IoT Blog


12 Jul 2016


The Nostradamus of Machines

MIT Technology Review writes about a new app developed by the Israeli company Augury. Auguscope, as the app is called, can be used to listen to the sounds a machine generates. The sounds are then sent to a server that analyzes and compares them with a database of how the machine is supposed to sound. If something is strange, the app hears it and can give the user a warning that the machine needs to be serviced. And this could be done before the machine starts to show any signs that something is wrong.

Augury is the first company we have heard of that troubleshoots machines using sound, but similar functionality already exists. We ourselves have outlined apps that could be used in the same way, but instead of sound we look at vibration. As long as the app has a chance to learn how the vibrations of a normally-functioning machine feel, it can then give a warning as soon as the sensor detects any events outside the ordinary.

This area is part of what is called predictive maintenance, a hot concept for factories where the cost of a stationary machine can be devastating. By getting an indication in advance of when and how a machine needs to be maintained – and maybe even ordering the right parts – downtime can be minimized or even eliminated.

As with many trends, the industrial sector is the leader, but you do not have to be Nostradamus to say that predictive maintenance will soon move to public transportation vehicles and all systems in the world of real estate. These are just a few examples.