Xilinx is a proud supporter of the ROS ecosystem, contributing to ROSCon and ROS-Industrial activities. While KRS focuses on adaptive computing, as part of Xilinx commitment with ROS 2, the ROS 2 Hardware Acceleration Working Group (HAWG) was created. With it, Xilinx is contributing back to the ROS 2 community and driving the creation, maintenance and testing of acceleration kernels on top of open standards (C++ and OpenCL) for optimized ROS 2 and Gazebo interactions over different compute substrates (including FPGAs and GPUs). The main driver behind this WG is summarized at the ROS Enhancement Proposal (REP) "ROS 2 Hardware Acceleration Architecture and Conventions". In here, Xilinx describes the architectural pillars and conventions required to introduce hardware acceleration in ROS 2 in a scalable and technology-agnostic manner.
As specified on its community announcement, the HAWG targets hardware acceleration in a) embedded (edge) devices, b) workstations, c) data centers and d) cloud. A roadmap with objectives is available in the community repository of the Working Group, which Xilinx is driving and executing disclosing results as they become available.
Through the HAWG, Xilinx is demonstrating hardware acceleration capabilities in ROS 2 on edge devices with simple examples at the application layer that roboticists could use as blueprint for their designs. In the future, the group will also target ROS 2 underlayers to optimize interactions between nodes within the ROS 2 computational graph.
To learn more about Xilinx contributions to the ROS 2 community, check out the following resources:
| Resource | Link |
|---|---|
ROS 2 HAWG GitHub organization |
ros-acceleration |
ROS 2 HAWG meeting minutes |
HAWG minutes |
ROS 2 HAWG meeting invite group |
ROS 2 Hardware Acceleration WG Google Group |
ROS 2 HAWG instant messaging |
Matrix community (Matrix is an open network for secure, decentralized communication). |
ROS 2 HAWG backlog management |
GitHub project |
ROS 2 HAWG discourse tag |
wg-acceleration |