Beyond URDF: The Universal Robot Description Directory for Shared, Extensible, and Standardized Robot Models

TL;DR

The paper presents the Universal Robot Description Directory (URDD), a modular, extensible representation that stores derived robot information in JSON/YAML modules to overcome the redundancy and fragmentation of traditional formats like URDF. A Rust-based URDF-to-URDD converter generates comprehensive modules (e.g., DOF mappings, kinematic chains, mesh decompositions) and includes Bevy-based visualization, while a web-based Three.js viewer enables browser-based inspection. The URDD’s modular structure supports incremental extensions and composite assemblies, reducing preprocessing effort and enabling immediate use in planning, control, and visualization. Empirical evaluation shows efficient generation across diverse robots, richer data content than URDF, and substantial reductions in FK implementation effort, underscoring URDD’s potential to unify robotics resources across frameworks. The work discusses limitations (scalability, module dependencies) and outlines future directions toward standardized validation and community-driven evolution of the shared robot description ecosystem.

Abstract

Robots are typically described in software by specification files (e.g., URDF, SDF, MJCF, USD) that encode only basic kinematic, dynamic, and geometric information. As a result, downstream applications such as simulation, planning, and control must repeatedly re-derive richer data, leading to redundant computations, fragmented implementations, and limited standardization. In this work, we introduce the Universal Robot Description Directory (URDD), a modular representation that organizes derived robot information into structured, easy-to-parse JSON and YAML modules. Our open-source toolkit automatically generates URDDs from URDFs, with a Rust implementation supporting Bevy-based visualization. Additionally, we provide a JavaScript/Three.js viewer for web-based inspection of URDDs. Experiments on multiple robot platforms show that URDDs can be generated efficiently, encapsulate substantially richer information than standard specification files, and directly enable the construction of core robotics subroutines. URDD provides a unified, extensible resource for reducing redundancy and establishing shared standards across robotics frameworks. We conclude with a discussion on the limitations and implications of our work.

Figures (4)

• Figure 1: Structure of the Universal Robot Description Directory (URDD). The URDD organizes derived robot information into modular subdirectories stored in JSON/YAML. Core modules include kinematic structure (chain, connections, DOF mappings), joint bounds, and preprocessed geometry (meshes, convex hulls, bounding volumes). Each module is version-tagged and independently extensible, enabling richer, reusable data for planning, control, and visualization without the redundancy of re-deriving information from raw URDFs.
• Figure 2: Our software tools enable seamless combination of multiple URDDs. In this example, a Robotiq gripper (left) is attached to a Unitree Z1 arm (center), which is then mounted onto a Unitree B1 quadruped (right). The resulting composite URDD integrates the information from all three platforms into a single, unified directory.
• Figure 3: The Bevy-based graphics front-end powers a proximity visualization, enabling users to observe distances between pairs of link shapes. In this instance, the visualization shows the distance between two convex decomposition shapes.
• Figure 4: In-browser visualization. a) Overall UI and visualization of the B1 robot. b) Oriented bounding boxes for each link. c) Bounding spheres for each link. d) Oriented bounding boxes for each element of the convex decomposition. e) Bounding spheres for each element of the convex decomposition. f) Convex hulls. g) Convex decomposition for each link.