Following from Hardware acceleration in ROS 2 and Gazebo survey, I’m happy to finally share the 2022 Hardware Acceleration Report in Robotics () which captures the state-of-the-art of hardware acceleration in robotics today and presents robotic architects with a resource to consider while designing their robot brains.
This report results from more than a year collecting data, distilling and speaking to roboticists to understand their views on hardware acceleration. The work presented in this report happened in two phases. First, a community survey conducted in both the ROS and the overall robotics communities helped grasp the interest behind the use of hardware acceleration in robotics. Input from this community survey was then used to drive the second phase, a hardware acceleration benchmarking effort. The most relevant results from these two phases are summarized in the report.
One aspect to highlight about this report is that its uses the Robot Operating System (ROS), as the common baseline in robotics and to establish reproducible benchmarks. Rationale comes from the fact that most companies building real robots today use ROS or similar event-driven software frameworks. This report uses ROS as the common baseline in robotics to conduct the study. In particular, we use ROS 2 which presents a modern industry-accepted framework for robot application development and consider both bandwidth and latency to benchmark performance in robotics using a grey-box and non-functional benchmarking approach.
The work in here aligns also with the efforts being driven at the ROS 2 Hardware Acceleration Working Group. Thanks to everyone who helped review the document, send feedback and constructive criticism.
If you’d like to be involved in future-related topics, join us at this WG!