A Facility Manager’s Crystal Ball: Predicting Equipment Failures Through Facility Maintenance


By Yaniv Vardi

Facility managers (fms) are charged with keeping equipment running efficiently. In many cases, however, they are assigned this responsibility without the proper tools necessary to fulfill it. To ensure the efficiency of equipment, fms need information, more than anything else. Beyond the reputed crystal ball, such information is the only way to predict and prevent equipment failures.

In most manufacturing plants, the demands on fms are high; but the data is insufficient or nonexistent. For example:

  • The energy consumed by individual components of the production line, or by individual machines, is not always measured or monitored;
  • The high amount of energy consumed by machines, conveyers, agitators, air compressors, etc., is not monitored and not correlated with the manufacturing plan;
  • Management has no way to evaluate the amount of energy required for the daily/weekly scheduled manufacturing plan;
  • The correlation between energy and machine performance is unknown;
  • In many cases, even the off/on status of the device is not monitored; or
  • Real-time data is rarely used to make decisions.

Without this kind of information, fms are left to rely on planned preventive maintenance (also known as PMM or Planned Maintenance) schedules, which can be expensive and can cause unnecessary downtime.

Predictive Maintenance Based On Data

Instead, equipment failures can be predicted by monitoring the energy demand of certain systems and devices. Monitoring, tracking, and analyzing device-level energy consumption data is much easier and less expensive than most fms realize; and the benefits of such granular data are immense.

When we install wireless, nonintrusive sensors on each of our devices, and aggregate the data through a cloud-based analytics engine, we can monitor, track, benchmark, report, and detect anomalies, all in real-time.

For example, a compressor that has begun to cycle more frequently than usual, or is out of sync with the external temperature or humidity, will display an energy profile that is characteristic of a particular failure mode. Or a conveyor motor that overloads and trips out can create a costly bottleneck in the process. Correcting these anomalies can increase operational efficiency and productivity, and it can also optimize and reduce energy consumption.

More subtly however, is when machinery begins to draw more and more electrical power. This upward slope, be it gentle or steep, is perceptible and can be monitored by real-time sensors. If a belt or bearing is causing this increasing draw, an alert notification would be triggered.

This shift from planned maintenance, which is costly and time consuming (often unnecessarily so), to predictive maintenance, which is based on real-time, device level data that enables fms to predict equipment failures, has become a necessity. By basing maintenance, repair, and retrofitting projects on the data that can determine which systems truly need it, we save time and money.

Predictive Maintenance: The Added Benefits

When fms transition to a predictive facility maintenance process, naturally, budgets and schedules are improved. Because planned maintenance programs often include repairs that may be unnecessary, its predictive counterpart saves money and time, and it can eliminate downtime that is not truly required.

Some fms may opt to keep waiting for a crystal ball that can predict equipment failures; the more realistic among them will instead opt for the crystal clear visibility into energy consumption, which makes the crystal ball obsolete.

Vardi is chief executive officer at Panoramic Power. Yaniv is a seasoned executive with close to two decades of executive leadership experience in the Enterprise Solution Industry. As CEO of Panoramic Power, he oversees the day-to-day operations of the company as well as provides vision, strategic direction, and focused execution for the company.