Physic Based Simulation

Physics-based simulation aims to accurately model the physical behavior of systems, primarily for predicting their responses to various conditions and actions. Current research emphasizes improving efficiency and accuracy, particularly in handling complex interactions like deformable objects and frictional contact, often employing techniques like model reduction (e.g., subspace methods), Gaussian splatting, and Material Point Methods (MPM). These advancements are crucial for diverse applications, including robotics (manipulation, locomotion), autonomous driving (human behavior modeling), and scientific discovery (extraterrestrial terramechanics, material science), enabling more realistic simulations and data generation for training and testing algorithms.