Robot Wiki

Classical Foundations

Classical robotics is the model-based stack that learned policies still run on: kinematics, planning, control, estimation, grasping, perception, mapping, calibration and ROS 2. In Intrinsic's 2026 AI for Industry Challenge, 90% of the top teams in the real-workcell phase inserted the connector with rules-based routines and force-torque feedback Intrinsic 2026.

Domain overview: 9 articles. Each article builds on the ones before it.

Related: Robot Learning Curriculum for ML Engineers, The Robot Learning Stack, Safety and Assurance

  1. Kinematics

    How joint angles place a robot's hand, and the reverse.

  2. Motion Planning

    Configuration space, RRT and its variants, and trajectory optimization.

  3. Control

    PID, LQR, MPC and whole-body QP, the stack under every learned policy.

  4. State Estimation

    Kalman filters, factor graphs, and pose estimation from noisy sensors.

  5. Grasp Planning

    Contact mechanics, grasp quality metrics, and force closure.

  6. Perception for Manipulation

    How camera images become an object's position and orientation.

  7. Scene Representation and Mapping

    From point clouds to Gaussian splats: what a robot remembers about space.

  8. Robot Calibration

    The measurements that make sensor coordinates agree with motion.

  9. ROS 2 for Machine Learning Engineers

    ROS 2 as the production boundary around a learned policy.