Hardware Acceleration Working Group #8
13 views
Skip to first unread message
Víctor Mayoral-Vilches
Apr 19, 2022, 2:19:50 AM4/19/22
to ROS 2 Hardware Acceleration WG
https://calendar.google.com/event?action=TEMPLATE&tmeid=MjNkcXQ2ZnJqMHU2MHM5ZWdoa3NvODRwb2ggYWdmM2thamlya2V0OGtoa3R1cG05Z283NDhAZw&tmsrc=agf3kajirket8khktupm9go748%40group.calendar.google.com
The meeting will cover updates over the last month. In particular, we'll touch on a) updates of REP-2008 to reflect cloud extensions (https://news.accelerationrobotics.com/ros-2-hardware-acceleration-architecture-and-conventions/), b) ROS 2 perception Nodes performance benchmarking, c) ROSCon 2022 Hardware Acceleration Workshop, d) the Robotic Processing Unit and e) progress on community ports of the architecture (`Ultra96-v2`)!
The session will also include a guest talk provided by Jeffrey Ichnowski, a post-doctoral researcher in the RISE lab and AUTOLAB at the University of California at Berkeley, which together with his team led "*FogROS 2: An Adaptive and Extensible Platform for Cloud and Fog Robotics Using ROS 2*". This work is specially exciting to me because I collaborated with this group for the last few months to align cloud and edge hardware acceleration capabilities with ROS 2. The results obtained in this research led to ROS VSLAM (`ORB-SLAM2`) latency reductions of 50%, grasp planning optimizations going from 14s to 1.2s, and motion planning speedups of 28x.
The meeting will cover updates over the last month. In particular, we'll touch on a) updates of REP-2008 to reflect cloud extensions (https://news.accelerationrobotics.com/ros-2-hardware-acceleration-architecture-and-conventions/), b) ROS 2 perception Nodes performance benchmarking, c) ROSCon 2022 Hardware Acceleration Workshop, d) the Robotic Processing Unit and e) progress on community ports of the architecture (`Ultra96-v2`)!
The session will also include a guest talk provided by Jeffrey Ichnowski, a post-doctoral researcher in the RISE lab and AUTOLAB at the University of California at Berkeley, which together with his team led "*FogROS 2: An Adaptive and Extensible Platform for Cloud and Fog Robotics Using ROS 2*". This work is specially exciting to me because I collaborated with this group for the last few months to align cloud and edge hardware acceleration capabilities with ROS 2. The results obtained in this research led to ROS VSLAM (`ORB-SLAM2`) latency reductions of 50%, grasp planning optimizations going from 14s to 1.2s, and motion planning speedups of 28x.
Víctor Mayoral-Vilches
Apr 19, 2022, 2:20:37 AM4/19/22
to ROS 2 Hardware Acceleration WG
As always, a discussion is available at https://discourse.ros.org/t/hardware-acceleration-wg-meeting-8/25147
Reply all
Reply to author
Forward
0 new messages