How Artificial Intelligence is Fighting Noise Pollution in Ireland

In the bustling metropolises of the world, where honking cars, roaring aeroplanes, and thunderous trains reign supreme, a silent invader lurks. Noise pollution, often overlooked and underestimated, is gradually emerging as a critical concern affecting our well-being and the environment. If we could visualise the sound we generate, it would resemble the mountains of litter that plague our cities. We’d see landfills of harsh sounds piled up to the clouds.  Camouflaged amidst chaotic urban life, noise pollution is causing severe harm. Although the problem of noise pollution is no easy fix, AI is a smart solution that we believe is a game-changer.


Noise Pollution:The Deafening Truth

Hearing loss, increased cardiovascular risks, mental health disorders, and developmental issues in children are just some of the negative health impacts caused by noise pollution. This is a problem all across Europe. While the World Health Organisations’ safety threshold for sound is 40dB, noise in Dublin exceeds 55 decibels at night and surpasses 70 decibels during the day. And, in London, residents are exposed to average levels of 86 decibels

Bruitparif, a non-profit organisation dedicated to combatting noise pollution in Paris, unveiled a groundbreaking report that paints a sobering portrait of the detrimental effects of noise pollution. This report combined medical projections from the World Health Organisation with “noise maps” based on data collected from its extensive network of acoustic sensors. The findings were staggering, revealing that residents living in the loudest areas of Île-de-France, encompassing Paris and its suburbs, lose more than three healthy life years over a lifetime due to ailments caused or worsened by the relentless noise generated by cars, trucks, aeroplanes, and trains.


The Earths Noisiest Neighbour: Airports 

One of the focal points for noise pollution mitigation efforts is airports. The incessant roar of jet engines during takeoff and landing can be highly disruptive and detrimental to the well-being of nearby residents. So much so that the economic cost of health damage caused by night flights alone is an estimated 36,000 euros per flight.

Amidst the global cacophony generated by more than 40,000 airports worldwide, it is no surprise that noise analysts play a vital role in monitoring and mitigating the adverse effects of airport sounds. However, the sheer volume of data they handle is overwhelming. Currently, these dedicated professionals painstakingly listen to hours of recordings and manually report on them. For instance, a mere 250-minute audio recording necessitates 6 to 8 hours of arduous manual processing, and even then, the process remains error-prone. 

With the number of flights growing year-by-year, how we tackle noise pollution needs to change.

Our Solution: AI-powered Sound Monitoring 

AI-powered systems can be deployed to monitor and analyse noise levels in real-time. By leveraging machine learning algorithms, these systems can distinguish between different types of sounds, including aircraft noise, and provide valuable insights into the impact on local communities. Such data empowers policymakers, urban planners, and airport authorities to make informed decisions regarding flight paths, noise reduction measures, and residential zoning to mitigate the harmful effects on public health. 

At Gemmo, our commitment to tackling noise pollution extends to the development of a comprehensive model zoo. This repository houses a wide range of pre-trained machine-learning models, built on diverse datasets, to address various sound analysis challenges. One particular gem in our collection is our specially designed model focused on aviation noises.  This specialised model encompasses a vast collection of sounds, ranging from the roar of jet engines during takeoff to the high-pitched whine of landing gears.

Towards a Quieter, Healthier Future

Our machine listening technology is key to the systematic, constant monitoring of noise pollution. By connecting our algorithms to a network of sensors, we can continuously gather data on noise levels.  This allows us to track noise patterns, identify sources of noise pollution, and assess the impact on surrounding communities. With this comprehensive and real-time information, stakeholders can take effective, information-driven action for noise mitigation. AI is the way toward a quieter, healthier future. 

By Johanna Walsh

Start your Ai journey today!

Gemmo's noise classification case study with Sonitus