“If you can find a path with no obstacles, it probably doesn’t lead anywhere.” 
— Frank A. Clark

"So easy that even a caveman can do it,” as stated in a popular TV commercial, could easily be used to describe today’s predictive maintenance tools because they work so well. However, to be truly competitive, a company’s goals should go further than being satisfied with marginal improvements in machine reliability.

Achieving equipment reliability that is required for maximum profits is both realistic and obtainable for any company. Proper use of predictive maintenance (PdM) tools is a key factor in realizing such goals. This article provides solutions to overcoming obstacles and issues associated with monitoring machinery and using predictive maintenance tools—such as precision shaft alignment and vibration instruments.

Training

Deciding who receives training is important. Many times predictive maintenance technicians are the only people trained in using predictive maintenance technologies. This can sometimes put them at odds with frontline supervisors who may not understand the importance of such methods. 

A solution is to train these supervisors in the basics of the predictive maintenance technologies used in the plant. Excellent, cost-effective training can usually be obtained from the  equipment vendors. All stakeholders in machine reliability should have a thorough understanding of the benefits and uses of predictive technologies.

Reports

Poorly written or non-existent reports are another factor that may limit predictive maintenance success. 

If detected faults are only reported by word of mouth, they may fall through the cracks and not be acted upon. Good, concise reports that provide monetary savings and correctly state problems in an easy-to-understand format are important. 

A good technician producing poor reports will not be successful.

Inaccurate Alarms

Alarms that are not accurately set will allow for missed faults or result in crying “wolf” when no faults are present. Predictive maintenance software is so well-developed that the computer can be good at gleaning faults. However, this can only happen if alarm levels are properly set. The curve fitting features of some software may even provide a good estimate of when a measured parameter may go into an alarm condition. 

Technicians may not know all the features of the software they use and may also not make use of every feature that could benefit their predictive maintenance process. An alarm should be thought of as a tolerance standard. Without standards, end users cannot know the condition of their equipment. An alarm standard can be treated as a measure of machine reliability.

Adequate and Organized Data Collection

Inadequate data collection may also result in poor predictive maintenance work. Attention to detail in data collection is paramount to success. There should never be any doubt about where and how previous data was collected. If those parameters are not known, the measurement is not repeatable to any degree of confidence. 

No aspect of predictive maintenance is more important than data collection. Only good data provides information for making good decisions. Sometimes, technicians may not keep their databases current. This can make the predictive maintenance process out of date if current information is not readily available. 

Within a week, new information will require that the databases are tweaked, whether a new machine is installed or an alarm setting is adjusted. Conducting predictive maintenance with out-of-date information can lead to making the wrong reliability decisions.

Lack of Compliance and Sustainability

Too often a company or plant will achieve success through their predictive maintenance efforts only to slide back into the bad habits of poor maintenance. Gaining initial success is usually easier than sustaining the effort. Just knowing this fact can help an organization continually improve its predictive maintenance efforts. One of the causes of this regress is bringing new managers into the organization who are not well-versed in the benefits and methods of predictive maintenance strategies. This is simply a training issue that can easily be addressed through proper training and by having a well-documented physical asset management strategy. 

Another reason for such regress is that managers new to the organization may be biased toward an entirely different asset management strategy, and they may want to toss out the old and bring in the new. Managers are brought into the organization with hopes that they bring new ideas with them. However, these new ideas should not be implemented to the detriment of established processes that have been proven successful. The best option in these cases is to have those new to the organization incorporate their ideas into the working process. The goal would be for the new input to add value to the strategies already in place and not to replace current, successful predictive maintenance processes.

Root Cause Analysis 
& Repair

Another obstacle is the employment of predictive maintenance tools that only detect defective parts, creating a vicious cycle of doctoring symptoms without addressing the root causes of failure. This is a common practice in many organizations. Good predictive maintenance is about solving problems. Predictive maintenance tools are a great benefit in helping identify the root causes of failures. 

For example, vibration analysis can detect unbalance, misalignment, electrical faults and lubrication faults, as well as many other issues. Suppose that through vibration testing a technician learns that a fault is lubrication related. This fact is a clue in getting to the root cause of a problem. Predictive maintenance tools can and should play a strong role in root cause analysis (RCA).

Not addressing failures will most certainly lead to profit losses, as well as loss of confidence in the predictive maintenance process. One of the best ways to identify potential failures before they happen is to look at past failures. A glance back may provide a clearer vision into the future. It is critical to track all failures. Of course, some failure modes may have never happened but are likely to happen, and tasks should be developed to detect these potential failures, as well. A large obstacle to achieving the required equipment reliability is not having good documentation on failure history.

Proper Machine Management & Operation

Another obstacle to success is failing to understand how to properly manage a machine to attain the required reliability. Having a well-defined and documented strategy is essential for success. Machines can be managed in many ways, and several different strategies may produce positive results, but there is only one best strategy for every machine. End users should continually strive to find and implement the best strategies.

Do not continue to operate equipment with known defects that may result in failure. Develop a defect elimination approach by identifying potential failures (failure modes) that are likely to occur. 

Develop tasks to detect them before they result in failure. Remember, it may not always be necessary to prevent a failure. Reducing or eliminating the consequences of a failure may be all that is required. If the consequence of a failure is negligible, then the best strategy may be run-to-fail. Always strive for the greatest value and safety with minimum impact on the environment.

A Culture for Success

Another obstacle that may limit success is not knowing how well your predictive maintenance processes are working. Even after years of hearing about the importance of measuring all the elements of the equipment reliability process, many companies still fall short of such measures. Excellent gauges have been developed for all aspects of machine reliability and just a little research will produce the measures needed in any process. The work has already been done by others, so why not take full advantage of their efforts?

Perhaps the greatest obstacle to predictive maintenance success is the cultural aspects and not the technical issues. A concerted effort must be maintained until predictive maintenance becomes an integral part of plant culture. 

A fact-based and well-documented predictive maintenance strategy leaves nothing to chance and is a key factor in ensuring that the process will keep plant machinery operating at the reliability required for producing the highest quality products at the lowest costs.

If a company can avoid or overcome the obstacles described in this article, the road to reliability will be satisfying, and the predictive maintenance process will be among the best in industry.

Pumps & Systems, May 2012