Humanoid robotics glossary

URDF

Also known as: Unified Robot Description Format, Robot URDF

An XML-based robot description used to represent links, joints, geometry and physical properties for robotics software.

Last verified Oct 13, 2026

What it means

URDF is an XML-based format for describing a robot's structure and properties. It represents links and the joints connecting them, with named coordinate relationships. ROS's urdfdom project provides core data structures and XML parsing for these descriptions. A URDF file is therefore a model of the robot, rather than a program that decides how it should move. [1]

Different parts of a description serve different purposes. ROS 2's visual-model tutorial starts with link geometry and fixed joints so the robot can be displayed in RViz. Its physical-properties tutorial adds collision and inertial information needed when using the model for more than appearance. [2, 3]

How it applies to humanoids

A humanoid model can describe its torso, limbs and attached devices through the same link-and-joint structure. Software can use those relationships to display poses and connect joint-state information to spatial transforms. The ROS tutorial demonstrates this workflow with robot_state_publisher and a joint-state publisher. [2]

Physics requires additional information. The physical-properties tutorial explains link mass and inertia and distinguishes collision geometry from visual geometry. A model that looks like a humanoid can therefore be useful for visualization while still lacking the properties needed for a credible dynamics simulation. [3]

Terminology and model traps

Visual detail is not physical accuracy. Collision shapes can be simpler than the visible meshes to reduce computational cost, and the tutorial treats the two representations separately. A realistic-looking foot does not establish that the contact geometry or inertial values match the hardware. [3]

A valid parser result means the file meets the parser's requirements, not that its measurements are correct. urdfdom supplies the parsing layer; the physical-properties tutorial still requires meaningful masses and inertia. File validity and model fidelity are separate checks. [1, 3]

Count independent robot motion carefully. The visual tutorial uses fixed joints to connect parts, so the number of joint elements is not automatically the number of movable degrees of freedom. A software model can include structural connections that add no independent motion. [2]

How to read it on humanoidversus.com

Read dof_total and weight_kg from the robot record's cited specifications rather than automatically extracting them from an arbitrary model download. A URDF may describe a particular version or attachment, and the record should identify that scope before using its values.

This entry has no live table or robot links. The open-source hub helps locate public developer interfaces, but a public SDK alone does not guarantee a published URDF or accurate simulation model. Follow the project's actual model repository and documentation, and leave undisclosed model availability unknown.

Further reading

urdfdom documents the parsing library. The ROS 2 Jazzy visual and physical-property tutorials explain how links, transforms, collision shapes, mass and inertia contribute to different uses of a robot description. [1–3]

Related data

Explore connected records

References

Sources

  1. [1]urdfdom: URDF Parser and Core Data Structures ↗ROS · official · accessed Oct 13, 2026
  2. [2]ROS 2 Jazzy Documentation: Building a Visual Robot Model with URDF from Scratch ↗ROS 2 · official · accessed Oct 13, 2026
  3. [3]ROS 2 Jazzy Documentation: Adding Physical and Collision Properties to a URDF Model ↗ROS 2 · official · accessed Oct 13, 2026