Optimized route suggestions for ships in polar regions using AI- based processing, Earth Observation data, and model forecasts

Kurzfassung

Climate change and subsequent retreat of sea ice favor the increase of maritime traDc in the polar regions. Despite the growing importance of the ice-covered Arctic Ocean for all types of global shipping and the availability of massive Earth Observation (EO) and weather data archives, navigational decision making is still limited. This is due to the fact that manual data pick up by the ship crews continues to be next to impossible for at least the following two reasons. (1) Communication bandwidth is limited in Arctic regions and (2) the conversion of such data into navigational decisions is inherently a big data process. The FastCast-2 project develops innovative solutions to enhance the value of satellite-based EO and weather data. Furthermore, we process that data into near-real time and on-demand route suggestions for stakeholders transiting through or around ice-infested waters. AI-techniques are used to accelerate underlying processing steps and state-of-the-art web technologies like ‘Progressive Web Apps’ are utilized for user interaction. The International Maritime Organization’s Polar Code requires ships to have up-to-date ice information on board when heading towards the polar regions. Although the ice is retreating, sea ice, icebergs and uncharted bathymetry remain hazardous for modern day shipping in the Arctic. The FastCast-2 consortium strives to combine improved sea ice drift forecasts with up-to-theminute ice and weather information, which are processed into continuously updated route suggestions. These route suggestions will also avoid shallow areas detected by satellite-derived bathymetry. This kind of service will not only help ships to fulfill the Polar Code requirements. It will also advance digitization of the shipping industry and enable ship crews and their companies to find the safest, fastest and most fuel-effient routes in and around ice-covered waters.