Saving Lives with AI-Enhanced Noise Suppression

AI can do many things. By now, we are pretty much aware of its capabilities and potential. Still, when we hear about saving lives with AI-enhanced noise suppression, we are bound to feel amused! However, it is not a matter of imagination or speculative reality anymore. The latest research shows a novel AI-based noise suppression system for efficient victim detection by UAVs. The system proves especially useful in events of natural disasters. 

AI-Enhanced Noise Suppression: Enhancing the Value Proposition of UAVs

Unmanned Aerial Vehicles have proven their worth by showing results in search and rescue missions, which become critical and difficult to efficiently accomplish after large-scale natural disasters like earthquakes. In the wake of such calamities, one of the most obvious aftermaths of an earthquake is the volume of rubble generated.

However, UAVs, with their propellers, generate a considerable amount of sound. This sound, often perceived as noise, unfortunately, drowns out human sounds that might come from under the rubble, asking for help. Addressing this issue, the scientific innovation we are discussing here helps effectively suppress this UAV noise and amplify human sounds. The cornerstone of this solution is a really powerful AI model known as GANs or Generative Adversarial Networks. 

Click here for a list of the five best drone companies to invest in.

How do GANs work?

The AI model that GANs use can learn a range of data types with precision. This property or feature of the model was leveraged by the researchers/developers for the model to accurately learn the various types of UAV propeller sound data. 

After it learns these sounds, the learned model could generate a similar sound to that of the UAV propellers. This sound was called the pseudo-UAV sound. 

The final solution was to subtract this pseudo-UAV sound from the actual sound captured by the onboard microphones in the UAV. The result was in the form of the operator clearly and correctly hearing and recognizing human sounds coming from the surroundings.  

The technique and its development mechanism received much praise from the scientific community. It had several advantages over the traditional noise suppression system. 

For instance, it could efficiently suppress UAV noise within a narrow frequency range much more accurately. It also turned out to be agile enough to adapt to the UAV noise fluctuations in real time. 

Moreover, testing the solution showed impressive outcomes in the fact that it could considerably enhance human sound levels. While summarizing the benefits of the new solution and why it could be a boon for future search and rescue missions, Professor Chinthaka Premachandra from the Department of Electronic Engineering at the School of Engineering in Shibaura Institute of Technology has the following to say:

“Suppressing the UAV propeller noise from the sound mixture while enhancing the audibility of human voices presents a formidable research problem. The variable intensity of UAV noise, fluctuating unpredictably with different flight movements, complicates the development of a signal-processing filter capable of effectively removing UAV sound from the mixture. Our system utilizes AI to effectively recognize propeller sound and address these issues.”

A significant amount of research has been carried out in the area of noise suppression. One of the most notable research among them came from Loughborough University. Named Sonobex, this solution from the university addressed the health risks that environmental noise poses to us, including tinnitus and hearing loss, low blood pressure, and cardiovascular diseases. It also leads to cognitive impairment in children and disrupts our sleep, mental and social well-being. 

Click here to learn all about investing in artificial intelligence.

Sonobex: Noise Canceling Structures for Urban Environments

Sonobex addresses one of the core challenges that most environment noise canceling solutions face: how to reduce without prohibiting the free flow of air and light in the room. The research explored acoustic metamaterials, especially sonic crystals, to create efficient noise barriers by arranging them in arrays of cylindrical rods. 

The researchers could see that it was possible to vary the range of the sound frequency blocked by this method. This finding led to greater applicability of the solution. The sound barriers could be used in construction, industry, transport, and power generation.

Since the spacing was adjustable, one could allow enough space between the two barriers to facilitate the flow of air and light. The flow also created scope for air-cooling of the equipment.

Oftentimes, architectural solutions appear as a roadblock to achieving aesthetic superiority. This solution is free of that in the sense that it could be manufactured from a range of materials. 

The architects could explore diverse options. For instance, the noise suppressor could be used as a window covering. It came as a significant value-add as wind loading often posed problems for conventional solid noise barriers. They required considerable support. The gaps in this system and the wind permeability factor helped reduce this problem. 

The solution proved its worth beyond the laboratory scene by becoming functional, cost-effective, environment-friendly, and suitable for one’s health and well-being. 

In fact, Sonobex received many awards during its innovation lifespan. It received research funds from the East Midlands Development Agency, Engineering and Physical Sciences Research Council (EPSRC), Innovate UK, and the Royal Academy of Engineering.

The practicality of the innovation for real-life application was clearly demonstrated when, in January 2017, the Dutch Engineering and Manufacturing Company Merford acquired it. Adding to the product’s credibility, Sonobex secured the ISO 9001:2005 quality assurance certification, further solidifying its market position. As a testament to its growing acceptance and effectiveness, Sonobex’s NoiseTrap acoustic panels are now finding increasing use at various sites across Europe.

According to Joost Vertooren, the Director of Corporate Development, Merford:

“Sonobex offered exactly what Merford – with more than sixty years experience in noise control – was looking for.”

Companies like Merford that are active in the business of sound and noise reduction are continuously in search of solutions like Sonobex. Now, we look into such company and business-led noise suppression and cancellation products in the coming segments. 

