Robot Motion

Use MotionGenerator to turn joint or end-effector goals into timed trajectories. It combines trajectory planners with kinematic solvers; simulation playback and Atomic Skills execute the resulting motion.

Choose a capability

Goal	Start here
Generate a trajectory from robot goals	MotionGenerator: strategies, batched inputs, results, and playback.
Choose a timing or collision-aware planning backend	Planners: backend selection and configuration.
Compute forward or inverse kinematics	Solvers: solver selection and shared FK/IK conventions.
Analyze reachability or sample reachable poses	Robot Workspace: workspace analysis and caches.
Generate variations of an expert trajectory	Trajectory augmentation API: candidates, variation operators, coverage, and generation budgets.
Collect several ways of executing one fixed set of waypoints	Trajectory Variants: path, posture, timing, and approach variation for imitation learning and reinforcement-learning post-training.

Trajectory augmentation requires host integrations for scene reset, planning, rollout, task validation, and dataset persistence.

Import APIs from their owning module, such as embodichain.lab.sim.motion.motion_generator or embodichain.lab.sim.motion.planners. See the motion API for the complete public interface.

• MotionGenerator
• Trajectory Variants
• Planners
• Solvers