Exciting week at #GTC. Physical AI is hitting an inflection point: scaling laws are emerging, but data pipelines are still bottlenecked by human labor.
Palatial is the only fully automated SimReady asset generator, create articulated objects ready for robot training in less that 10 minutes.
V0.2 is rolling out next week to wave 3 on our waitlist. Try it out, sign up at palatial.cloud/join
@TheRobotSimGuy@stevenRSim@nvidiaomniverse Amazing to see PhysReady assets working seamlessly in Isaac Sim! The fork-to-dishwasher manipulation is exactly the kind of real-world task our pipeline enables. Thanks for sharing the full workflow - this is the future of sim asset generation 🚀
Automated real2sim pipeline is finally real
by @PalatialSim
I tested it with Isaac Sim and Franka Panda Arm— placing a fork into a dishwasher.
Both assets? Generated from scratch by Palatial in minutes. @StevenRSim@PalatialSim@nvidiaomniverse
Here's the full workflow:
Sunday AI: robots don't need teleoperation anymore to perform well. Data collection for big models such as VLA is changing from teleoperation to egocentric human collected data!
#teleoperation#egocentric#datacollection
What is Newton?
The new physics engine created by Nvidia, Deepmind and Disney! It is ready, it is better than anything else available in simulation space! Begin your jouney with Newton!
#Simulation#Roboticstraining#physicsengine
Which factory size should you target for VLA data collection?
A) Large factories
B) Medium-sized factories
C) Small factories
Take a guess!
The answer is... C! Small factories.
[REVEAL - Explain why]
Here's the thing...
Small factories need cash.