Robotics

Agent + reusable skill library for LIBERO manipulation tasks (no RL / VLA / world model; no LLM calls at runtime)

Robot-agent harness with a CLI and local Web console for LIBERO short evaluation

An open-source end-to-end Autonomous Mobile Robot simulation stack for ROS 2 Jazzy and Gazebo Harmonic.

[RSS 2026] Learning Point Cloud Geometry as a Statistical Manifold: Theory and Practice

2026年 东北大学T-DT实验室 RoboMaster超级对抗赛 哨兵定位开源

Reproducible Microduck RL experiments, policies, simulation assets, and hardware add-ons

[ECCV'26] VIGS-SLAM: Visual Inertial Gaussian Splatting SLAM

Robô de bancada com ESP32: rosto animado no OLED, conversa por voz em pt-BR, jogos e controle do Mac

Modern C++ frontier exploration package for autonomous mobile robots, with a core C++ library for custom integrations alongside ROS 2 Jazzy & Humble, Nav2 support.

Rethinking World-Action Model for Compositional and In-context Robotic Manipulation

AI agent operating system for Blender 5.2: bpy knowledge base, AGENT_OK execution contract, 3D-print production gate, worked keyboard and desktop-robot builds.

Generative dynamics on manifold. For Robotics Planing. Control. Learning.

Official codebase for Unitas: A 3D-Native World Action Model for Embodied Manipulation

ROS Noetic 两轮差速自主移动机器人：ESP32 底盘(PID/编码器/IMU) + J1900 车载(LiDAR/相机) + PC(SLAM/导航/人体跟随)，三机分布式全链路开源实现

[T-IM 2025] A real time LIO system focused on lifelong localization

Real time UWB localisation for small spinning robots on a tabletop play field, from ranging hardware to live position display. Includes calibration for each anchor, Kalman tracking, a browser dashboard and offline replay, together with the experimental data.

VerhoBot is a compact curtain-moving robot designed to automatically close your curtains when you need darkness and open them again when it is time to wake up, no rope,no pulley,deploy and let the bot move your curtain. The goal is simple: improve sleep quality while letting natural sunlight become part of your morning routine.

A GPU-Parallel analytical physics engine with mujoco API compatibility

Open ROS/PX4-based UAV development platform for research, education, and competition.

Use a android phone as the controller for a LeRobot mobile robot instead of a Raspberry Pi. It costs less and already has a screen, a USB camera, GPS, Wi-Fi, 5G, and an NPU.

로보틱스 개발자가 되기 위한 오픈소스 독학 커리큘럼: 강의 지도, 숫자와 그림으로 푸는 강의, 면접형 퀴즈

RL Token pipeline for the SO-101 arm: π0.5 fine-tuning, RL Token training, critical-phase classifier, online RL and deployment.

Fine-grained robot control with Jev, physics previews, and configurable LIBERO tasks.

Source-linked examples of AI models used for 3D modeling, industrial design and CAD, and robot control, with archived media and the model named for every case

Official code for ForesightSafety-VLA (IROS 2026): a unified diagnostic safety benchmark for vision-language-action models on RoboTwin

Your real-life buddy duck: an open-source embodied AI robot duck that walks, hears, sees, and has a life of its own. 让 DuckBuddy 成为你真正的生活伙伴鸭：开源具身智能小鸭机器人，运动、语音、视觉与生命层全部跑在机器人自己身上。Feetech servos + Orange Pi Zero 3W, built on Pollen Robotics' Microduck (unofficial).

🦀 Appearance and scene design for Jumper, an open-source crab robot.

超能花钱鸭 Fanduck · Microduck 已复刻成功！分享实机演示、物料清单、3D 打印、板件、制作工具与 3D 镜像工作台。Build resources from our completed Microduck replica.

[Assembly NIST] Assembly policy learning for a Franka arm in MuJoCo. Independent implementation of AlphaNIST with PPO, training code, and pretrained policies.

Toward Recursive Self-Improvement in Physical AI: Foundations, Mechanisms, and Open Problems. Survey overview and open reference library.

Best Open WHEN/DO Rule Engine for Web micro:bit Roblox 2026

A harness for embodied intelligence — connecting agents, robot skills and the physical world through execution, feedback and recovery.

FlashDexRetarget: Accelerating Dexterous Manipulation Data Generation through Multi-Motion Retargeting

An mjlab integration for the Booster K1

Every motion dataset in one place: 3D previews, measured hours, quality red flags. Human bodies, hands, interaction, animals, objects and robots.

OmniRisk: omnidirectional trajectory-risk learning for agile quadrotor dynamic avoidance. One forward pass scores every maneuver onboard; real-world evasion at up to 15 m/s.

Grounded Action Model: 3D Grounding as a Foundation for Robotics

A robot-painting benchmark for multimodal language models: reference images, text prompts, and custom goals. Built on AgentArk.

RoboWorM: Evaluating Movement Realism, Consistency and Acceleration Affects in Embodied Video World Models

Advanced Lightweight C++26 DLL Injection Framework for Windows Memory Patching 2026

First recursive non-uniform B-spline state estimation framework, supporting a wide range of multi-sensor odometry systems with optional loop closure correction.

Independent Jumper geometry recovery and assembly references: STL, faceted STEP, Blender exploded views, and pilot-fit 3MF. Physical replica not yet validated.

simulation-ready chemistry lab instruments: part-level geometry, joints, executable behaviour models and machine-readable contracts

[CoRL 2026] EgoWAM: World Action Models Beyond Pixels with In-the-Wild Egocentric Human Data

ROS2 LiDAR-inertial SLAM and multi-session alignment

Official Unitree H2 deployment of TRACE: SONIC whole-body control with motion replay, planning, and VR teleoperation.