3DSignals, when innovation listens to machines

Published on Wednesday, 18 October 2017

“The day offered an opportunity for discussion in which we wanted to be more practical and less theoretical in order to understand the potential uses and advantages of this technology and how to employ it in different areas of Operation & Maintenance”

– Iacopo Mazzantini, Innovatino Hydro, Enel

Workshop participants shared this sentiment: In a survey that followed the workshop, most of the participants said that they believe that acoustics monitoring technology should be applied to maintenance-related applications and processes, and that this workshop was relevant and useful to their job.

The workshop began with a demonstration by 3DSignals of how the system functions: microphones and commonly found electronic components were placed near hydroelectric turbines, generators and alternators in three Hydro plants. The sounds recorded continuously by the microphones were sent to a server where they were archived and then analysed by artificial intelligence algorithms. This allowed the creation of a data bank of sounds available to our technical experts who, unable to be present at the plants, could listen over and over to the moment in which the machines malfunctioned or broke down and thus better understand the event that caused the fault or anomaly. More importantly, thanks to machine learning tools, this technology learns to read sounds and understand when the machine is working well, or rather, understand the ‘normal’ sound of its proper functioning and detect when there are variations in this sound. Over time, the software learns to recognise which sound variations relate to an actual malfunction or indicate an imminent fault.

At the hydroelectric plants in Cencenighe and Soverzene in northern Italy (Veneto), followed by the plant in Riba-roja in Spain (Catalonia), the pilot activities to apply the technology have provided the first positive results, identifying during the first three weeks some malfunctions of the machines and components.

This type of predictive diagnostic and maintenance tool enables the optimization of maintenance costs, while ensuring maximum efficiency in energy generation.

“Utilizing the acoustic monitoring technology and deep learning systems, 3DSignals is transforming the use of sound into a reliable source to better understand the health of machines and to extend their service lives,” underlined Amnon Shoenfeld, CEO of the Israeli startup. 

“Enel’s Innovation department is providing us with a great opportunity to verify a use for our technology in the real world. We are currently trying to understand how to extend it to other types of plant and machine. I hope that the Pisa workshop has offered other Enel business lines the possibility to see the added value that we can bring to maintenance procedures and the equipment that they use”

– Amnon Shoenfeld, CEO 3DSignals

The workshop in Pisa, in fact, offered a group of around 20 internal experts from different areas of business at Enel, the chance to work together to brainstorm which fields of operations and maintenance activities sound monitoring could be applied to. The result was the identification of over 15 potential machines, devices or operating environments, some of which are already subject to automated controls, in which to roll out this technology in order to evaluate its effectiveness and determine the cost-benefit relationship.

The experience of the Hydro division of the Global Business Line Renewables is a clear example of the Open Innovation approach that our company is using in order to simplify the production of electricity. The technological partnership with 3DSignals is the result of the scouting activities carried out by our Innovation Hub in Tel Aviv. At the many meetings held at the Israeli headquarters, Innovation managers from different segments of our Business Lines identified the technology of 3DSignals, together with that of other startups. The O&M technicians of the Hydro area glimpsed the potential of the sound monitoring system and met with the startup to establish a technology partnership agreement and move the project on to a pilot phase that led to the workshop in Pisa in October.

The meeting proved useful to put into practice Enel’s innovation model that, in the interests of creating shared value, aims to generate opportunities for our business and for the startups with which we come into contact, through a transversal multidisciplinary approach.

Open Innovability®

Use our crowdsourcing platform to put forward your technological solutions