I built RoboOps, a small fleet-operations prototype around @NVIDIARobotics NVIDIA Isaac Sim.
The demo connects an AWS-hosted operator UI with an Isaac Launchable running on an L40S GPU.
The repo is in the comments.
Had a blast at yesterdays Robot Learning Paper Club by @DominiqueCAPaul.
Thanks for organizing and bringing together ETH wizards and robotic people of Zurich 👏 Looking forward to the next event.
Yesterday, we hosted the first robot learning paper club in Zurich.
Five papers and one open-source project. Next one coming up in 3-4 weeks.
Thanks for this very cool edit by @mrza_ch.
NVIDIA showcased Newton at GTC again earlier this year.
If you are familiar with MuJoCo or Isaac Sim, Newton is a new open-source, GPU-accelerated physics engine aimed at robotics and contact-rich manipulation, built with DeepMind and Disney Research. It also runs as a physics backend inside the Isaac ecosystem.
I finally had time to dig through the examples, and the RJ45 plug simulation (example_contacts_rj45_plug.py) caught my attention. The latch deflects and clicks, and the cable behaves like a real 1D deformable. We've been working on cables, connectors, and other contact-heavy interactions, so seeing this as a first-class example was a pleasant surprise.
I extended it to test insertion and removal, with and without pressing the latch, and visualized the result in Rerun. Pull without pressing the tab, and the latch holds the plug in place; press it, and the plug comes free.
It also interested me how code-first the examples feel. We've spent real time trying to drive both MuJoCo and Isaac Sim with coding agents, and the recurring friction is being able to reason spatially to modify the simulation setup.
It feels like there's real potential here. I'm curious if Newton might be better suited for agent-generated robotics simulations.
If you're looking for something like this too, Newton is definitely worth a try.
Historic moment.
A few days ago, people were joking about Dario deciding who gets to use powerful LLMs.
Now it’s Trump.
Today it’s a universal ban. Tomorrow it will be specific groups and countries, chosen by the US government, that are allowed to access this technology.
In awe of SpaceX and its story - past, present and the future. You can think about it in 10+ different ways and continue re-blowing your mind in circles. Huge congrats to the team! 🚀
@DominiqueCAPaul Love the Design!
Be sure to check Google PageSpeed Insights though, some small improvements in regards to accessibility and performance are still possible.
Dominique is an impressive hands-on builder, love to see it, follow him if you are not already!
Which other builders from Switzerland and the DACH area are on twitter?
First checkpoint out of 10 evaluated.
> All evals at night to keep lighting consistent across multiple hours
> I’m doing 40 trials per checkpoint (one box)
> Boxes are in same position for all checkpoint. I marked the position with tape.
The space data center thing is great example of normal people not having read the proper sci-fi cannon.
To them it seems needlessly complicated and dumb. To those of us with sufficient sci-fi psychosis, it's perfectly logical and even crucial.
@10_X_eng@NVIDIARobotics Same here, I only recently went down that rabbit hole 😄 Isaac Sim is pretty impressive once you see how much robotics tooling NVIDIA has built around it.
I built RoboOps, a small fleet-operations prototype around @NVIDIARobotics NVIDIA Isaac Sim.
The demo connects an AWS-hosted operator UI with an Isaac Launchable running on an L40S GPU.
The repo is in the comments.
@adelbucetta@NVIDIARobotics Exactly. The stack is only useful if the boundaries between components are clear.
The GPU Infra side still matters here, though. Isaac Sim has serious runtime requirements, especially for more complex scenes. In this setup, I used an L40S with around 48GB of GPU memory.