Listening to reefs: How Reef Pulse is using AI and underwater sound to protect coral ecosystems.

In this episode of BlueTech Around the World, we travel to one of the most fragile and fascinating frontiers of ocean innovation: coral reefs.

They cover less than 0.1% of the ocean surface, yet they support an extraordinary share of marine biodiversity, protect coastlines, sustain fisheries, fuel tourism economies, and provide vital ecosystem services to millions of people worldwide. And still, they are disappearing before our eyes.

Under the combined pressure of climate change, heatwaves, pollution, acidification, overfishing, and coastal development, coral reefs are under siege. In some places, the collapse is no longer a distant scenario. It is already underway.

But while the crisis deepens, new tools are emerging.

My guest in this episode is Yann Bayle, PhD in artificial intelligence applied to sound analysis, co-founder and CTO of Reef Pulse, a pioneering company based in Réunion Island that quite literally listens to the ocean.

Together, we explored a powerful idea: what if we could monitor reef health not only by looking at it, but by hearing it? What if sound could become both a diagnostic tool and a restoration lever?

The result is one of the most original and inspiring conversations I have had on the podcast so far: a dialogue at the crossroads of AI, coral ecology, acoustics, entrepreneurship, and ocean regeneration.

A startup born between science, island roots, and a shared mission.

The story of Reef Pulse begins on Réunion Island, in the Indian Ocean.

For Yann, the personal connection runs deep. Born and raised on the island, he left to study in mainland France and later specialized in artificial intelligence applied to sound analysis, with a background spanning audio signal processing, music analysis, and computer science.

His future co-founder, Simon, came from another world: marine biology.

Simon had completed a PhD between 2016 and 2019 on passive acoustic monitoring applied to coral reef soundscapes, recording underwater environments across different regions, from Réunion to the Pacific. But instead of keeping this work confined to research papers and laboratory archives, he wanted to turn it into something more operational, more useful, more deployable in the real world.

That is where the complementarity clicked.

As Yann explained, “He’s a biologist… I got a PhD in computer science. So we just had to work together to create something great, something bigger than both of us.”

And that is how Reef Pulse was born in 2021.

Like many ocean startups, the company emerged from a familiar but difficult transition: moving from scientific research to entrepreneurship. Reef Pulse did not simply have to build a technology. It also had to learn how to structure a business, hire a team, develop services, find clients, and create a viable business model, all while staying true to its ecological mission.

It is a path many researchers hesitate to take. And that is exactly why this story matters.

First things first: what is a coral reef, really?

Before talking about AI, algorithms, and hydrophones, this episode takes the time to go back to basics, something I find essential in these conversations, because not everyone listening comes from a marine science background.

So let’s start there.

A coral is an animal.

That is the first thing Yann wanted to make clear.

As he put it very simply: “A coral is an animal. So that’s the basic.”

But when we talk about coral reefs, we are not just talking about individual corals. We are talking about an entire ecosystem. A reef is a living underwater city, inhabited by fish, invertebrates, crustaceans, algae, and countless other organisms interacting in complex ways.

Yann used a comparison I loved: it is like the difference between a tree and a forest.

Corals themselves live in symbiosis with microscopic algae, which provide them with nutrients and contribute to their bright colors. When ocean temperatures rise too much, this relationship breaks down. The algae leave. The coral turns white. That is coral bleaching.

And if the stress lasts too long, the coral may die.

There are some nuances, of course. Some species can survive without algae for a short period. Some can recover if stressful conditions ease quickly enough. But when heatwaves repeat too often or last too long, their resilience reaches its limits.

That is where we are increasingly heading.

More than biodiversity: why coral reefs matter to people too.

One of the strengths of this conversation is that it never treats coral reefs as an abstract environmental issue. Yann keeps bringing the discussion back to what reefs mean not only for marine life, but for human societies too.

And the numbers are staggering.

According to figures he referenced, up to one billion people globally depend on coral reefs for food supply and livelihoods.

Reef ecosystems support fisheries, help stabilize coastlines, and underpin tourism economies in many tropical islands and coastal nations.

As Yann reminded us, their value cannot be reduced to a balance sheet. Yes, there is food. Yes, there is tourism. Yes, there is coastal protection. But there is also something else: beauty, wonder, and the irreplaceable richness of living systems.

“They are beautiful and it would be a shame to see them disappear,” he told me.And honestly, that sentence says a lot.

Because there is something deeply moving about coral reefs. They are not only useful. They are extraordinary.

And that is part of why they are worth fighting for.

Do we still have time to act?

This is one of the hardest questions in any conversation on coral reefs today.

