The robots are coming faster than you think

Bernt Bornich says 1X will launch the early adopter program this year, for their home-use humanoid robot, NEO. He sets expectations right by saying it’s going to be useful but not perfect. “What you’re buying here is kind of a ticket to be part of this transformation.”

DHL is scaling robotics to meet the future of supply chains! ⛓️‍💥 The logistics giant just announced a £550M (€637M) investment to expand automation across UK & Ireland operations, supporting fast growth in e-commerce and life sciences healthcare. 🇬🇧 By 2030, DHL plans to

imo a problem in early-stage robotics: > Awesome demos on X > Yet, in reality, there is a big deployment gap (i.e. not as autonomous + scalable as in demo) > Very few can talk about the deployment gap (undermines reputation) > deployment gap stays underserved, harder to scale

🎉 Very happy to have our paper on predictive red teaming accepted to Conference on Robot Learning! Writing a first-author paper again was a good reminder of all the hard work students put in on details that I've forgotten about over the years. Project page: rb.gy/buih3o

Introducing Masquerade 🎭: We edit in-the-wild videos to look like robot demos, and find that co-training policies with this data achieves much stronger performance in new environments. ❗Note: No real robots in these videos❗It’s all 💪🏼 ➡️ 🦾 🧵1/6

We are continuing to tap into the massive resource of in-the-wild human videos for teaching our robots how to manipulate the world around them. In Masquerade 🎭 we replace the human hands with robot hands and train a visual encoder that is far more effective than prior work. 👇

Learning from human videos is all the rage in robotics right now. But, to truly create large datasets, we need methods like Masquerade that can scale pretraining to in-the-wild videos on the internet!

A day in the life of our warehouse robots: 8-hr shift, 1,000+ orders packed

Two things to note: 1) This isn’t a cherry-picked demo. Our robots have been running 5+ days a week for months now. That’s the bar for deploying robots into the world. 2) Part of what makes this possible is our application-specific hardware+software “wrapper” that bridges the

Full episode dropping soon! Geeking out with Ritvik Singh on DextrAH-RGB: Visuomotor Policies to Grasp Anything with Dexterous Hands dextrah-rgb.github.io Co-hosted by Michael Cho - Rbt/Acc Chris Paxton

Dexterous manipulation is a challenging task for robots, in no small part because it's hard to collect data for this task at scale. One option is to use simulation; simulation data, after all, scales readily with compute, and is much easier to collect. We talked to Ritvik Singh

Physical AI Weekly Report 🧵#2 Several cool videos this week from Ultra, Figure, and Boston Dynamics Plus a few other updates 👇

First from Ultra showing off their warehouse robot packaging over 1,000 packages in an 8-hour shift. This video highlights the potential for significant labor efficiency in warehouses where labor shortages are a persistent issue and where demand will only increase in

Carbon Robotics posts a mesmerizing video showcasing their LaserWeeder G2 technology which uses lasers to precisely eliminate weeds in cilantro crops, demonstrating its ability to target weeds with sub-millimeter accuracy without disturbing the soil or crops, a significant

Top AI Papers of The Week (August 18-24): - ComputerRL - Beyond GPT-5 - Chain-of-Agents - Parallel Text Generation - Retrieval-Augmented Reasoning - Has GPT-5 Achieved Spatial Intelligence? - Open Foundations for Compute-Use Agents Read on for more:

The NVIDIA Jetson Thor is here. 🎉 This powerful new robotics computer is designed to power the next generation of general and #HumanoidRobots in manufacturing, logistics, construction, healthcare, and beyond. It’s a massive leap forward for physical AI. Early adopters

“Fold the towels”