Good vibrations

Thoroughly analyzing the vibrations made by a machine can help diagnose its mechanical state of health.

In Turin, Italy, a European research project aims to understand the vibrations made by machine tools. The idea is to use the data gathered by a network of sensors to forecast and therefore prevent breakdowns.

Alessandro Zanella, a physicist at the Fiat research Centre, said: “Vibrations come from rotation and moving parts inside the motors, and make up a unique signature for each machine. Using them, we can make a predictive diagnosis of the machine’s mechanical state. If certain vibrations are generated, they can indicate mechanical faults.”

A ball bearing that is about to break generates a peak vibration. This means the fault is detected before the machine breaks down due to vibration sensors, which use kinetic energy, meaning they are self-powered.

The graph representing the analysis of the vibrations can be read just like an electrocardiogram. Complex algorithms produce detailed information about the state of the
machine.”

Located between the sensors and the central unit, the nodes reduce the complexity of the data collected by sending the server only significant data, as features.

Mihai Marin-Perianu, a computer science researcher, with Inertia Technology, said: “We analyze the data stream in real time and use some sophisticated methods of checking the data in time and frequency domains. Then we extract the features that can pinpoint the presence of defects and send only these features upstream to the central point.”

We will rely on this distributed intelligence more and more in order to be able to monitor virtually any piece of equipment in industrial sites. This will be at an unprecedented scale but at an affordable cost.

Fewer malfunctions makes machinery more efficient, and produces higher quality goods, but also makes machinery safer.

Paul Havinga, a computer science professor at the University of Twente and a WiBrate Project Coordinator, said: “Vibrations are everywhere in the world. They can be found in planes, helicopters, machinery, and trains. And vibrations give an impression of the quality and the status of the machinery itself, and whether a machine is likely to break down.”

In the coming years, technology for monitoring vibrations will appear in trains, planes and factories all over the world.

www.wibrate.eu