@srigopal@positronic_ro@lukas_m_ziegler@nebiusai You can explore the datasets on https://t.co/tWuqgWJGk0. It has pretty detailed labels with timing, including grasping, controlling and fallbacks. Happy to chat more or give you API to try your models if you are interested
Robotic manipulation is far from solved—and the data proves it.
Positronic Robotics has teamed up with Nebius to launch the Physical AI Competition, a global open challenge bridging simulation to real hardware ($20K+ in GPU/evaluation credits).
With https://t.co/J6tmoO9lT3 showing top open models topping out at ~70% task completion, the playing field is wide open for builders.
Read the full breakdown & register: https://t.co/GaCDxfJRTh
@ThomasScott_Co@lukas_m_ziegler@nebiusai@positronic_ro there will be fewer tasks and different set of them (despite preserving overall DROID shape), nevermind the sim-real diff. Direct comparison would be hard. But relative order is interesting: more overfitted models will suffer more
There are quite a few submissions beating the best baseline on the MolmoSpaces sim we have for the first stage. Looking forward to seeing how adding RoboLab to the mix will shuffle the results.
JUST IN:
@nebiusai and @positronic_ro just launched a Physical AI Competition, and the prize pool is $20K+ in GPU and evaluation credits.
On the DROID: Everyday Tasks leaderboard at https://t.co/zX8nRG9j6c, over 1,000 episodes, 100+ common tasks, all on the same Franka arm — the best open models complete only ~70% of tasks on average. The weakest: ~12%.
Robotic manipulation is far from solved. And the gap between models is wide enough for a small team with a good idea to make a real difference.
That’s the whole point of this competition.
→ Open to all teams — register at https://t.co/rpI90l4hRj
→ First 50 teams to submit in week one get $100 in Nebius GPU credits
→ Models evaluated first in simulation
→ Top 10 advance to testing on a real physical robot arm
→ Competition runs October to December 2026
→ Powered by Nebius GPU infrastructure and Physical AI Workbench
The historical parallel the organisers draw is apt. In the 2004 DARPA Grand Challenge, no autonomous vehicle finished the course. A year later, five did. Open competitions with shared rules create turning points.
Robotic manipulation needs its DARPA moment.
🔗 Register here: https://t.co/qyaS0lbISh
~~
♻️ Join the weekly robotics newsletter, and never miss any news → https://t.co/GoA3ZuwWF9
Is your model a top model?
Submissions for the @nebiusai × Positronic Physical AI Competition are open!
Submit your policy to run in sim first. The top 10 models advance to run on a physical robot arm in our lab.
➡️ https://t.co/X9edZslLT5