Robot Wiki

About

Robot Wiki covers modern robot learning, from Manipulation & Learned Policies and World Models to the Classical Foundations underneath. The industry appears as data in the Market Map; company news and investment advice are out of scope.

Reading the wiki

Articles stand alone, but inside a domain the reading order is the contents order: later articles assume the earlier ones. Start with Action Chunking (ACT and ALOHA), which shows the format every article follows.

The one prerequisite is fluency in machine learning. An article that needs a classical result links to the article that derives it, and terms are defined where they first appear and collected in the glossary.

Technical claims trace to cited evidence. A citation chip links each claim to its source, and every article ends with its bibliography. Sources range from research papers to first-party documentation and labelled community estimates; a citation tells you where a claim comes from, and the kind of source tells you how far to trust it. Where researchers disagree, the text names who holds which position.

Reading paths

Robot learning and VLA

  1. Robot Learning Curriculum for ML Engineers
  2. Major Datasets
  3. Behavior Cloning Foundations
  4. Action Spaces for Robot Learning
  5. Diffusion Policy
  6. Vision-Language-Action Models
  7. The Evaluation Crisis
  8. Industrial Deployment

Classical robotics foundations

  1. Kinematics
  2. Robot Calibration
  3. Perception for Manipulation
  4. State Estimation
  5. Motion Planning
  6. Control
  7. ROS 2 for Machine Learning Engineers

World models for robotics

  1. What Is a World Model?
  2. Model-Based Robot Learning
  3. Latent-Dynamics World Models
  4. JEPA and the Non-Generative Counterargument
  5. Generative Video World Models
  6. Generative Simulation
  7. World Model Evaluation for Robotics
  8. World Models vs Physics Simulators

Corrections follow the editorial policy. The author and every image licence are listed under credits.