We hear alarming figures everywhere. Up to 90% of coral reefs could disappear by 2050. Ocean temperatures keep rising. Ocean acidification has crossed another worrying threshold. Severe bleaching events are becoming more frequent.

So naturally, people ask: is it already too late?

Yann’s answer was honest, but not hopeless.

Coral reefs do have some resilience. They have survived disturbances before. But the real danger today is the accumulation of threats.

Heatwaves do not come alone. They can be followed by cyclones. Pollution adds pressure. Overfishing weakens the ecosystem further. Recovery windows shrink. The reef is not only stressed, it is attacked from multiple directions at once.

Yann gave the example of Mayotte, where bleaching and cyclone impacts hit in close succession, with devastating consequences. In some areas, the damage is already so severe that full recovery looks uncertain.

And yet, he refuses fatalism.

“If we don’t act today, it’s obvious that we can’t go anywhere with that,” he said.

But he also insisted: “We have to fight for them.”

That balance, realism without resignation, runs throughout the whole episode.

From visual surveys to soundscapes: how coral reefs are usually monitored

To understand what makes Reef Pulse different, we first explored how coral reefs are traditionally monitored.

Historically, the answer was simple: divers go underwater and look.

That alone was already a big step. Underwater exploration is relatively recent in human history. But visual surveys come with many limitations. They are episodic, subjective, intrusive, and highly dependent on the diver’s experience.

As Yann explained, one diver may estimate fish size differently from another. One may notice more than another. Some species are nocturnal. Others are hidden. Some behaviors only happen at dawn, dusk, or during very short seasonal windows. If you are not there at the right moment, you miss them.

And there are other tools too:

• satellite monitoring, useful for broad observations like bleaching but limited by clouds, turbidity, and depth;

• eDNA, which can detect species from water samples but may reflect organisms from far away because of currents;

• and, increasingly, acoustic monitoring, which adds another layer of understanding.

Yann was very clear that these methods are not in competition. They are complementary.

That matters. Reef Pulse is not claiming that sound alone can explain everything. What it offers is something different: a way to capture continuous, non-intrusive, time-rich information that other tools may miss.

Bioacoustics or ecoacoustics? From one musician to the whole orchestra.

One of the most fascinating moments in the episode was when Yann explained the difference between bioacoustics and ecoacoustics.

Bioacoustics focuses on one source. One species. One signal. One sound-producing animal at a time.

Ecoacoustics takes a broader approach. It listens to the entire soundscape of an ecosystem.

Yann’s metaphor was perfect: “Bioacoustics is just one musician and ecoacoustics is the whole orchestra.”

And coral reefs are noisy places.

You hear the crackling of snapping shrimp, the grazing of parrotfish, the vocalizations of groupers, the background of waves and currents, the sound of rain or thunder, and increasingly, the presence of human activities too: boats, engines, infrastructure.

Reef Pulse works at the level of that full acoustic tapestry.

The idea is not always to identify every single sound source individually. Sometimes the value comes from the bigger picture: the volume, diversity, frequency balance, daily rhythm, and long-term patterns of the reef soundscape.

A healthy reef sounds different from a degraded one. And in many cases, that difference is measurable.

What a reef sounds like when it is alive… and when it is fading.

During the episode, we discussed audio examples from Reef Pulse’s work, comparing healthier and more degraded reefs.

In broad terms, healthier reefs tend to have a richer, denser, more diverse soundscape. You hear more variety, more activity, more layers in the acoustic environment.

Degraded reefs, by contrast, often sound flatter, poorer, less balanced.

As Yann explained, in some degraded states the higher frequencies from snapping shrimp may become dominant, while lower-frequency fish activity decreases.

In his words, “The ecosystem is really doing better” when you hear more acoustic diversity and a more balanced soundscape.

This is where sound becomes more than a curiosity. It becomes a form of ecological intelligence.

It reveals hidden activity. It captures species we do not see. It records the rhythm of life at times when no diver is present. It creates a memory of the ecosystem that can be reanalyzed later as new algorithms emerge.

And that is one of Reef Pulse’s major strengths.

Data is gold: why archived sound matters

If there is one thing Yann insisted on repeatedly, it is this: data is gold.

Reef Pulse does not only monitor in real time. It also archives sound recordings, which means old datasets can be revisited with new algorithms later on.

This is a huge advantage.

A recording captured ten years ago may suddenly reveal new information once a better detection model becomes available. A species that could not be identified before might become recognizable. A new ecological indicator might be extracted from an old sound file.

As Yann put it, “We can still analyze the data afterwards. Whereas with visual survey, we cannot.”

This archival dimension gives acoustic monitoring a long-term value that is often underestimated. The reef is not only being observed. It is being documented in a form that remains scientifically alive.

