Menu
ROBOT DESCRIPTION FORMATS
Structure, scenes, simulation and semantic embodiment are different jobs.
URDF, USD, SDFormat, MJCF and RGD overlap in useful ways, but they are not interchangeable. The important question is not “which format wins?” — it is “which layer owns which responsibility?”
RESPONSIBILITY MAP
Five formats, five different centers of gravity.
Links, joints, geometry, inertials, collision and joint constraints in a ROS-centric XML model.
Not a general scene-composition system or semantic embodiment graph.
USDComposable scene descriptionLarge hierarchical scenegraphs, references, variants, time-sampled data and physics schemas.
Not a robot-specific semantic capability or governance model by default.
SDFSimulation-oriented model and world descriptionRobots, environments, models, sensors, physics, lights, plugins and simulation context.
Not designed as a general AI embodiment-governance or provenance graph.
MJCFMuJoCo-native dynamic-system modelNested bodies, joints, actuators, tendons, sensors, defaults and simulator-native dynamic modeling.
Optimized for MuJoCo rather than cross-stack semantic embodiment.
RGDSemantic embodiment graphIdentity, physical evidence, operational constraints, capability semantics, provenance and lifecycle context.
Does not replace scene composition, a physics engine, middleware or a physical runtime.
WHY THIS MATTERS
Interoperability improves when layers stop pretending to be each other.
URDF
Excellent for robot kinematic structure and common ROS robot-model workflows.
USD
Strong for scalable scenegraphs, composition and rich scene/physics data.
SDF
Strong for simulation-oriented robots, models, sensors and whole environments.
MJCF
MuJoCo-native dynamic modeling with rich simulator-specific bodies, actuators, tendons and sensors.
RGD
Focused on semantic embodiment, provenance, operational context and machine-readable grounding.
DEEP COMPARISONS
Compare the layers directly — including the choices developers actually make.
URDF vs SDF
ROS robot model vs simulation-oriented model/world.
Open comparison →COMPAREURDF vs USD
Robot model vs composable scenegraph.
Open comparison →COMPAREURDF vs MJCF
Portable robot structure vs MuJoCo-native modeling.
Open comparison →COMPAREURDF vs RGD
Robot structure vs semantic embodiment.
Open comparison →COMPAREUSD vs RGD
Scene composition vs semantic embodiment.
Open comparison →COMPARESDF vs RGD
Simulation description vs semantic embodiment.
Open comparison →FORMATS → SEMANTICS
A robot file tells you about structure. Semantic layers tell you what those facts mean together.
Complement, do not collapse.
OpenRGD is designed to consume source-supported evidence from robot descriptions, preserve provenance and add semantic context. It is not trying to become a full USD scene engine, a Gazebo world format, ROS middleware or a physical runtime.