Replacing the Engineer’s Ear with Sound Detection

Augury’s condition-monitoring system detects the health of equipment by comparing machine sound to a database of machine noises.

The sounds machines make reveal quite a bit about whether they’re working properly – what’s wrong and what’s not. The first thing an engineer does when a machine doesn’t seem to be operating properly is to listen to it. In the past, the machines themselves had no way of listening to their own sound to determine health, and therefore they had no way to alert the automation staff that something is wrong. That’s changed.

ultra sound, condition monitoring, industrial equipment, automation and control


Recently, Augury, a predictive machine diagnostics company, added sound sensors to its range of condition monitoring. The company designed tools that listen to any machine – from a jet engine to a household appliance – and run it through an algorithm to identify what’s wrong and even predict what’ll break next.

In developing sound analysis, Augury hones in on specific types of equipment. “We focus on three types of machines: pumps, fans, and chillers or compressors. Within these three families we can diagnose any of the machines we encounter,” said Saar Yoskovitz, CEO of Augury. “We don’t need to generate an algorithm for this pump versus others. The sounds these machines make have a unique fingerprint and we can detect it across any model. Like a fan belt squeaking. It doesn’t matter whether it’s a Mercedes or a Toyota, a fan belt sounds like a fan belt.”

ESC logoSecuring the Internet of Things. Today's IoT devices are under increasing attack. Device manufacturers and embedded software designers must be vigilant if they are to provide a secure system for applications to do their work. Learn more about securing IoT devices and applications in the Connected Devices track at ESC Silicon Valley . December 6-8, 2016 in San Jose. Register here for the event, hosted by Design News ’ parent company UBM.

Ultrasonic and vibration sensors are common in predictive maintenance. The difference with Augury's sensor processing is that it's Internet- and algorithm-based. The data is sent to servers where it is compared with previous data collected from that machine, as well as data collected from similar machines. The platform then detects changes. “We rely on a specific model for your specific machine and analyze it during its operational day so we can alert you to changes,” said Yoskovitz.

The analysis is done in real time and the results are returned to the control engineer within seconds. Sound readings are taken from the machines through Internet of Things technology. The readings are put through a mechanical diagnostics layer, and the results are transmitted back to the user through a smart device.

The automation industry has seen an increase in interest in condition monitoring since Baby Boomer engineers with decades of experience have retired. Those engineers could listen to a machine and instantly determine its operating health. That’s a hard skill to pass on to a new engineer. “You only have two or three years to infuse the expertise that is leaving with the Boomer,” said Yoskovitz. “Technology is


Jerald Cogswell's picture
Such expert listening is a great use for those tiny MEMS microphones (See my Gadget Freak #240 article). Coupled to an opamp, their signal can be run through either analog or digital filters to an analysis machine. Analysis can be as simple as a band pass filter that detects a squeal or an FFT with a comb filter looking for specific combinations of frequencies. Lots of uses. But beware of info overload in this IoT. We need smart machines to filter our info. Think of uses in your home or car.

The first diagnostic vibration monitoring system we sold was back in 1985. At least one version used that wonderful Spectral Dynamics model SD210 analyzer. That wonderful device not only did the spectral analysis, it also reported if the amplitude in any of the multiple frequency bands exceeded the programmed limits. It was a very handy device for testing. I have not been able to find any of the SD210 systems available recently, no idea where SD went.

Actually, asking the machine operators about the sound of the machine has been a useful service technique of mine for many years. Some operators are able to deliver a great deal of useful information, and they appreciate their input being considered worthwhile. Of course, there are also those operators who either never listen or whose attention span is so short that they have no clues about anything changing. Their observations are less useful.

Add new comment

By submitting this form, you accept the Mollom privacy policy.