Resources

Resources

Tools, dataset platforms, robotics guides, repositories, and community resources for Physical AI

Overview

Here is a curated directory of tools, repositories, documentation, and community channels across the Humaan ecosystem to help you build, train, and deploy embodied intelligence.

Core Platforms & Tools

Humaan Loop

The continuous improvement cycle for robot intelligence. Connect dataset ingestion, policy training, and evaluation in one closed-loop system so every real-world interaction makes your robots more capable.

  • Explore Humaan Loop
  • Capture observations, actions, and outcomes from physical hardware.
  • Benchmark and evaluate policy iterations against standardized tasks.

Humaan Index

The data foundation for robot learning. Review and curate robot experience into learning-ready datasets.

  • Explore Humaan Index
  • Episode timeline scrubbing and multi-camera stream inspection.
  • Filter, label, and export training batches with high data quality.

Interactive 3D Robot Arm

Interactive 3D simulation and model viewer running directly in the browser.

  • Launch Robot Arm Simulation
  • Test inverse kinematics (IK) and joint limits in real time.
  • Inspect coordinate frames, end-effector poses, and mechanical linkages.

Guides & Learning Material

The Robotics Manual

A foundational, hands-on reference covering the core mechanics and systems behind robotics.

  • Open The Robotics Manual
  • Foundations: What defines an embodied robot and how the sense-think-act loop operates.
  • Sensors & Actuators: Perception layers, cameras, depth sensors, IMUs, brushless motors, and gearboxes.
  • Kinematics & Control: Joint coordinate systems, degrees of freedom, and closed-loop feedback.

Platform Documentation

Official technical guides for setting up environments, working with datasets, and integrating robot hardware.

Hardware & Making

Accessible Experimentation

Our approach combines affordable, modular hardware with open-source software and 3D-printable components.

  • Standardized off-the-shelf actuators and high-torque servomotors.
  • Open 3D printable designs (STL and STEP files) for rapid iteration in maker spaces and garages.
  • Lowering the barrier to entry so you don't need a multi-million-dollar lab to advance physical AI.

Code & Open Source

GitHub Repositories

Explore source code, contribute tools, and download hardware models.

  • GitHub Organization: Core platform, drivers, and dataset utilities.
  • Model Checkpoints: Reference policies and baseline models for imitation learning and teleoperation.
  • CAD & Mechanical Files: Printable STL files and assembly schematics.

Community & Support

Join fellow roboticists, AI researchers, and builders: