Launchpad Leaders: How AI sonar is Transforming Marine Monitoring
AI-powered sonar systems could change how marine environments are monitored, particularly in the turbid waters along the UK’s coastline.
PEBL, a community interest company (CIC), is working to unlock this potential with a goal to make marine monitoring more accessible and affordable for researchers, industry and coastal communities.
In conversation with Managing Director Christian Berfer, JOMEC students uncover how the technology works, the impact of the South West Wales Net Zero Industry Launchpad funding and what’s next for PEBL…
What is PEBL?
PEBL develops tools that monitor marine environments; we currently have two main products. The first is an underwater camera, similar to a GoPro, that can be programmed to switch on at specific times, tides or seasons — allowing long-term monitoring of changes underwater. The second is a sensor that measures water quality alongside factors such as temperature, and speed and direction of the current.
Alongside selling these tools, we use them to carry out biodiversity monitoring around the UK coastline — by combining video and sensor data, we can assess whether biodiversity is improving or declining. We also work closely with coastal communities, encouraging citizen science and supporting local conservation and restoration projects.
Why can’t traditional cameras always be used underwater?
In Welsh waters, visibility is often limited to two or three metres due to murky conditions. This makes cameras ineffective for monitoring marine developments, especially over long periods of time or during storms.
How does an AI-powered sonar system help you see underwater?
Instead of relying on light, sonar uses sound to ‘see’ underwater. It sends out sound waves that bounce off objects and return as echoes, which are then used to build up an image of what’s around it — similar to how an ultrasound scan works.
The device scans in 360 degrees and can be placed on the seabed to monitor changes over time. AI then helps interpret these grainy acoustic images, learning to recognise patterns and detect damage more quickly than traditional methods.
How does using AI change the way underwater data is analysed?
Interpreting acoustic images is similar to reading an X-ray — it usually requires a highly trained specialist, which is expensive and time-consuming.
If something goes wrong on a mussel farm, for example, operators can’t afford to wait weeks or months for expert analysis. AI allows much faster interpretation, enabling quicker decision making and responses.
How does this technology help protect marine environments?
The goal is to create an alert system that notifies operators when something changes. For example, if a buoy breaks loose, the system could alert the farmer so they can deal with it immediately.
Responding quickly helps reduce pollution, as lost equipment often ends up scattered along the coastline. It also helps protect marine habitats as when developments fail, anchors can drag across reefs or seagrass beds, whilst tangled ropes can pose a risk to dolphins and porpoises.
By identifying problems early, the damage can be limited, which has a positive impact on the wider marine ecosystem.
How has funding from the South West Wales Net Zero Industry Launchpad supported PEBL?
As a self-funded CIC without large investors, we wouldn’t normally be able to undertake research of this scale so the Launchpad funding has allowed us to explore this project using our existing skills in building cameras and sensors.
It’s also given us the opportunity to develop a potential new product that could benefit both industry and marine conservation.
How do you ensure that AI is used responsibly?
Our system isn’t public facing and operates in a very specific sector. The data stays on the device and with the developer, so privacy concerns are minimal.
More broadly, we believe AI should be used for public benefit — which means only using data we have explicit permission for or that we generate ourselves.
How do you see AI-powered sonar evolving in the future?
Cost is the biggest barrier. While these systems work well, they often cost tens of thousands of pounds, limiting their use.
Ideally, they would become affordable enough for local councils, or even available through community schemes — like borrowing one from a library. That would allow people to investigate concerns themselves, such as the impact of a ferry port on nearby seagrass beds. This kind of accessibility is the future of marine stewardship.
Could AI-powered sonar influence policy or conservation efforts?
I hope so. Strong acoustic evidence showing that certain marine developments are safe and well-managed could help shift public opinion and give policymakers more confidence in supporting sustainable projects.
What’s the next milestone for PEBL?
Our next big milestone is a full demonstration of the system for stakeholders. We’ve spoken to offshore wind farm developers and would like to show the technology working in real conditions, with clear images and reliable performance. Achieving that would be a major step forward.