FAST-CAST 2 — Calculation of optimal shipping routes through polar waters based on AI supported evaluation of earth observation data and weather forecasts.

The goal of FAST-CAST 2 was to make a decisive contribution to transforming data from satellite-based earth observation and weather forecasts into route suggestions through ice-covered areas on request from a ship via IcySea. AI techniques were used to accelerate the underlying processes and modern web technologies such as Progressive Web Apps were used to interact with the user.

FAST-CAST 2 combined findings from FAST-CAST and MAPBORealis and transferred them into the development and validation of a comprehensive route optimization system for ice-covered areas.

The project was led by the Zentrum für Technomathematik at Universität Bremen. The consortium included project partners Deutsches Zentrum für Luft- und Raumfahrt e.V. (DLR) and EOMAP GmbH & Co. KG (EOMAP).

In the framework of FAST-CAST2, Drift+Noise worked together with DLR in validating ice drift forecasts from the TOPAZ4 model, and extended the PRIIMA algorithm, to assimilate ice drift data from the higher resolution model neXtSIM. Another important aspect was the integration of the route optimization user interface within the IcySea app to provide the best possible and most intuitive user experience. Drift+Noise also organised two successful test cases on ships along the western coast of Greenland and the central Arctic Ocean.

DLR performed ice drift forecast validations using their newly developed algorithm based on the phase correlation method for high-resolution ice drift tracing from Sentinel-1 image pairs. Furthermore, they developed new algorithms to detect and indicate ice convergence zones. Those are dangerous areas that may even trap a ship in the ice.

EOMAP provided bathymetric information in polar regions, which are not covered as well as in the middle latitudes. They used multispectral satellite data and lidar.

Interested in finding out more? Contact us via info@driftnoise.com.