so101

Control AI robots. Community-driven UI middleware for controlling robots, recording datasets, training action models. Compatible with SO-100 and SO-101

Isaac Lab external project for SO-ARM100/101 arm robot.

Squint: Fast Visual Reinforcement Learning for Sim-to-Real Robotics [PyTorch, SO-101 Robot Arm, ManiSkill3, Sim-to-Real]

Autonomous Robot that builds Electrical Circuits. 4th at Hugging Face Global Robotics Hackathon

实机部署so101流程和踩坑全记录，使用lerobot框架so101真机。在本地 ubuntu22+5070显卡 搭建环境，部署，录数据，训练smolvla，推理全流程。

URDF-based forward and inverse kinematics helpers for robot arms — pluggable IK backends (Placo, PyRoki, RoboPlan), Viser 3D visualization, Rerun logging, and LeRobot/ROS 2 real-arm integration.

实机部署so101流程和踩坑全记录，使用lerobot框架以及isaaclab（issac sim）在虚拟环境跑通so101。在本地 ubuntu22+5070显卡 搭建环境，部署，录数据，训练smolvla，推理全流程。

End-to-end SO-101 tic-tac-toe: printable hardware, local UI, LeRobot dataset, SmolVLA policy and guarded Strands agent

Manual real‑world benchmarking of ACT/SmolVLA on a physical SO101 robot using LeRobot.

Voice Controlled SO101 (lerobot) robotic arm

Viam support for SO-101 robotic arm

MCP server that turns a sentence into a 3D-printed part: parametric CAD with build123d, slicing with PrusaSlicer, printing via OctoPrint on an Ender-3 V3 SE — with a human approval gate at every step that moves the machine.

World Action Model for LeRobot SO-101 control implemented from scratch, using OXE, DROID, and BRIDGE for pretraining and task-specific actions for finetuning, able to accomplish a range of dextrous tasks.

Reproducible no-sudo SO-101 VLA training, evaluation and HITL data-loop pipeline

early experiments on vlm + graspnet on so101 real robot, done by pathonai.org open source contributors

Personal SO-101 + FlexiTac reproduction/adaptation based on Evo-RL upstream.

ROS2-based visual-tactile grasping system for the SO-101 robot arm with vision, FK/IK, trajectory planning and FlexiTac feedback.

AMD Robotics Hackathon 2025 submission (3rd place): Connect 4 robot opponent for SO-101 with vision-based board perception, game-state reasoning, move planning, and LeRobot/SmolVLA execution on AMD ROCm

Blender仿真同步SO-101机器臂

Show a robot once. One-shot task teaching for robot arms: narrate (NVIDIA Cosmos on Nebius) → rehearse in Isaac Sim (Antioch) → execute → verify → remember → compose by language → human-verify (Toloka). Programs first, policies later.