Code Generation from Unified Robot Description Format (URDF) for Accelerated Robotics

This conference talk explores how to accelerate motion planning for robots through compile-time optimizations, focusing specifically on forward kinematics calculations. Learn how to translate Unified Robot Description Format (URDF) directly into optimized C++ code using code generation techniques. Discover methods for generating hardware-specific code like CUDA for parallel computations to achieve significant performance gains. Compare benchmarks between compiled URDF and naive implementations of forward kinematics, and evaluate how design decisions like memory layout and compiler flags impact performance. The presentation demonstrates integrating code generation into the C++ build process using CMake and Jinja templates for easy adoption. Gain valuable insights into compile-time optimization benefits for performance-critical robotics applications from Paul Gesel, a Robotics Scientist at PickNik Robotics with a Ph.D. in Computer Science focused on robotics.