SANTOS is a physics-based human digital twin platform developed to support human modeling and simulation (M&S) for applications in human systems integration, product development, and human performance. Central to the platform are two digital human models, Santos and Sophia, designed to simulate human physiology and motion with a high degree of fidelity. The system builds on over two decades of research and is intended to provide a computational alternative to physical testing for evaluating individual performance and designs, early in the cycle. SANTOS employs Human Predictive Dynamics (HPD), an optimization-based approach that predicts human motion without explicitly solving differential equations of motion. This methodology accounts for mass, inertia, joint velocities and accelerations, external forces, muscle strength, fatigue, and other physiological parameters, producing realistic movement trajectories under task-specific constraints. The musculoskeletal model incorporates 215 degrees of freedom and integrates biomechanics, physiological modeling, artificial intelligence components, and extensive empirical data derived from military and industrial sources. The platform’s utility is demonstrated through its application in assessing performance on the Army Combat Fitness Test (ACFT), among other military use cases. This paper presents the mathematical framework of the HPD approach and discusses its implications for enhancing Soldier readiness through predictive human performance modeling.