Live Demo · Open Access

See the platform live, right now

A working build of the OceanStar monitoring platform — seven simulated Canadian facilities, real outdoor weather, and a live AI prediction engine. No login, no waiting.

7 facilities across Canada

Atlantic Salmon · Pictou · Rainbow Trout · Winnipeg · Arctic Char · Iqaluit · White Shrimp · Calgary · Snow Crab · Bonavista · Nile Tilapia · Brampton · Barramundi · Vancouver

Live outdoor weather

Environment Canada / Open-Meteo data pulled in real time, translated into building heat-load and freeze-risk metrics for each site.

AI reasoning chain

Every prediction traces back to a published biological tolerance rule with mechanism, severity, and reversibility. No black box.

26 trained ML models

XGBoost forecasters (1-hour and 6-hour horizons) and Isolation Forest anomaly detectors trained on facility-specific synthetic data.

Pick your entry point

Three ways into the live platform

Start here01

National Operations Centre

Live fleet view across seven Canadian land-based aquaculture facilities. Real-time outdoor weather, AI risk classification, and cross-facility alerts.

Launch
AI reasoning02

Per-species facility dashboard

Open the Atlantic Salmon site in Pictou, Nova Scotia. Live sensor channels, AI reasoning chain, weather → habitat impact, and 1-hour forecasts.

Launch
Interpretability03

Species biology reference

Every physiological tolerance range, biological outcome rule, and combined-stressor cascade the model uses to make predictions — fully auditable.

Launch

Opens in a new tab · Data refreshes live · No account required

What you'll see

Six water-quality parameters, seven species, one live system

●Live sensor channels: Temperature, dissolved oxygen, pH, conductivity, turbidity, salinity — refreshing every five seconds.
●Live outdoor weather: Real Open-Meteo data for each Canadian city, mapped to facility heater load and freeze risk.
●AI reasoning chain: Every active prediction shown as sensor → forecast → biology rule → outcome, with physiological mechanism.
●Forecast confidence: Every XGBoost prediction includes the model R² so you can judge reliability for each parameter.
●Combined-stressor cascades: When two parameters drift together (e.g. warm water + low DO), severity multipliers trigger compound alerts.
●Species biology library: Optimal / stress / lethal ranges with sources for every rule the model uses — fully auditable.

Jump directly to a species

All seven facilities are running live

Atlantic Salmon · Pictou, NS→Rainbow Trout · Winnipeg, MB→Arctic Char · Iqaluit, NU→White Shrimp · Calgary, AB→Snow Crab · Bonavista, NL→Nile Tilapia · Brampton, ON→Barramundi · Vancouver, BC→

Demonstration data: Sensor readings are generated by a physically-modelled simulator (seasonal, diurnal, feeding-driven, with injected failure events). All weather is real and pulled live from Environment Canada via Open-Meteo. The AI reasoning, models, and biology rules are identical to what runs against real sensor feeds.

Ready to deploy this in your facility?

The live demo shows the software stack. Pair it with the OceanStar Hub + sensor suite to get the same view on real tanks. We'll walk you through the deployment in a 30-minute call.

Book a deployment call

Ready to Transform Your RAS Operation?

Join the next generation of aquaculture operators using AI-powered monitoring to protect their fish and their bottom line.

Request a Demo →