And as with many AI applications, the true differentiator is not only the model itself. It is the combination of data + labels + expertise.

Reef Pulse has spent hundreds of hours listening, annotating, tagging, and structuring recordings. That work is invisible from the outside. But it is what makes the technology meaningful.

How Reef Pulse works in practice: from underwater tripods to stakeholder dashboards.

So what does the operational chain actually look like?

It starts with the client.

Reef Pulse first discusses needs: what area should be monitored, what the priorities are, what questions the client wants to answer, and what budget is available. Then the team selects sites and deploys its equipment.

That equipment includes:

• a tripod,

• a recorder with batteries and electronic board,

• an SD card with very large storage capacity,

• a cable,

• and a hydrophone, which is essentially an underwater microphone.

The system is deployed underwater for weeks or months. Then the data is collected, uploaded to Reef Pulse servers, cleaned, processed, and analyzed. The team filters out anthropophony and geophony when needed, isolates relevant biological signals, computes ecoacoustic metrics, and makes results available through a dashboard.

Clients can then explore maps, time series, trends, sound pressure levels, species indicators, and correlations with environmental variables such as temperature.

One example Yann gave was particularly striking: parrotfish grazing.

Because parrotfish feed on algae, their activity is closely linked to sunlight and reef functioning. Acoustic monitoring can help detect where this grazing activity is stronger, which provides useful insight into herbivory dynamics, reef resilience, and possible management actions.

From there, the applications multiply.

From whales to fish spawning: the many things sound can reveal

What can stakeholders actually learn from reef soundscapes?

The answer is: quite a lot.

Depending on the project, Reef Pulse can provide insights on:

• key species presence,

• underwater noise pollution,

• biological activity patterns,

• seasonal changes,

• whale arrivals,

• spawning periods,

• and even reproductive windows that deserve temporary protection measures.

Yann shared the example of a fish species whose vocalization peaks at the same time each year. This kind of information can help define temporary no-go zones or targeted protection windows, in a way that is useful not only for conservation but also for fisheries.

If fish populations are allowed to reproduce successfully, everyone benefits in the long term.

The same logic applies to whale watching. Acoustic data can detect whale presence earlier than visual methods, helping tourism operators or regulators adjust seasons more accurately.

That is what I find particularly compelling in Reef Pulse’s approach: the data is not abstract. It can lead to very concrete decisions.

Noise pollution: the threat coral reefs hear but law still overlooks.

One of the most important themes in the episode was underwater noise pollution.

We often talk about noise in relation to marine mammals, offshore construction, or shipping lanes. Much less often do we discuss its impact on coral reef ecosystems. Yet the evidence is growing.

As Yann explained, excessive sound can alter fish behavior, reduce feeding efficiency, disrupt predation and mating, cause animals to hide, and in extreme cases even injure or kill marine life.

And yet, around coral reefs, legal protection remains very limited.

That is why Reef Pulse’s role increasingly goes beyond data production. The company is also involved in awareness-raising and policy discussions, trying to push for more robust frameworks.

It is a reminder that ocean innovation is not only about creating tools. Sometimes it also means creating the conditions for those tools to matter.

A new frontier: acoustic restoration.

And then comes the part of the story that feels almost magical.

Because Reef Pulse does not only listen to reefs. It now also works on restoration through sound.

The principle is based on an increasingly well-documented ecological insight: coral larvae and fish larvae do not drift entirely at random. They use environmental cues to choose where to settle. And one of those cues is sound.

Healthy reefs are noisy. They crackle, grunt, pulse, and hum with life. Degraded reefs are often quieter.

So what happens if you play back the sound of a healthy reef in a degraded area?

That is the premise behind Reef Pulse’s restoration solution. The company deploys underwater speakers that broadcast ambient sound recorded from healthier ecosystems, with the aim of attracting larvae and encouraging settlement.

Yann described it beautifully: the speaker acts almost like a beacon.

Or, as I told him during the conversation, like a kind of underwater GPS signal saying:here is the party, here is life, come settle here.

The image may sound playful, but the science behind it is serious. Previous studies have shown significant increases in larval recruitment after relatively short periods of acoustic enrichment.

From proof of concept to field-scale deployment in the Indian Ocean.

What makes Reef Pulse especially interesting here is that it is not stopping at a lab demonstration.

The company is now pushing this work toward real-world deployment, notably through the REACT program in Réunion Island, with the ambition of becoming the first to test such restoration at scale in the Indian Ocean.

And not in isolation.

This project brings together Reef Pulse, university researchers, local companies, scientific partners, and other expertise across fields such as sound propagation, underwater operations, photogrammetry, and ecological monitoring.

