Co-founder @DynaRobotics Prev: @GoogleDeepMind, @NVIDIAAI, @Penn, @Harvard.
- Tweets 1,059
- Following 994
- Followers 123,272
- Likes 7,680
Jason Ma retweeted
Sharing new work on understanding articulated object manipulation from monocular human videos!
We explore how far SOTA vision models combined with optimization can take us for articulated object reconstruction and dexterous re-targeting
track-articulate-act.github.…
3
19
102
8,446
Eureka was so ahead of its time
GPT-6 Astra test 4/n
Prompt: "Implement pen spinning with a dexterous hand. Use Isaac Lab for RL training, use the Sharpa hand, and create the pen mesh yourself. Give me a trained RL policy and a visualization video. You are free to search the web and download papers or anything else you need."
After running autonomously for a day and a half (including training the policy), this is what we got:
Credit: my student Chengyang Li
8
12
316
55,855
Now, with Astra, the possibilities are so much greater, and some new paradigms will emerge again. What an exciting time for robotics
4
548
Later, we followed by taking the idea to the real world and taught a robot dog how to balance on an inflated yoga ball!
Introducing DrEureka🎓, our latest effort pushing the frontier of robot learning using LLMs!
DrEureka uses LLMs to automatically design reward functions and tune physics parameters to enable sim-to-real robot learning. DrEureka can propose effective sim-to-real configurations for several robots and tasks, and we even got a bit creative with it: Let’s make a robot dog walk and balance on a yoga ball! Check out these fun videos, and follow the thread for a deep dive!
5
990
3 years ago, we just let GPT-4 keep cooking in Isaac Gym and got some amazing pen spinning results! by today’s terminology, it was perhaps the first rsi in robotics?
Can GPT-4 teach a robot hand to do pen spinning tricks better than you do?
I'm excited to announce Eureka, an open-ended agent that designs reward functions for robot dexterity at super-human level. It’s like Voyager in the space of a physics simulator API!
Eureka bridges the gap between high-level reasoning (coding) and low-level motor control. It is a “hybrid-gradient architecture”: a black box, inference-only LLM instructs a white box, learnable neural network. The outer loop runs GPT-4 to refine the reward function (gradient-free), while the inner loop runs reinforcement learning to train a robot controller (gradient-based).
We are able to scale up Eureka thanks to IsaacGym, a GPU-accelerated physics simulator that speeds up reality by 1000x. On a benchmark suite of 29 tasks across 10 robots, Eureka rewards outperform expert human-written ones on 83% of the tasks by 52% improvement margin on average. We are surprised that Eureka is able to learn pen spinning tricks, which are very difficult even for CGI artists to animate frame by frame!
Eureka also enables a new form of in-context RLHF, which is able to incorporate a human operator’s feedback in natural language to steer and align the reward functions. It can serve as a powerful co-pilot for robot engineers to design sophisticated motor behaviors.
As usual, we open-source everything! Welcome you all to check out our video gallery and try the codebase today: eureka-research.github.io/
Paper: arxiv.org/abs/2310.12931
Code: github.com/eureka-research/E…
Deep dive with me: 🧵
1
24
6,526
Jason Ma retweeted
"Folding towels is a real operational bottleneck. We fold around 3,600 towels a day and Dyna is meeting that throughput."
- Hotel Manager, Best Western
Get ready for exponential growth in commercial robot deployments.
Dyna Robotics
6
22
167
13,195
Jason Ma retweeted
Love to see this! I want to see more bragging about deployments from everyone
After so many demos, models, pilots, robots are still struggling to land real deployments with real customers. Until now.
Today we’re excited to share that our robots have successfully crossed the ROI threshold, and Din Tai Fung, one of the highest revenue per location restaurant chain in the US, is rolling out Dyna robots across its extensive restaurant network.
This brings our rollouts across hotels, logistics, data centers, and many other use cases to a fleet that reaches hundreds of robots by the first half of 2027. And we’re just getting started.
It’s been a wild year, and today we’re double clicking on the battlefield stories and sharing a few learnings about scaling robot deployments. We are just scratching the surface.
Read the full blog post: dyna.co/news/scaling-custome…
2
3
30
4,260
Jason Ma retweeted
Generational moment
After so many demos, models, pilots, robots are still struggling to land real deployments with real customers. Until now.
Today we’re excited to share that our robots have successfully crossed the ROI threshold, and Din Tai Fung, one of the highest revenue per location restaurant chain in the US, is rolling out Dyna robots across its extensive restaurant network.
This brings our rollouts across hotels, logistics, data centers, and many other use cases to a fleet that reaches hundreds of robots by the first half of 2027. And we’re just getting started.
