

ITI is working on SOROLL-IA, a project funded by IVACE and ERDF funds.
Noise is one of the most significant environmental factors in terms of public health. Today, one in four people in Spain is exposed to high levels of noise pollution. The sources of noise are very diverse, ranging from road traffic, construction work and industrial plants to air traffic. And while it is true that the European Commission has established various Directives to mitigate its effects, there is still a lack of specific data on the sources of this noise that could be turned into action plans.
In order to address this situation, ITI, a technology centre specialising in ICT, is working on the SOROLL-IA project, which aims to create a database of generic audio recordings in a port environment.
Along these lines, ITI will collaborate with the Port of Valencia to advance the research, for which a series of acoustic nodes have been deployed at strategic locations around the port. The work, in which the Valenciaport Foundation is also taking part, focuses on collecting these sounds and then extracting information from them using “Machine Listening” techniques, also known as computer audition.
This is a technology from the field of Artificial Intelligence that works on and designs algorithms and models capable of extracting knowledge and relevant information from audio signals, specifically from environmental or generic sounds. This learning enables computers to carry out specific tasks autonomously, that is, without needing to be programmed.
SOROLL-IA, a project funded by IVACE and ERDF funds, will make it possible to improve various applications based on Artificial Intelligence (AI) systems, such as the identification of noise sources, with the aim of locating and acting on the noisiest and most disruptive elements in the environments where they occur.
The purpose of the project is to obtain data and structure it so that it is useful for the various tasks within the computer audition field. As Pedro Zuccarello, principal investigator of ITI's Soroll-IA project, explains, “to deploy a real machine listening solution, the first step is to build a database of sounds related to the context in which you are working”.
Industrial application
The creation of the Soroll-IA database will enable ITI to tackle future computer audition problems efficiently and effectively, particularly in the industrial environment.
In the field of safety and the identification of noise sources, this technology would be a key factor in protecting not only the health of workers, but also that of people and other living beings in the surroundings of the facilities.
Among the possible applications of sound recognition at industrial level, the use of acoustic sensors to carry out predictive maintenance also stands out. It will make it possible to detect problems effectively.
One of the main characteristics of industrial machines is that they make noise; analysing these sounds using AI can give us information about the condition of the machine and show whether it is drifting from its normal operation. Predictive maintenance is expected to eventually replace classic preventive maintenance, owing to the cost savings its use would bring. This initiative will enable, in subsequent projects, much more efficient and accurate development of R&D projects aimed at providing industrial solutions based on AI models within the area of computer audition, adds Pedro.
Finally, the SOROLL-IA project, file number IMDEEA/2022/68, is funded by the Valencian Institute for Business Competitiveness (IVACE) and the European Regional Development Fund (ERDF).
A short summary of new resources and the latest news from our editorial team