That collaborative dimension is central. Acoustic restoration is not a one-size-fits-all miracle fix. It is a complementary lever that can strengthen existing coral restoration projects, nurseries, and field efforts.

As Yann explained, the goal is also to better define the conditions under which acoustic enrichment works best:

• how degraded can a reef be and still respond,

• what noise thresholds matter,

• how long the intervention should last,

• and what effects can be observed over longer timescales.

That last point is crucial. Previous studies often looked at a few weeks or months. Reef Pulse wants to understand what happens over years.

Across three oceans: building a global library of reef sound.

Another major milestone for the company is its geographic expansion.

Reef Pulse has already deployed or worked on projects across three ocean basins, including Réunion, Mayotte, Guadeloupe, and now New Caledonia through a major project supported by BestLife2030, the IUCN, and the Parc naturel de la mer de Corail.

This matters scientifically and strategically.

Different reef regions have different sound signatures, different species, different acoustic communities. Some elements are shared. Others are highly specific. By collecting data across oceans, Reef Pulse is building something rare and valuable: a global library of underwater biological sound.

That dataset is not just useful for Reef Pulse. It has the potential to support a broader shift in how reef ecosystems are monitored and understood worldwide.

The business model: from B2G foundations to international expansion.

Reef Pulse began with a B2G model  (business to government) which made sense given its origins in public funding, public research, and ecosystem monitoring missions of collective interest.

As Yann put it, “The sea belongs to everyone.”

That public dimension remains central. But the company is now broadening its scope.

Its future clients may include:

• governments and marine protected areas,

• research labs,

• resorts and tourism actors,

• infrastructure developers,

• and organizations needing environmental baseline or impact monitoring.

That diversification is smart. It reflects a wider truth in blue tech entrepreneurship: environmental missions and business sustainability are not opposites, but they do require thoughtful positioning.

What I appreciated in Yann’s explanation is how gradual and disciplined the approach is. Start with one client. Validate the proof of concept. Add new use cases. Expand step by step.

As he said, it is an agile method.

And notably, the company name is Reef Pulse, not Coral Pulse. That is a clue. The long-term vision may eventually extend beyond coral reefs to other reef ecosystems too.

Protecting nature without ignoring the footprint of technology.

One question I really wanted to ask Yann was about something many ocean tech conversations still avoid: the footprint of the technology itself.

AI consumes energy. Servers consume energy. Data transfers consume energy. Hardware requires materials. So how do you build a company that aims to protect ecosystems without blindly increasing its own environmental burden?

I was impressed by how seriously Reef Pulse takes this.

The company works to optimize model size and energy performance, tracks electricity use, prioritizes local servers where possible, limits unnecessary bandwidth-heavy workflows, and chooses repairable equipment and more responsible materials whenever possible.

Some of this is invisible to the client. But it is part of the company’s DNA.

As Yann made clear, the goal is not just to help coral reefs by any means. It is to do so without worsening the situation elsewhere.

That coherence matters.

What comes next: from digital twins to intelligent ocean monitoring.

At the end of our conversation, I asked Yann what technologies he thinks will reshape coral reef protection in the next five to ten years.

His answer was not about one single miracle innovation. It was about combination.

Better algorithms. Better hydrophones. Lower-energy AI. More intelligent drones. More connected sensors. More integration across acoustics, video, satellite data, and environmental data.

In short: a move toward something like an ecosystem digital twin, where multiple technologies work together to create a richer and more actionable picture of the underwater world.

That vision is ambitious. But after listening to everything Reef Pulse has already built, it no longer sounds far-fetched.

The bigger picture: innovation that gives nature a voice.

What stayed with me most after this episode is something I said to Yann near the end:

innovation is not just about technology. It is about giving nature a voice, and listening carefully to what it tells us.

That is exactly what Reef Pulse does.

It listens.

It translates hidden signals into ecological intelligence.

It helps scientists, governments, and coastal stakeholders understand what is happening below the surface.

And now, with acoustic restoration, it is even trying to use sound not only to observe the reef, but to help life return.

In a time when coral reef conversations can easily become overwhelming, tragic, or paralyzing, I found this episode deeply energizing.

Because it reminds us that there is still room for action.Still room for invention.Still room for collaboration.

And perhaps that is the true pulse of this story.

Listen to the full episode

Listen to the full conversation with Yann Bayle from Reef Pulse on your favorite podcast platform, or watch the episode on YouTube.

To learn more about Reef Pulse and follow their work in coral reef monitoring and restoration, stay connected through their channels.

If you enjoyed this story, subscribe to BlueTech Around the World for more conversations at the intersection of ocean science, entrepreneurship, innovation, and regeneration.