Introduction

This section introduces the FiveAges Sim ecosystem — what it is, how its components fit together, and how to navigate the documentation.

What is FiveAges Sim?

FiveAges Sim is a collection of ROS 2 packages and workspaces that enable:

  • Robot Description — URDF/xacro models with ros2_control integration

  • Hardware interfaces — ros2_control hardware plugins for real or simulated robots

  • MPC Controllers — OCS2-based motion planning and control (分体控制 / 全身控制)

  • Simulation — Gazebo Harmonic, and FaSim (Isaac Sim high-fidelity + the same ROS 2 运控 as the real robot, plus simulation ground truth)

  • Teleoperation — VR, isomorphic teleop, and glove-based control

  • Python Applications — High-level APIs for data collection and autonomous tasks

  • Synthetic data — Isaac Sim USD → interface → composer → LeRobot record/export (Synthetic Data)

Terminology

In this documentation, humanoid means a wheeled-arm humanoid (mobile base + arms, e.g. FiveAges W2/W2R). It does not mean a bipedal or footed humanoid.

Public vs Internal

The ecosystem has two entry points:

Path

Workspace

Audience

Robots

Public

open-deploy-ws

Open source users

Dobot CR5, ARX Lift 2S, Acone (dual-arm), Galbot, HighTorque Panthera HT, etc.

Internal

fa-deploy-ws

FiveAges team

W2, W2R, S2, S2R, dual-arm CCS

If you’re new, start with the public path. It uses only public submodules and can run demos without special access.

Real Hardware on Public Path

ARX Lift 2S (full-body mobile manipulator, including chassis) and HighTorque Panthera HT (dual-arm manipulator) have dedicated open-deploy-ws branches. Acone / AC One is dual-arm, not the Lift 2S platform. See ARX Lift 2S and HighTorque Panthera HT.

Documentation Map

  • Architecture — Layer diagram and dependency relationships

  • Terminology — 驱动层 ↔ Hardware Interface, 分体 / 全身, 夹爪 / 灵巧手, simulation keys

  • Learning Path — Day-by-day onboarding guide

  • Repository Map — Complete list of repositories by layer

  • Public vs Internal — Detailed comparison of the two paths

  • Synthetic Data — Isaac datagen pipeline (USD → orchestration → LeRobot export)

Next Steps

  1. Review the Architecture to understand component relationships

  2. Follow the Learning Path for structured onboarding

  3. Install Environment, then Quick Demo