#1. ThinkLabs’ ThinkBox: A Stethoscope that Can Listen to Sounds In the Most Noisiest Environment 

Stethoscopes are vital in diagnosing a patient at the right time in the most time- and resource-efficient way possible. However, in high-noise environments, in a makeshift treatment center in a battleground, for instance, or on an airplane, it becomes difficult for the medical professional to hear the patient’s bodily sounds. 

ThinkLabs, a well-known manufacturer of stethoscopes, solved this problem by repurposing the stethoscope technology in a highly advanced way. The device, among other features, comes with a filtering technology that involves five different filters. It can efficiently filter out certain sounds and leave only what is required to be heard. 

According to Charles Harding, a biomedical engineer involved in the project: 

“Of the five filters, there are two heart filters, which work to isolate only heart sounds; two lung filters, which target the inhale and exhale movement of air sounds; and a broadband filter, which allows medical crews to record sounds for second opinion diagnoses as well as training.”

The broadband filter serves three purposes: telemedicine, telemonitoring, and telementoring. Through the stethoscope, it is possible to record the sound and send it off to another medical professional to get their assistance or advice. The device has been helping the Department of Defense aeromedical professionals to more effectively treat injured warfighters. 

Based in Centennial, Colorado, United States, Thinklabs is a private entity established in 1991 by Clive Smith. The company’s estimated annual revenue is around US$16 million.

#2. Qualcomm Proprietary Echo Canceling and Noise Suppression Technology

Qualcomm’s cVc Echo Cancelling and Noise Suppression Technology proves effective when making a voice call using a Bluetooth Stereo Headset or earbuds. It transmits crystal clear voice through its clear voice capture (CVC) technology by negating the ambient noise, road noise, and noise generated by people talking nearby. 

Qualcomm’s cVc technology is a suite of algorithms that work on the transmission and receiving path of voice calls and deliver optimum voice quality irrespective of the location of the receiver. Its echo-canceling property ensures that the acoustic echo is reduced for a more natural two-way conversation.

The solution also ensures a consistent voice quality, overcoming the challenges of packet loss and bit error concealment. It ensures that compromised RF connectivity does not come in the way of the voice quality. It has several other features that enhance user experience, including automatic gain controls, send/receive equalizers, comfort noise properties, howling control, nonlinear processing capabilities, adaptive equalizers, auxiliary stream mixing, frequency-enhanced speech integrability, and much more. 

As per the latest financial data available, Qualcomm earned US$9.9 billion in GAAP revenue in the quarter ended on December 24th, 2023. The GAAP EPS stood at US$2.46. 

#3. Silentium’s Active Noise Control Technology (ANC) 

Silentium is known as a high-end innovator in the active noise cancellation technology research space. Its solutions are capable of cutting out up to 90% of a sound, irrespective of its source, which could be an oven hood, a ventilation system, a car engine, or a noisy train. 

The Active Noise Control technology developed by Silentium is a disruptive sound management solution. Its algorithms can adaptively follow changes in the noise spectrum on a real-time basis. 

The final solution is small in size, low cost, and highly effective for low/medium-frequency noise. It is capable of producing successful commercial solutions across many applications, including home, office, and industrial workplaces. It can cater to the needs of both the automotive and transportation industries. 

The solution is unique in its delivery mechanism in the sense that it can reduce the active noise either from the source or by creating personal zones of quiet. These zones are called Quiet bubbles. 

As per the latest available information, Silentium is continuously scaling up the development and roll-out levels of its Active Acoustics software. 

The company closed its Series D funding round at around US$63 million in total. Of this US$63 million, US$20 million came as part of its extended Series D round, which was led by Menora Mivtachmin and Meitav Dash. 

Global electronics leader Molex was also one of the investors involved. While speaking about the investment influx, Yoel Noar, the CEO of Silentium, said: 

“We welcome new partners that share the same passion for acoustics and see the value in enhancing privacy and wellbeing through personal sound spaces. With this new investment, we will scale up the development and roll-out of our Active Acoustics software solutions for vehicle manufacturers and automotive suppliers – especially as NVH requirements evolve in the new era of electric and autonomous mobility.”

Arresting the Menace of Noise

The European Environment Agency (EEA) has recognized noise pollution as a major problem, both for human health and the environment. EEA’s report on environmental noise shows that twenty percent of Europe’s population is exposed to long-term noise levels that are harmful to their health. 

This translates to more than 100 million people on the continent. The report says that long-term noise exposure can result in a range of health effects, including negative effects on the cardiovascular and metabolic systems. 

As per estimates, environmental noise is responsible for 48,000 new cases of ischaemic heart disease a year, as well as 12,000 premature deaths. It also leads to as many as 22 million people suffering from chronic high annoyance and another 6.5 million from chronic high sleep disturbance. It is astounding to find that 12,500 schoolchildren suffer from reading impairment in school as a result of aircraft noise. 

According to studies done by the World Health Organization, noise is the second largest environmental cause of health problems, a close second to air pollution. 

Safeguarding our population from the harmful effects of noise pollution is a necessity we will deny at our peril. There is an urgent need to strengthen noise-suppression research by creating synergies with other developing technological paradigms like AI. The good news is that significant efforts are already being taken in that direction.