Underwater and surface cameras track fish populations, feeding behavior, and health indicators across ocean pens and inland tanks. Species-specific models for tuna, tilapia, salmon, and more.
From Lab to Ocean - The same rigor that measures mouse behavior — now measuring fish welfare.
ConductScience has spent over a decade building automated behavioral analysis for research organisms. ConductVision extends that expertise to aquaculture — where population estimation, feeding optimization, and mortality detection directly impact yield and sustainability.
Pen APen BSalmon 94%Salmon 91%S-01S-02S-03
Population Estimation
Automated fish counting using underwater cameras. Estimate biomass, stocking density, and population changes without manual sampling.
Pen APen BSalmon 94%Salmon 91%S-01S-02S-03
Feeding Optimization
Monitor feeding intensity in real-time. Detect appetite levels, optimize feed conversion ratios, and reduce waste by stopping feeds at satiation.
Pen APen BSalmon 94%Salmon 91%S-01S-02S-03
Health & Mortality
Detect dead fish, abnormal swim patterns, and early disease indicators. Alert operators before problems spread through the population.
Capabilities - What ConductVision sees beneath the surface.
Purpose-built models for underwater and surface aquaculture environments — not repurposed terrestrial detection.
Population estimation (counting)
Frame-by-frame fish counting with species classification across multiple pens and tanks
Biomass estimation (size + density)
Combine individual size estimates with density measurements for total biomass tracking
Feeding intensity scoring
Real-time appetite scoring based on surface activity and feeding behavior patterns
Feed conversion ratio tracking
Correlate feeding events with growth metrics to optimize feed utilization
Mortality detection (dead fish)
Automatic detection and counting of dead fish for rapid response
Abnormal swim pattern detection
Identify erratic swimming, listing, or lethargy that may indicate disease or stress
Size distribution analysis
Track growth rates and size uniformity across cohorts over time
Stocking density monitoring
Continuous density estimates to maintain optimal conditions per species
Use Cases - Built for every aquaculture environment.
Pen APen BSalmon 94%Salmon 91%S-01S-02S-03
Ocean Net Pens
Population estimation across multiple pens
Feeding behavior monitoring from surface cameras
Mortality alerts for rapid response teams
Net integrity monitoring via camera feeds
Pen APen BSalmon 94%Salmon 91%S-01S-02S-03
Inland RAS Facilities
Tank-level population and density tracking
Growth rate monitoring across cohorts
Automated feeding optimization per tank
Water clarity scoring from camera feeds
Pen APen BSalmon 94%Salmon 91%S-01S-02S-03
Research & Breeding
High-throughput phenotyping for selective breeding
Behavioral response studies under varying conditions
Longitudinal growth tracking with frame-level data
Automated welfare scoring for regulatory compliance
How it works - From camera to insight in three steps.
1
Capture
Deploy underwater and surface cameras in pens, tanks, or raceways. Works with standard aquaculture camera systems.
2
Analyze
ConductVision's aquaculture models detect fish, classify species, score feeding behavior, and flag health anomalies frame by frame.
3
Act
Get real-time dashboards, feeding alerts, and health reports. Integrate with your farm management system.
Why ConductVision - Built by scientists, for aquaculture professionals.
A decade of behavioral research heritage. ConductScience has equipped research labs worldwide with automated behavioral analysis. ConductVision adapts that same precision to aquaculture species and environments.
Species-specific models. Not a generic fish detector. Models trained per species — salmon, tilapia, tuna, trout, sea bass, shrimp — with species-appropriate behavioral baselines.
Underwater-first design. Handles the unique challenges of underwater vision: variable lighting, turbidity, refraction, and dense schooling behavior.
Outputs - Data your team can act on with confidence.
Converts underwater and surface video into structured production data that integrates with your farm management workflows.
Per-pen population and biomass estimates
Feeding intensity timelines with appetite scoring
Mortality detection logs with camera evidence
Growth rate tracking and size distribution curves
Feed conversion ratio calculations
CSV, Excel, and API exports for management software
Deployment - Fits into your existing aquaculture infrastructure.
Designed for aquaculture environments — works with your existing camera systems, runs on standard hardware, and integrates with farm management platforms.
Camera assessment and underwater deployment planning
Species model selection and calibration
Baseline measurement with your stock
Staff training and dashboard setup
Ongoing support and model refinement
Our Commitment - We measure fish behavior the way scientists do. Because we are scientists.
ConductVision was born in the research lab. Every metric we report — from population counts to feeding scores — is held to the same standard as a peer-reviewed publication. That rigor is what makes the difference in production aquaculture.
FAQ - Questions we hear from Aquaculture teams
Can't find the answer you're looking for? Reach out to our support team.
We have species-specific models for Atlantic salmon, tilapia, bluefin tuna, rainbow trout, sea bass, and shrimp. Additional species can be added through our model training pipeline.
ConductVision works with standard aquaculture camera systems. We can advise on optimal camera placement and specifications for your specific environment.
Accuracy depends on species, density, and water conditions. In typical net pen conditions with moderate stocking density, our models achieve population estimates within 5-10% of manual counts.
ConductVision detects behavioral indicators of disease — abnormal swim patterns, lethargy, surface gasping, and mortality. It is designed as an early warning system, not a diagnostic tool. Consult your aquaculture veterinarian for diagnosis.
The system scores feeding intensity in real-time by analyzing surface activity and fish movement. It can signal feed operators when appetite drops, reducing waste and improving feed conversion ratios.
Start with a conversation about your operation — species, environment, camera setup, and goals. We will design a pilot program tailored to your facility.
Still have questions? We are happy to talk it through.