hurt in an accident.
Preventive maintenance
Industrial preventive maintenance programs typically involve a planned shutdown on a regular interval to perform maintenance on equipment. The interval is determined by experience or by the manufacturer’s recommendations, and could be based on time or some other measurable attribute, such as the number of parts run. Some preventive maintenance items can and should be performed without a stoppage in production. Lubrication is a good example of a preventive maintenance process that may be performed while equipment is still running.
Additional pros of preventive maintenance
Scheduling. Preventive maintenance permits scheduling of manufacturing and maintenance personnel to optimize their productivity. It also allows companies to proactively manage production quotas. For example, a company may plan on overtime expenses to produce enough product, accounting for losses during maintenance.
Purchasing. Purchasing of spare parts can be planned to coincide with the scheduled maintenance and decrease spare parts inventory.
Cons of preventive maintenance
Excessive maintenance. As with the smoke detector battery, PvM may dictate performing maintenance earlier than necessary. This is especially true if many items are addressed at the same time. Some may require maintenance more frequently than others, but to save on shutdown costs and improve maintenance efficiencies, may be performed at the same time. It may be desirable to schedule multiple preventive maintenance intervals; some to address more frequent, quicker maintenance items (lubrication, belts, etc.) and others for longer, more involved tasks such as bearing or seal replacement.
Maintenance programs should also take technology improvements into consideration. Changing to a different type of lubrication or different belt material may provide a longer service life, increasing the interval for maintenance shutdowns.
Cells and associated processes and equipment. If a machine is part of a cell, or if the company follows a kanban or just-in-time manufacturing process, shutting a machine down for maintenance can have a domino effect. Upstream and downstream equipment and processes may be affected and need to be considered.
Resource allocation. Employee and equipment resources will need to be planned for an efficient maintenance program. For example, multiple machines requiring the same mechanics or maintenance equipment should not be planned for the same time.
Startup costs. Implementation of a PvM program may require additional funding to get off the ground. For example, if it is desirable to perform maintenance on multiple machines at the same time, additional personnel and equipment resources may be required. Outside resources may be required if the maintenance cannot be performed with existing personnel and tools.
Predictive maintenance
Source: ITT Inc.
As we've seen, predictive maintenance shares many of the advantages of preventive maintenance programs. A key difference is that predictive maintenance requires additional information to institute compared to a preventive maintenance process.
Sensors and other condition monitoring solutions detect changes in equipment operation, indicating that a maintenance item needs attention. Instead of shutting down based on an estimated lifecycle maintenance schedule, PdM often relies on sensor data to determine the real-time condition of the equipment and maintenance is scheduled based on predictive failure indicated by these readings.
Monitoring machine condition with the following sensors and equipment is common to a good PdM program.
- Temperature sensors detect excess heat due to lubrication, wear or component failure issues.
- Vibration sensors detect excess vibration levels that indicate problems with bearings, shafts, pumps and motors.
- Ultrasonic sensors detect flaws or cracks that can develop due to degradation of welded joints or material fatigue.
- Thermal imagers show heat patterns in the infrared wavelength spectrum. These hotspots may be an indication of friction due to component wear.
- Oil sensors and analyzers, also known as ferrography instruments, sense oil levels and check for contamination, particulates, viscosity or temperature. Contamination or particulates may be an indication of seal failure, leaks or wear. An increase in oil viscosity or wear will increase operating temperature due to friction.
Some condition monitoring products, such as ABB's smart sensors for motors and Ability smart sensor for mounted bearings, incorporate multiple sensors and wireless technology to provide a comprehensive health check on equipment that can be viewed remotely on a smartphone or computer.
Additional pros of predictive maintenance
Cost savings. In addition to the costs shared by preventive maintenance, PdM will save money by eliminating the costs associated with performing maintenance when it is not required as dictated by a PvM program. This saves both spare parts and labor expenses.
Cons of predictive maintenance
Resource allocation. Since maintenance is dictated by machine condition and not the calendar, resource planning is more difficult than preventive maintenance programs. However, it is superior to breakdown maintenance plans as problems can be detected before a catastrophic failure.
Startup costs. Purchasing and installation costs for sensors and instrumentations can be costly.
Education. Predictive maintenance requires personnel training and expertise to understand machine condition as well as sensor and instrumentation use and data interpretation. Although this is an additional cost and time expense for the company, the result is employees who are more skilled and in tune with machine condition and operation.
Choosing the approach
Deciding on and implementing a good maintenance program can save a company significant costs associated with unexpected breakdowns.
The pros and cons of preventive maintenance and predictive maintenance should be considered, as well as the costs to implement each program. The correct solution is different for each company and application.
For example, the same company may decide that a predictive maintenance plan may be best for a high-volume cell while a preventive maintenance plan is better for low-volume equipment. Engineering and manufacturing firms should perform careful planning and research before deciding the proper course of action. If needed, a condition monitoring and machine maintenance service provider can help develop, document and implement the best program.
I dunno. Seems to me the only real difference is based on the accuracy of the predictor. A calendar-based prevention plan is someone's guess at predicting. Hopefully it's in-house by someone who knows his stuff, and not a third party who assumes you use it like "everybody else" does. Granted, sensors take a lot of guess-work out of the equation. But, what's the real difference between the two? In both cases, you try to predict a failure that you want to prevent. A distinction with no real difference.
As long as you do the best you can with what you got, then you can't expect any more. It's a matter of due diligence.
In reply to #1
Technically, I believe you are correct. In theory, the accuracy of predictive maintenance is greater. The differences are subtle and both are really preventive maintenance, it's just how you determine and when you plan to perform the maintenance. Also, another major benefit of predictive maintenance is in cases where unexpected problems arise. For example, a temperature sensor may pick up a rapid increase in temperature or a sensor bearing may detect vibration due to contamination or some other unexpected issue. This type of predictive maintenance can save a lot of money compared to a catastrophic failure if the condition change had not been monitored or noticed.