Physical AI Lab

Build robot sensor fusion with timestamp and frame alignment, covariance, EKF or UKF estimation, outlier rejection and fault validation.

Understand robot whole-body control through contacts, task priorities, momentum, constraints, optimization, learned commands and validation.

Compare robot motion planning and trajectory optimization by search, initialization, collision, costs, dynamics, timing and feedback execution.

Compare inverse kinematics and inverse dynamics by inputs, outputs, constraints, singularities, torque limits, contact and feedback control.

Separate robot reducer backlash, lost motion, hysteresis and torsional stiffness and compare them with a controlled torque-angle test.

Understand robot motor driver architecture including DC bus, inverter, gate drivers, current sensing, control MCU, protection and regeneration.

Compare series elastic and quasi-direct-drive actuators by force sensing, backdrivability, bandwidth, impacts, thermal duty and validation.

Understand field-oriented control for robot motors, including phase current, rotor angle, d-q axes, PI control, PWM, limits and torque validation.

Understand robot joint backdrivability by reduction ratio, reflected inertia, friction, motor electrical state, active control and output tests.

Compare robot actuator torque density by peak and continuous output, mass, volume, transmission, cooling, speed and measured duty cycle.