Development of a Real-Time Track Solver for Digital Twin of the Underwater Tracked Vehicle

This study focuses on developing a real-time track modeling approach to improve the computational efficiency and accuracy of tracked underwater robots operating in challenging seabed environments. The real-time track model is designed using a 3-wheel configuration and computes interactions with the terrain through triangular patch contact calculations. The research primarily addressed the interaction between the robot and seabed terrain under controlled land-based conditions, providing a foundation for real-time underwater applications. A virtual seabed terrain model was reconstructed from marine terrain data of the Jeju Sea, and simulations validated the real-time track model against a multibody dynamics (MBD) model. The results demonstrated that the real-time track model achieved comparable accuracy to the MBD model while significantly enhancing computational efficiency. Additionally, experimental validation was conducted using a physical seabed surface and sensor data to evaluate the model's performance in replicating real-world conditions. These findings highlight the model's suitability for integration into Cyber-Physical Operation Systems (CPOS), enabling real-time digital twin applications. Future research will explore its applicability in diverse underwater environments and expand its use across various tracked vehicles to further enhance underwater robot operations.