The Lab. of Off-road Machinery System and Terramechanics focuses on advancing next-generation intelligent off-road machinery by integrating terramechanics with Physical AI. Our research aims to understand and model the complex interactions between terrain and machinery in unstructured and extreme environments, enabling robust and high-performance system operation even under harsh conditions.

We develop high-fidelity Digital Twin environments through multi-physics integrated modeling, providing a realistic virtual platform for system analysis, design optimization, and control development. Based on terramechanics principles, we investigate soil–tool and soil–traction interactions to derive optimal design and control strategies using advanced approaches such as reinforcement learning.

In addition, we conduct research on Prognostics and Health Management to enhance system reliability by predicting component fatigue life and detecting potential failures using virtual sensing data. Our work also includes high-precision computational fluid dynamics and intelligent hydraulic systems to improve energy efficiency and optimize thermal management for electrified off-road machinery.

Furthermore, we develop physics-based multimodal perception technologies that integrate vision and proprioceptive signals to estimate terrain properties in real time, enabling adaptive and reliable control of machine subsystems. To address the limitations of real-world data in off-road environments, we build a simulation-based synthetic data platform to support AI training and accelerate system development.

Through these research efforts, the Lab. of Off-road Machinery System and Terramechanics bridges physics and artificial intelligence, pioneering innovative solutions for future agriculture and off-road industries.