Victoria’s MarineLabs has launched accurate and hyper-localized coastal weather forecasts delivered through MarineLabs’ new machine learning models.
Forecast AI is a much-anticipated addition to MarineLabs’ flagship product, CoastAware, a real-time weather intelligence solution that enhances safety, efficiency, and sustainability in maritime operations and bolsters the climate resilience of coastlines.
MarineLabs, a leader in cutting edge maritime weather intelligence technology, raised a $4.5 million seed funding round earlier this year, marking the largest investment of its kind in Canadian history for an ocean tech company.
Ports, harbour masters, and others in maritime industries rely on weather forecasts to make decisions that impact operations, crew and staff safety, costs, and flows of essential goods for consumers. However, there are challenges with the weather forecasting options available.
RELATED: MarineLabs is hiring on Techtalent.ca
Most weather models provide forecasts at vast global scales, relying on coarse grids to make the computations manageable, which limits their ability to account for local coastal features.
Using extensive data from their fleet of real-time sensor nodes, MarineLabs’ Forecast AI generates profoundly more accurate wind and wave forecasts that account for each location’s unique characteristics and geography.
“Forecast AI represents a giant stride forward in our mission to revolutionize marine mobility,” said Dr. Scott Beatty, CEO of MarineLabs.
“With CoastAware, we set a new standard in real-time weather awareness. Now, by leveraging our unique observations to train thousands of machine learning models, we have the ability to more accurately predict weather in a way that addresses critical gaps in traditional forecasts. This empowers our subscribers and partners with the precise, forward-looking, hyper-local wind and wave forecast information they need to make safe and informed decisions in the face of increasingly volatile weather.”
In an in-depth study over a two-month period from December 2023 to February 2024 at a site on the East coast of Canada, MarineLabs’ Forecast AI produced impressive results versus currently available models.
Breaking the test period into 182 eight-hour work shifts, Forecast AI made 23% more correct calls due to significant wave height exceeding critical safety thresholds for port operations than a trusted publicly available forecast model at a decision threshold time of 24 hours.
To date, in aggregate over 32 locations in MarineLabs’ sensor network with forecasting available, Forecast AI has 50% lower average errors in predicting weather parameters compared to publicly available maritime forecast models.
Access to this kind of tactical, highly reliable, hyper-local forecast data with customizable alerts specific to weather thresholds offers new efficiencies and levels of trust in decision-making for mariners and ports.
A total of 31 buoy locations from MarineLabs’ sensor network along Canadian coastlines in the Atlantic and Pacific now have Forecast AI, with an additional 12 sites in the US and Canada coming online throughout the fall.
Leave a Reply