- March 07, 2017
By Bill Lydon, Editor, Automation.com You know that feeling when you walk past a machine, and something just doesn‚Äôt quite sound right? This article discusses the use of audio analysis sensors and deep learning technology to predict equipment problems and generate alerts for preventative maintenance actions.
By Bill Lydon, Editor, Automation.com
You know that feeling when you walk past a machine, and something just doesn’t quite sound right? Most of us have, yet what if we’re not around to hear that malfunction? This was the basis of a discussion I recently had with Amnon Shenfeld, CEO of 3DSignals, an Israeli-based innovator. We discussed Shenfeld’s company, and how they utilize audio analysis sensors and deep learning technology to predict equipment problems and generate alerts for preventative maintenance actions.
Amnon Shenfeld started thinking about the idea for 3DSignals in 2013 while on a train ride from Cambridge to London. The train started to make a lot of strange noise and he thought if there was a technician on board, then he or she would almost immediately know what was happening based on the sound alone. This idea spurred the creation of 3DSignals and has attracted a number of investors including Grove Ventures. founded by Dov Moran. Moran, of course, is the inventor of the USB flash drive, and former CEO and founder of msystems flash data storage, which was sold in 2006 to SanDisk for $1.6B.
This sound-based maintenance discussion was particularly interesting to me since I worked on many projects, back in the 1970s, with intercoms mounted on control equipment cabinets. The intercoms were designed for operations and maintenance to use in order to communicates, as wireless communications was very expensive in those days and not as reliable as today. Many operations and maintenance people also relied on the intercoms to remotely listen to equipment, in order to diagnose problems and check on the health of equipment.
Today, the common predictive maintenance methods for monitoring the condition of industrial and manufacturing equipment, utilize temperature, vibration, power consumption, and visual inspection. Yet, the one important sense that engineers and technicians have used for years, sound, has been curiously left out. The 3DSignals offering is designed to go beyond human listening limitations. The sensors are mounted physically next to each machine, monitoring at all times during operation, and deploy sophisticated filtering that separates out interfering sounds, especially noise from other machines.
Figure 1: Acoustic monitoring of machine sound patterns is holistic accounting for a wide range of issues opposed to a sensor measuring one aspect of equipment operation.
Innovative Applications of Modern Technology
3DSignals’ business is based on a patent-pending system that makes use of ultrasonic airborne sensors along with deep learning software algorithms coupled with predictive analytics. The ultrasonic sensors are mounted near rotating equipment and communicate with an embedded computer which detects abnormal audio patterns. This device communicates to the cloud using LTE cellular communications.
3DSignals cloud software leverages deep learning algorithms, including the application of predictive analytics for sound signal processing, the definition of normally-operating audio patterns, the aggregation of data, along with the creation of alerts and report generation. In addition, the device communicates with on-site systems, including asset management and automation, over a standard Ethernet.
Ultrasonic airborne acoustic sensors then become the ears of the system, which predicts electro-mechanical machine failures by detecting anomalies in sound when compared to optimally running machine patterns.
The software can be hosted in public or private cloud servers, using cybersecurity solutions to maintain system integrity.
Figure 2: Acoustic patterns are used to analyze equipment health
VIDEO: 3DSignals Metal cutting application reduces production downtime by preventing unexpected machine failure.
The 3DSignals acoustic monitoring solution is designed to be nonintrusive and does not physically contact the monitored equipment. Mounting the ultrasonic sensors makes retrofitting old assets simple without requiring the addition of internal sensors.
3DSignals’ current customer base includes names like MacSteel International, one of the largest international steel conglomerates; and Enel Green Power, an Italian multinational renewable-energy corporation and subsidiary of the power generation firm Enel.
Figure 3: Sensors areconnected wirelessly and software can be deployed on either a public or private cloud. The directional microphones used eclectic acoustic data does not capture people talking by design.
3DSignals provides a bright new example of how out-of-the-box thinking is leveraging the technological advancements in exciting new ways.
- InTech - IoT impact on manufacturing
- Manufacturing: Feed Innovation or Starve
- Internet of Things Driven Manufacturing
- Industry 4.0: Intelligent and flexible production
- Manufacturing Competitiveness Requires Awareness
Did you enjoy this great article?
Check out our free e-newsletters to read more great articles..Subscribe