←
Tuomas Haarnoja46 posts
Tuomas Haarnoja avatar✓

Tuomas Haarnoja ✓

@haarnoja

Co-Founder of Eka Robotics, previously Research Scientist @DeepMind, PhD @berkeley_ai and Student Researcher @GoogleAI

◷ Joined August 2017▧ 6 Photos and videos
Pulkit Agrawal avatar
⟳ Tuomas Haarnoja retweeted
Pulkit Agrawal@pulkitology·Oct 20

@EkaRobotics has raised a $150M Series A. Generality v/s Performance is the fundamental tension at the frontier of robotics. Generality without performance is just a prototype. Performance without generality is just a single-purpose tool. Customers shouldn’t have to choose. Eka breaks the tradeoff. Powered by our Vision-Force-Action (VFA) model, we deliver the performance and versatility customers need to put robots to work every day. Eka is built on three convictions: ⚡️Force is the language of manipulation. 🚀 Reinforcement learning is how dexterous manipulation scales. 🎂 System 1—fast, reliable physical execution—is the bottleneck to solve. With this funding, we will scale our model and accelerate customer deployments. Grateful to be building alongside a dream team and investors who share our ambition.

0:24
Pulkit Agrawal avatar
⟳ Tuomas Haarnoja retweeted
Pulkit Agrawal@pulkitology·May 8

Safe human-robot interaction with Eka's force-first approach to robotics: Vision-Force-Action Models (VFA).

0:05
Humanoids daily avatar
⟳ Tuomas Haarnoja retweeted
Humanoids daily@humanoidsdaily·Apr 29

🧵In the world of physical AI, we’re used to seeing videos sped up to hide how slow robots actually are. @EkaRobotics is doing the exact opposite. They’re showing their robots at 1/25x speed because the movement is so fast and the force-sensitivity so precise, you’d miss the "superhuman" dexterity in real-time.

0:14
Tuomas Haarnoja avatar
Tuomas Haarnoja@haarnoja·Apr 29

I'm really proud to finally share what we've been up to at Eka. Manipulation is all about forces (literally!), and we've built the first force-native robot learning stack. It's time to bring robotics back to its roots. 🤖🦾

0:27
Pulkit Agrawal avatar
⟳ Tuomas Haarnoja retweeted
Pulkit Agrawal@pulkitology·Apr 29

Eka means unity -- “one,” in Sanskrit and “first” in Finnish. We’re building intelligence for the physical world in its native language: forces. Until now, robotics faced a tradeoff — generality or speed. The real world requires both. Robotics also faced a data problem. Our Vision–Force–Action (VFA) model — the first of its kind — breaks the generality-speed tradeoff and the data barrier. It's a new foundation uniting performance, generality, and safety for putting capable robots in everyone's hands. Today, I am excited to share our journey of pushing robots beyond human limits. Today, dexterity becomes scalable. Today, I welcome you to the Era of Eka. Co-founded with @haarnoja, and so thrilled and grateful to be working with a dream team at @EkaRobotics. Learn more: ekarobotics.com

0:27
Pulkit Agrawal avatar
⟳ Tuomas Haarnoja retweeted
Pulkit Agrawal@pulkitology·Apr 28

Forces are the language of the physical world. At @EkaRobotics, we're building a Vision-Force-Action Model (VFA).

0:18
Markus Wulfmeier avatar
⟳ Tuomas Haarnoja retweeted
Markus Wulfmeier@m_wulfmeier·8 May 2024

Can deep reinforcement learning enable autonomous #robots 🤖🦿 to play real-world soccer ⚽️? sites.google.com/view/vision… Thrilled to share our latest step towards learning multi-agent robot soccer purely with onboard computation and sensing. We're extending prior motion capture based work that you might have seen in our @SciRobotics publication earlier this year @GoogleDeepmind science.org/doi/10.1126/scir… Why is this exciting?! A thread 🧵

Tuomas Haarnoja avatar
Tuomas Haarnoja@haarnoja·11 Apr 2024

Very proud to be part of this team! And looking forward to seeing more robots being... well, less robotic

0:27
Tuomas Haarnoja avatar
Tuomas Haarnoja@haarnoja·27 Apr 2023

We just released our work on robot soccer. I've been working on this for quite some time with my amazing colleagues at DeepMind. It's exciting how deep RL can produce such beautiful behaviors with low-cost robots. Full paper is available at arxiv.org/abs/2304.13653 Enjoy!

1:02
Tuomas Haarnoja avatar
Tuomas Haarnoja@haarnoja·17 Apr 2023

A sneak peek to our latest work; more details to follow soon. Stay tuned.

2:21
Sergey Levine avatar
⟳ Tuomas Haarnoja retweeted
Sergey Levine@svlevine·22 Jul 2019

L2 distance is a bad proxy for how long it takes to reach a goal. Can we learn dynamical distances? DDL can shape rewards and help learn from preferences. Below: robot learning w/ DDL from images w/o reward, only prefs. sites.google.com/view/skills… w K Hartikainen, X Geng, @haarnoja