SeaLice-3D is an AI-driven acoustic system that watches the water around your pens and works continuously to stop sea lice from ever attaching to your fish.
ASWES AquaIntel ApS is a Danish technology company developing SeaLice-3D — an acoustic, AI-driven prevention system for commercial open-water salmon farming. We work at the intersection of marine acoustics, computer vision, and aquaculture science to give farmers a way to stop sea lice pressure before it becomes a welfare or production problem.
Sea lice remain one of the most persistent and costly challenges in salmon farming — driving fish stress, treatment costs, and risk to wild salmonid populations in the surrounding fjords and coastlines. Most tools available today act after an infestation has already taken hold. We build systems that intervene earlier, while lice are still searching for a host rather than after they've found one.
We develop and operate the full system ourselves — the acoustic sensors, the AI verification models, and the deterrent hardware — and work directly with commercial salmon farmers and marine research institutions to test, validate, and refine it under real field conditions.
SeaLice-3D does not remove sea lice that have already attached to a fish. What it does is work continuously to prevent that attachment from happening in the first place — keeping lice away from the pen before they ever reach the salmon.
Around the clock, in the water surrounding the pen — not just at intervals.
Using AI to confirm what's actually a sea louse before anything reacts.
Deterring lice from reaching the fish — never treating them after the fact.
SeaLice-3D is built around a continuous three-part cycle that runs around the clock at every pen it protects.
Seabed-mounted sensors continuously scan the water surrounding the farming rings, watching for the presence of small particulate matter moving through the water column around the pen.
An onboard camera system, guided by AI image recognition trained specifically on sea lice, decides whether what's been detected is actually a sea louse — filtering out plankton, debris, and other harmless particles before any deterrent response is triggered.
Once the system decides lice are present, it responds with a gentle, non-harmful acoustic signal that discourages lice from approaching the pen — creating a protective barrier around the farm without any physical contact with the fish.
Sea lice reach a fish farm as free-swimming larvae, searching the surrounding water for a host to attach to. That searching phase — before attachment — is the only window where a farm can stop an infestation from starting at all.
SeaLice-3D is positioned at a distance from the farming rings, anchored to the seabed around the perimeter of the pen rather than on the nets themselves. This lets it monitor and respond across the water lice must pass through to reach the fish, deterring them during that searching phase — before they ever find a host.
Because the system runs continuously and reacts in real time, it functions purely as a preventive layer: it does not treat fish that are already infested, it works to stop that infestation from occurring in the first place. Over time, this is designed to reduce how often — and how severely — a farm needs to rely on delousing treatments.
The system operates entirely from the seabed perimeter. Nothing is attached to the nets or the fish themselves.
Detection and deterrence run continuously, not on a schedule — matching the pace at which lice pressure actually builds.
A supporting monitoring unit tracks current, temperature, and water conditions around the farm, helping optimize how the system is positioned at each site.
SeaLice-3D is being piloted at a commercial salmon farming site in the Faroe Islands. An initial observation phase mapping lice presence and local water conditions has been completed, with installation of the full detection and deterrence system planned for later in the year.
We're working alongside marine research institutions to ground our detection and deterrence approach in peer-reviewed fjord and sea lice science, helping tune the system to the specific conditions of Nordic and North Atlantic waters.
A patent application covering our detection and verification approach for particulate matter in seawater is currently pending, protecting the core methodology behind SeaLice-3D.
Lepeophtheirus salmonis and Caligus elongatus spend part of their life cycle as tiny, free-swimming zooplankton, drifting through the water column alongside countless harmless planktonic organisms. Detecting them means being able to continuously read the water column itself — not just check the surface.
Together with Akvaplan-niva and DTU Aqua, we've grounded our detection method in echosounder-based acoustic screening: pulses reflect off particulate matter suspended in the water, letting the system build a live, depth-resolved picture of zooplankton presence around the pen — rather than an occasional snapshot.
Acoustic backscatter alone can't tell a sea louse apart from other plankton, so an AI-driven camera system cross-checks what the echosounder detects. Trained specifically on sea lice morphology, it confirms both what an object is and precisely where it sits in the water column.
Only once a sea louse is identified and precisely located does the system respond — directing a low-frequency acoustic signal to that exact point in the water column, rather than running an undirected, constant barrier. It's a targeted response, built on knowing exactly where in the water the risk actually is.
Every echo, every image, and every verified detection is logged — building a continuous, high-resolution record of zooplankton and sea lice presence at each site, hour by hour, depth by depth. That dataset is, in itself, one of the most valuable outputs of the system: it's the ground truth our AI models train and improve on, and a longitudinal picture of lice pressure and water-column behaviour that neither a single sample nor a manual count could ever produce.
For our research partners, this turns every deployment into an ongoing field study. For farmers, it means decisions about a site are increasingly informed by real, site-specific data rather than general assumptions — and the more the system runs, the more valuable that data becomes.
We're working with commercial salmon farmers on new pilot sites. Get in touch to talk through your farm and where SeaLice-3D could fit.
jm@sealice3d.com