ROS News for the Week of August 24th, 2026

ROS News for the Week of August 24th, 2026


I ran into @01binary and his Str1ker mech robot at Open Sauce and Teardown. Quite frankly, it is one of the most over-the-top ROS personal projects I’ve ever seen. This week the source code and CAD files dropped so you can take a peek under the hood.



We’re about three months out from the Lyrical Release and the ROS package ecosystem is starting to catch up. This week @smac released Nav2 for Lyrical! There’s a host of new features and it is worth checking out the release notes.


smolvla
This week a team in Germany released ROS2SmolVLA a ROS package that enables small vision-language-action models with industrial-grade cobots running ROS 2. I had a lot of great conversations this week about “small models” for robot deployment and this seems to scratch that itch.


IBVS-admittance
IBVS-rot
IBVS-trans

@aitor-ibarguren just released a set of ROS 2 visual servoing controllers for ROS 2 Jazzy that includes both a position-based visual servoing (PBVS) and an image-based visual servoing controller (IBVS). The controllers include an action server providing a parametrizable PBVS/IBVS task, mimicking the joint trajectory controller’s functionality.


autodock
Our colleagues over at Hello Robot just published a blog post on building autodocking for the Stretch 4.



Our colleagues in Japan just released this adorable 3D printed mini ROS 2 quadruped.

Events

ROSCons

News

ROS

Got a Minute? :clock:

We’re in the midst of updating the ROS 2 documentation. We could really use some help from the community reviewing all the recent pull requests.

2 Likes

Would it be possible to add this meetup to the events list as well?

Thanks!

Sure thing! I must have mis-read it as the other Nordic ROS event and skipped over it!

One for a future roundup, if it fits: Provael 0.38.1 is out. Open-source, Apache-2.0, CPU-first harness that red-teams vision-language-action policies in simulation and reports an attack success rate against a benign control arm.

The release itself is a correction rather than a feature, which is why I am mentioning it: we had published a zero-width confidence interval for three null arms for 21 days. 0 of 50 is consistent with a true rate as high as 7.1%, and printing [0%, 0%] claims a certainty the data does not have. Fixed and written up.