It’s been a wild year, and today we’re double clicking on the battlefield stories and sharing a few learnings about scaling robot deployments. We are just scratching the surface.
Read the full blog post: dyna.co/news/scaling-custome…
3
3
29
4,602
Jason Ma retweeted
Dyna is equal parts research and commercial. They've been sharing more about their research recently, and I'm glad they're now sharing more of what they've been up to on the commercial side.
After so many demos, models, pilots, robots are still struggling to land real deployments with real customers. Until now.
Today we’re excited to share that our robots have successfully crossed the ROI threshold, and Din Tai Fung, one of the highest revenue per location restaurant chain in the US, is rolling out Dyna robots across its extensive restaurant network.
This brings our rollouts across hotels, logistics, data centers, and many other use cases to a fleet that reaches hundreds of robots by the first half of 2027. And we’re just getting started.
It’s been a wild year, and today we’re double clicking on the battlefield stories and sharing a few learnings about scaling robot deployments. We are just scratching the surface.
Read the full blog post: dyna.co/news/scaling-custome…
2
5
21
4,738
Jason Ma retweeted
Congrats Dyna on real world deployment.
Too many companies hyping and too few actually shipping.
Dyna Robotics
2
5
21
2,262
Jason Ma retweeted
SOTA Metric: Happy customers 👀
After so many demos, models, pilots, robots are still struggling to land real deployments with real customers. Until now.
Today we’re excited to share that our robots have successfully crossed the ROI threshold, and Din Tai Fung, one of the highest revenue per location restaurant chain in the US, is rolling out Dyna robots across its extensive restaurant network.
This brings our rollouts across hotels, logistics, data centers, and many other use cases to a fleet that reaches hundreds of robots by the first half of 2027. And we’re just getting started.
It’s been a wild year, and today we’re double clicking on the battlefield stories and sharing a few learnings about scaling robot deployments. We are just scratching the surface.
Read the full blog post: dyna.co/news/scaling-custome…
1
6
14
2,098
Jason Ma retweeted
Real world deployments >> demos
After so many demos, models, pilots, robots are still struggling to land real deployments with real customers. Until now.
Today we’re excited to share that our robots have successfully crossed the ROI threshold, and Din Tai Fung, one of the highest revenue per location restaurant chain in the US, is rolling out Dyna robots across its extensive restaurant network.
This brings our rollouts across hotels, logistics, data centers, and many other use cases to a fleet that reaches hundreds of robots by the first half of 2027. And we’re just getting started.
It’s been a wild year, and today we’re double clicking on the battlefield stories and sharing a few learnings about scaling robot deployments. We are just scratching the surface.
Read the full blog post: dyna.co/news/scaling-custome…
2
1
12
2,181
Jason Ma retweeted
3 weeks in, this blog is the least technically or academically interesting, yet it's the one that matters most: people.
We're building intelligence that's useful to people, and robots that the general public understands and wants.
Read the full blog to see why customers love Dyna robots: dyna.co/news/scaling-custome…
And we are hiring!
After so many demos, models, pilots, robots are still struggling to land real deployments with real customers. Until now.
Today we’re excited to share that our robots have successfully crossed the ROI threshold, and Din Tai Fung, one of the highest revenue per location restaurant chain in the US, is rolling out Dyna robots across its extensive restaurant network.
This brings our rollouts across hotels, logistics, data centers, and many other use cases to a fleet that reaches hundreds of robots by the first half of 2027. And we’re just getting started.
It’s been a wild year, and today we’re double clicking on the battlefield stories and sharing a few learnings about scaling robot deployments. We are just scratching the surface.
Read the full blog post: dyna.co/news/scaling-custome…
4
3
14
2,494
Jason Ma retweeted
A robot that can make money is the truth... It's not just about collecting real data and training models. Entering reality means truly crossing the threshold of large-scale deployment.
After so many demos, models, pilots, robots are still struggling to land real deployments with real customers. Until now.
Today we’re excited to share that our robots have successfully crossed the ROI threshold, and Din Tai Fung, one of the highest revenue per location restaurant chain in the US, is rolling out Dyna robots across its extensive restaurant network.
This brings our rollouts across hotels, logistics, data centers, and many other use cases to a fleet that reaches hundreds of robots by the first half of 2027. And we’re just getting started.
It’s been a wild year, and today we’re double clicking on the battlefield stories and sharing a few learnings about scaling robot deployments. We are just scratching the surface.
Read the full blog post: dyna.co/news/scaling-custome…
1
12
59
8,157
Jason Ma retweeted
2026 is shaping up to be the inflection point for robotics.
We’re finally seeing the shift from impressive demos to real-world deployments.
These deployments will change the trajectory of the entire industry.
