Facilitating Safe Automated Navigation in Ice with Cooperating Autonomous Vehicles

Autonomous shipping has gained much interest in the last decade stemming from several desirable safety and environmental benefits. A significant boost to Canada's autonomous shipping drive would be adequately monitoring ice conditions in shipping environments with unmanned assets and calculating safe navigation routes for autonomous ships. A workflow was created to simulate an unmanned aerial vehicle (UAV) cooperating with a maritime autonomous surface ship (MASS) to facilitate the MASS's safe navigation through an icy waterway. Both vehicles were modeled as MATLAB Simulink's Sim3D objects operating in an icy shipping environment representing a section of Canada's St Lawrence Seaway and built with Unreal Game Engine version 4.26.2. The UAV performed regular ice-condition-monitoring flights over the shipping environment with its vision and LiDAR sensors. Maps of Ice Numerals (IN) pertaining to the MASS's class were developed from processed data from the UAV's sensors. They served as inputs to a Rapidly Exploring Random Tree (RRT) algorithm that estimated safe obstacle free routes for the MASS through the shipping environment. Colour-coded hazard maps specifying safe and hazardous waterway regimes were also developed from IN maps. Hazard maps could serve as navigational aids to operators monitoring MASS operations in hazardous environments. Results obtained from testing the workflow on the MA TLAB Simulink + Unreal Engine simulation environment produced the desired INs and hazard maps with safe navigation routes for the MASS. These results show the possibility of having autonomous assets corporate to facilitate safe autonomous navigation in ice.