Open X-Embodiment
A large cross-embodiment collection of robot trajectories from many institutions.
Practical index
Datasets, benchmarks, simulators, and frameworks
A curated set of primary resources for studying and building Physical AI systems.
Open resource
DROID
A diverse real-world robot manipulation dataset collected across many scenes and operators.
LIBERO
A benchmark for lifelong robot learning with manipulation task suites.
CALVIN
A language-conditioned long-horizon manipulation benchmark.
NVIDIA Isaac Sim
A robotics simulation platform for synthetic data, training, and testing.
NVIDIA Halos for Robotics
A full-stack functional-safety platform spanning IGX Thor, Halos OS, safety application blueprints, and inspection support.
NVIDIA Halos Outside-In Safety
Infrastructure-camera safety agents for robot zoning, perception monitoring, event integration, and safety decisions.
MuJoCo
An open-source physics engine widely used for robotics and control research.
LeRobot
Open models, datasets, and tools for real-world robotics in PyTorch.
OpenPI
Open-source models and tooling from Physical Intelligence.
NVIDIA Cosmos
World foundation models, guardrails, tokenizers, and tooling for physical AI data generation.
NVIDIA Isaac GR00T
Robot foundation model and data pipeline platform for humanoid and general-purpose robot learning.
Isaac Lab
A unified framework for robot learning research built around Isaac Sim.
Gemini Robotics
Google DeepMind's embodied AI model family for visual reasoning and robot action.
Ego4D
Large-scale first-person video dataset for hand-object interaction, memory, forecasting, and social understanding.
Ego-Exo4D
Synchronized first-person and third-person skilled activity dataset with language, audio, gaze, pose, and 3D signals.
Project Aria Gen 2 Pilot Dataset
Open egocentric wearable-sensor dataset for contextual AI, robotics, and smart-glasses research.
World Labs World API
API for generating, rendering, and exporting navigable 3D worlds from images, text, panoramas, and video.
Encord Physical AI Data Layer
Multimodal data operations for video, robotics, world models, and VLA training datasets.
Labelbox Sensors and Robotics
Labeling and curation workflows for sensor, robotics, video, and computer-vision datasets.
Scale AI Data Labeling Guide
Reference for image, video, sensor, and temporal annotation workflows used in AI data engines.
Label Studio
Open-source multimodal data labeling and evaluation tool for images, video, text, audio, and time series.