After so many demos, models, pilots, robots are still struggling to land real deployments with real customers. Until now.
Today we’re excited to share that our robots have successfully crossed the ROI threshold, and Din Tai Fung, one of the highest revenue per location restaurant chain in the US, is rolling out Dyna robots across its extensive restaurant network.
This brings our rollouts across hotels, logistics, data centers, and many other use cases to a fleet that reaches hundreds of robots by the first half of 2027. And we’re just getting started.
It’s been a wild year, and today we’re double clicking on the battlefield stories and sharing a few learnings about scaling robot deployments. We are just scratching the surface.
Read the full blog post: dyna.co/news/scaling-custome…
1
4
37
4,406
Jason Ma retweeted
Most Physical AI companies are still doing lab demos. A key factor in our investment in Dyna last year was their world class post training expertise/results and their deployment focus.
More deployments begets better data begets better models begets faster time to deployment
We saw this loop play out for multimodal models like ChatGPT/Claude and autonomous driving and we're seeing it play out for robotics
5
5
50
13,056
This is our release that I am actually most excited about because it shows and proves what truly matters. Deployment is the ultimate prize and the only reliable eval for robotics. After having worked on so many research projects and models in my career, where I saw so many new capabilities I had never seen before, there was always a question in the back of my mind: do they matter and how?
At a time in robotics where the signal-to-noise ratio is so low, where there’s so many demos, models, pilots, you see everyday, the conclusion I arrived at has always been shipping our models, robots, and entire systems to real customers and proving that they actually fulfill a need for real people. Showing a demo is easy, proving that the robots create sustainable value for customers over a long period of time is extremely hard. We know it's hard, but we also know it's necessary to get right. This mindset is also why I decided to start Dyna with @Lindon_Gao and @YorkYang5050 in the first place, because our heart has been focusing on the right problems from day 1. It’s great to see the whole team’s effort culminating in this first scaled deployment release of its kind for the industry. But we are just getting started, and I am more excited about the future of the physical economy than ever before.
After so many demos, models, pilots, robots are still struggling to land real deployments with real customers. Until now.
Today we’re excited to share that our robots have successfully crossed the ROI threshold, and Din Tai Fung, one of the highest revenue per location restaurant chain in the US, is rolling out Dyna robots across its extensive restaurant network.
This brings our rollouts across hotels, logistics, data centers, and many other use cases to a fleet that reaches hundreds of robots by the first half of 2027. And we’re just getting started.
It’s been a wild year, and today we’re double clicking on the battlefield stories and sharing a few learnings about scaling robot deployments. We are just scratching the surface.
Read the full blog post: dyna.co/news/scaling-custome…
11
13
115
12,067
Jason Ma retweeted
The heroes behind deployment are often underrated.
A strong, general model is not enough to get robots into production. Anyone who has really deployed knows that most of the important lessons only show up in the field — and you need serious infrastructure to capture them.
At Dyna, we’ve spent a lot of time building observability, auto-labeling, feedback, and evaluation systems that turn deployment into a continuous learning loop, with humans guiding the loop where they add the most value.
That loop is critical. It tells us where models actually fail, what needs to be solved fundamentally, and where our foundational research should go next — instead of patching problems one deployment at a time.
Deployment isn’t just a business use case for us. It’s what powers the data and feedback flywheel behind scalable robotics.
More here:
dyna.co/news/scaling-custome…
There’s probably a lot more behind it than you’d expect.
An hour of lab evals catches a model that doesn't work. It won't catch one that fails once every two hundred trials, or degrades over a week, or runs fine on this robot and badly on the one beside it.
So the eval moved to where the work is. Every episode, every site, graded on the customer's definition of good. Over a terabyte a day, autolabelled into SOP steps, outcomes, and failure modes.
Today, a new Dyna deployment goes from setup to production ROI in as little as three days.
This is just scratching the surface.
Also - we're hiring across deployment research: dyna.co/careers
2
10
55
7,461
Jason Ma retweeted
So much to appreciate in this blog
“Deployment is the eval”
After so many demos, models, pilots, robots are still struggling to land real deployments with real customers. Until now.
Today we’re excited to share that our robots have successfully crossed the ROI threshold, and Din Tai Fung, one of the highest revenue per location restaurant chain in the US, is rolling out Dyna robots across its extensive restaurant network.
This brings our rollouts across hotels, logistics, data centers, and many other use cases to a fleet that reaches hundreds of robots by the first half of 2027. And we’re just getting started.
It’s been a wild year, and today we’re double clicking on the battlefield stories and sharing a few learnings about scaling robot deployments. We are just scratching the surface.
Read the full blog post: dyna.co/news/scaling-custome…
2
5
38
5,018