Modern Predictive Maintenance (PdM) has come a long way in recent years, with ease of use and installation both having increased significantly, while cost has similarly decreased. Conventional PdM was an expensive and time consuming technology and, because of this, it was reserved for only the most critical facilities. The problem with this is that less costly equipment, though seemingly peripheral, can be just as essential to continuing operations, and it is instead relegated to a preventive maintenance (PM) schedule.

Predictive vs. Preventive Maintenance

Preventive maintenance relies on a schedule of inspections and part replacements to keep equipment up and running. The primary issue with PM is that downtime is often unplanned. PdM, by comparison, offers a real-time view into the status of your equipment and allows you to plan convenient times to take machines offline and perform maintenance. Furthermore, PM involves little to no upstart costs, and as such, is the default method for facilities maintenance. PdM, by comparison, has traditionally cost thousands upon thousands of dollars to implement, and the cost only increases as more machines are monitored.

Until recent developments, implementing PdM involved installing individual monitors on each and every device you wanted to monitor. As such, technicians might have to ascertain systems over others in order to save time and money. In a silicon fabrication facility, for example, Ultra-Purified Water (UPW) is pumped continuously to rinse and cleanse silicon wafers. While the machines that produce the silicon are often monitored, the fans and pumps that keep the room cooled are not. If the pumps or fans fail, however, the plant is unable to produce silicon and thus must absorb the cost of downtime.

With mobile hardware and cloud-based computing, PdM is now available at a fraction of the cost of systems that were once hardwired and required highly-trained technicians to analyze results. This means that you no longer have to choose one system over the other and can simply work to keep all systems running smoothly and online.

When considering PdM versus PM, there are simple questions you might ask yourself to justify the move:

1. If the auxiliary systems fail, what effect would this have on overall operations and what would the cost be?
   
   A fan or a pump may be a simple and cheap replacement, but only if you have the right parts available when needed. That simple and inexpensive replacement can dramatically increase in cost when downtime is unplanned and your facility is taken down with it. With this in mind, would the costs of PdM versus PM be justified?


2. What are your common failures?
   
   Look at the most common equipment failures in your facility and compile a list. Could any of them be missed by the kind of human inspections performed during PM? If they were detected, could they be found in time to schedule a repair outage without incurring emergency downtime?


3. What systems am I overlooking as secondary that would ultimately affect the bottom line?
   
   It can be easy to overlook equipment as essential to keeping operations running smoothly until it fails and you find yourself with unforeseen problems. For example, power generation facilities have very expensive gas turbines that generate the power, but the closed water system is driven by pumps. Data centers similarly rely on pumps to keep systems cooled. Maybe these have gone down in the past and were quickly fixed, but now these same parts are on a long wait time, and would take the whole system down for an extended time. With PdM, you can get insight into the condition of these seemingly secondary systems and plan ahead.