Predictive maintenance is a critical component of a comprehensive reliability program. The practice of monitoring equipment for defects and potential failure has been shown to increase uptime, reduce maintenance cost, and prevent catastrophic equipment failure. However, the role of the predictive team has evolved beyond just monitoring equipment for signs of defects. The new role of the predictive team is to follow up on root cause analysis and forensic analysis, then to provide recommendations on how to eliminate the conditions that cause defects. In other words, predictive teams now hold a proactive role.
The Traditional Role of Predictive Maintenance Team
Traditionally, the role of the predictive technician was to focus on defect detection and recommend corrective action. For instance, if a vibration analyst detected a defect in a motor bearing severe enough to require replacement, they would write the observations in the report and recommend replacing the bearing. However, if the condition that caused the defect, such as electrical fluting, was not addressed, the issue would likely reoccur. In this scenario, the predictive technician would repeat the cycle of detecting and replacing the bearing, which results in increased costs and decreased equipment reliability.
The New Role of Predictive Maintenance Team
In the evolving role of a predictive maintenance team, the focus is on conducting thorough root cause analysis to identify and eliminate the conditions that cause defects. For the same scenario, if a predictive analyst detects a defect in a motor bearing that may be caused by electrical fluting, the predictive team would follow up. They would verify the presence of fluting marks by physically inspecting the removed bearing. Then the team would review the motor nameplate to ascertain if the bearing has protection against fluting and whether the motor runs on a variable frequency drive—a drive which is known to generate electrical fluting on motor bearings.
In this scenario, the predictive maintenance team could also recommend the use of a voltage meter. Measuring the voltage on the shaft can confirm if current is flowing from the motor to the shaft through the bearing, which causes damage to the bearing. If the electrical fluting is confirmed, the predictive team will suggest installing a bearing with protection against electrical fluting. After installation, they would measure the voltage again to ensure that the issue is resolved. Going even further—the predictive maintenance team could proactively recommend the examinations of other critical motors in the plant with similar operating conditions. That would be mean inspecting all motors running on variable frequency drives with no electrical fluting protection.
This is just one example of how a predictive maintenance team could approach the issue of electrical fluting. The same forensic analysis follow-ups can be used to identify other defects, which can be compared to data or other processes. One example is personnel skills and procedures for alignment in case a defect due to misalignment is found physically in the bearing supported by vibration data.
The new role of the predictive maintenance team includes the identification and elimination of root causes, thereby reducing downtime and increasing equipment reliability. A team's approach should be meticulous and comprehensive, and its recommendations should be based on data and engineering principles.
Continuous Improvement
As part of continuous improvement, the predictive analysis team must consider the percentage of repetitive detected defects and if there are patterns across all defects found. For example, a pattern could be found in machines with belt tension and misalignment issues. Worn-out sheaves could result from multiple causes, such as lack of a preventive maintenance to measure and replace sheaves. Maybe inspections are in place, but personnel aren’twell trained well. Perhaps there is no accountability to ensure work effectiveness. The predictive team identifies trends. They understand that patterns and issues found through vibration analysis and predictive analysis should be solved at the root.
Challenges
Despite the significant benefits of predictive maintenance, there are still several challenges that need to be addressed. One of the major challenges is the lack of skilled personnel. Predictive maintenance requires specialized expertise in data analysis, root cause analysis, and mechanical forensic analysis. Unfortunately, these skills are often in short supply. Companies need to invest in training their current employees or hire new ones to fill this gap.
Another challenge in implementing predictive maintenance is that not all companies have an in-house predictive technician. They might have an external contractor who performs data collection and analysis instead. This can result in a disconnect between the external service technician and the in-house reliability team who need to follow up on the forensic analysis from the removed component. To overcome this, the in-house reliability team needs to share feedback with the predictive analysis service contractor—letting them know the findings. Then they can verify if the analysis was correct, determine trends, then identify areas of improvement together. This approach may deter companies due to the cost increase. In the long-term, the benefits pay off when repetitive failures are avoided and the root causes of problems are eliminated. The predictive analysis service contractor needs to be informed of findings after components are replaced and conditions are fixed, and the predictive team needs to have a collaborative attitude and a proactive role.
