Robot Motion 101 (1/3)
From Classical Planning and Control to Modern RL/VLA
A practical guide from tabletop robot hardware and software setup to first motion and custom-robot validation — 3 Parts, 9 Chapters
First published: 2026-07-14 | Last updated: 2026-07-14
Four Purpose-Built Starter Cells
Choose from a low-cost learning setup to a contact-rich dexterous-manipulation research cell.
Understand the Stack by Role
See where ROS 2, MoveIt 2, Isaac Sim/Lab, and MuJoCo connect—and where they do not.
From First Day to Sim-to-Real
Follow staged checkpoints for first motion, teleoperation, a small learned policy, and custom-robot calibration.
Part I: Design the Experimental Cell
What Will You Move? Start with Task, Budget, and Safety
Translate tabletop manipulation goals into requirements and define safety, space, power, and network boundaries.
→ 02Choose the Arm and Hand — Four Purpose-Built Configurations
Compare UR5e/Franka-class arms, parallel grippers, and five-finger hands through purpose-specific BOMs and decision gates.
→ 03Add Eyes, Controls, and Compute — Cameras, Teleoperation, GPUs, and Fixtures
Integrate VLA-ready cameras, VR/glove/tracker/joystick interfaces, GPU tiers, and safety hardware.
→Part II: Understand the Robot Software Stack
How ROS 2 Becomes Motion — From Nodes to the Real Controller
Explain ROS 2 communication, frames, timing, and drivers, plus the alternative of using vendor SDKs directly.
→ 05Classical Planning and Control — IK, MoveIt 2, and ros2_control
Follow how joint and Cartesian goals become collision-free trajectories and safe hardware motion.
→ 06There Is No Single Simulator — Roles of Isaac Sim, Isaac Lab, and MuJoCo
Select simulators for digital twins, sensing, RL, and contact-rich manipulation, then connect them through ROS 2 and data pipelines.
→Part III: Make First Motion and Cross into Reality
The First-Motion Runbook — From Simulation to the Real Robot
Use gated checks for URDF validation, dry runs, small joint moves, Cartesian motion, and grasping.
→ 08Teleoperation and a First Learned Policy — Build the Leader–Follower Loop
Connect VR, tracker, and glove inputs through IK and retargeting, then run a small imitation/RL exercise safely.
→ 09Make a Custom Robot Trustworthy — URDF, Calibration, Identification, and Sim-to-Real
Import URDF into USD/MJCF and identify kinematics, inertia, friction, actuators, and sensors to reduce reality gap as a measured error budget.
→