Design News is part of the Informa Markets Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Predictive Maintenance Is Replacing the Plant’s Retiring Knowledge Worker

RRAMAC Connected Systems, Tom Craven, predictive maintenance, artificial intelligence, machine learning, sensors
Sensors and artificial intelligence are beginning to replace retiring Baby Boomers.

The retiring Baby Boomer at the plant may get replaced by predictive maintenance software. Just as robots are stepping up to do the mind-numbing and dangerous repetitive manual labor jobs in manufacturing, we’re now seeing that sensors and artificial intelligence begin to replace the plant’s knowledge workers who can smell a failing motor at 50 yards.

This screen shot shows a predictiv.e maintenance system monitoring equipment health. Image courtesy of RRAMAC Connected Systems.

Those highly knowledgeable workers with 20 to 40 years of experience with pant equipment are retiring in a stream that will soon become a flood. These workers have gained years of expertise ferreting out the weakness of plant equipment, judging its health through sound, vibration, even smell. You can’t hire a new grad to replace this expertise. But you can hire software to do it.

Using Sensors to Replace the Five Senses

These experienced workers have been doing predictive maintenance on plant equipment by using their five senses. “The retiring knowledge worker has been doing the same job for 20 or 30 years. They can hear things in the equipment,” Tom Craven, VP of product strategy at RRAMAC Connected Systems, told Design News. “What they do gets into operational efficiency, but also, it’s predictive maintenance. They’ll hear a rattle and know what to do about it. Or they can detect a specific smell that can be a motor current problem. They smell the electrical burn and know that something bad is going to happen.”

Craven will present the session, Best Practices in Successfully Performing Predictive Maintenance, at the Advanced Design and Manufacturing Expo in Cleveland on March 8, 2018.

Craven noted that software is getting developed to detect anomalies in plant equipment, with the goal of catching the failing equipment before it causes any stoppage. “That rattle that the knowledge worker hears has other symptoms that can be picked up by a vibration sensor,” said Craven. “The predictive maintenance system is a combination of sensors and machine learning. The knowledge worker is responding to his five senses and he has the experience to know what to do to correct the problem. The machine learning feeds the system data that knows what to do to correct the equiopment.”

Using Data to Analyze Equipment Health

Predictive maintenance systems take equipment health monitoring into the realm of data analysis, creating a data picture of a healthy machine and searching for anomalies that may be out of sync with that picture. “The machine learning is artificial intelligence applied to a machine. What happens over time is you record data, which includes multiple vibration points, motor current, and temperature,” said Craven. “You look for anomalies. In some cases, it’s obvious – motor current issues are easy to associate with a failure.”

The analysis of machine health becomes complicated when the machine – over its normal course of performance – goes through changes in its vibration, voltage, or temperature. “Where data analysis becomes more complex, is when an anomalous vibration may be normal. Vibrations may vary during the cycles of the machine,” said Craven. “What happens is in these cases, is you look at multiple variables and run a mathematical calculation that can flag the anomaly.”

The artificial intelligence can learn the difference between anomalous machine readings that are healthy and anomalous readings that are unhealthy. “The predictive maintenance system detects when something has changed and is atypical. You can teach the system that this atypical change is OK,” said Craven. “The system learns that one particular anomaly can cause this failure, another anomaly can cause a different failure, and yet another anomaly does not result in a failure at all.”

Rob Spiegel has covered automation and control for 17 years, 15 of them for Design News. Other topics he has covered include supply chain technology, alternative energy, and cyber security. For 10 years, he was owner and publisher of the food magazine Chile Pepper.

For the second year,  Advanced Design & Manufacturing Cleveland  is back at the Huntington Convention Center, March 7-8, 2018.  Register today  for loads of free, can’t-miss education focusing on Smart Manufacturing, 3D Printing, Battery technologies, Medtech, and more!

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.