back

Robotics Bible

A from-scratch guide to robots: the body, the math, classical autonomy, reinforcement learning, and the vision-language-action models driving the current wave of general-purpose machines.

Table of contents

  1. Chapter 1Anatomy of a RobotWhat a robot actually is: the sense–plan–act loop, joints, degrees of freedom, actuators, sensors, and why parallel parking is hard.Jun 30, 2026
  2. Chapter 2The Language of Space: Frames, Rotations & TransformsCoordinate frames, rotation matrices, gimbal lock, quaternions, and the 4×4 transforms that let a robot translate between everything it sees.Jul 13, 2026
  3. Chapter 3Kinematics: Forward, Inverse & the JacobianTurning joint angles into hand positions and back again; why inverse kinematics is hard, why 6 DoF is magic, and what the Jacobian buys you.Jul 20, 2026
  4. Chapter 4Control: From PID to Whole-BodyMaking the body actually track a command against gravity and friction: P, I, D, feedforward, and force/impedance control.Jul 27, 2026
  5. Chapter 5Estimation, Localization & SLAMFighting drift and lying sensors with probability: Bayes filters, Kalman, particle filters, mapping, and loop closure.Aug 3, 2026
  6. Chapter 6Motion Planning & NavigationFinding a path through configuration space: Dijkstra, A*, RRT and RRT*, trajectory optimization, and the global/local planner split.Aug 10, 2026
  7. Chapter 7Reinforcement Learning for RobotsLearning the policy when you can't hand-code it: MDPs, value functions, policy gradients, PPO, sim2real, and teacher-student.Aug 17, 2026
  8. Chapter 8Perception: Vision & Vision-Language ModelsThe eyes of a modern robot: attention and Q/K/V, ViT, CLIP, SigLIP, and the diffusion/flow-matching machinery that generates behavior.Aug 24, 2026
  9. Chapter 9Vision-Language-Action ModelsThe models running today's robots, as a lineage where each fixes the last: RT-1/2, OpenVLA, Octo, π0/π0.5, FAST, MolmoAct, GR00T, and Gemini Robotics.Aug 31, 2026
  10. Chapter 9.5Interlude: Flow Matching & Diffusion Models, From ScratchThe generative engine inside modern robot policies, built from zero: ODEs and SDEs, probability paths, the flow matching loss, score functions, guidance, the networks and latent spaces that host it all, and why a VLA is sure about now and vague about later.Jul 15, 2026
  11. Chapter 10Learning from Demonstration & Data CollectionWhere robot data comes from: behavior cloning, action chunking (ACT), Diffusion Policy, teleop, UMI, and the LeRobot/Rerun tooling.Sep 7, 2026
  12. Chapter 11World Models & SimulationEscaping the data bottleneck: learned world models, Dreamer, video-generation models, Gaussian splatting, sim2real, and phys2real.Sep 14, 2026
  13. Chapter 12Manipulation, Humanoids & Deploying Real SystemsAssembling the whole stack: grasping, dexterous hands, humanoids, ROS 2, evaluation, and safety, plus where the field goes next.Sep 21, 2026