This project demonstrates Diffusion Model Predictive Control (DMPC) for safe and efficient navigation in stochastic ice environments. Our approach leverages real-world data to create an accurate simulation environment for autonomous ship navigation in ice-covered waters.
Key Features
- Real-World Data Integration: Utilizes real satellite imagery data to generate ice maps, providing realistic representations of ice floe distributions in the Arctic region
- Ice Ridge Detection: Employs satellite imagery to detect ice ridges and generate real ice ridge density maps, enabling precise modeling of ice structure and navigation hazards
- Dynamic Environment Modeling: Incorporates real ocean current data to accurately model the dynamic marine environment
- Comprehensive Simulation: Together, these real-world data sources create a comprehensive and realistic testing environment for autonomous navigation algorithms
Demo & Resources
Authors
Diego Calanzone, Gabriel Sasseville, Akash Karthikeyan
References
This work is based on research from de Schaetzen et al. (2024) on AUTO-IceNav: A Local Navigation Strategy for Autonomous Surface Ships in Broken Ice Fields, and incorporates Diffusion Model Predictive Control techniques from Pan et al. (2024).