Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

Robots Doing Stuff is back. Ready? Robots Doing Stuff #60 | Welding Policy Repetitive, quick, and precise. Automating this manufacturing welding task with Physical AI required the successful execution of all of the above in order to achieve a consistent weld quality at production-level speed. Visit to learn more... show more

Sanctuary AI

10,144 subscribers

32,886 views • 2 months ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

8 Comments

Akshay Bhagat2 months ago

Why do we need physical AI for welding steel in fixtures? Can't we just numerically program coordinates and adjust for drift every few thousand cycles. That's how most automotive companies have been doing it for decades now

Ferbin2 months ago

most failures aren't the welding. they're part setup. if the part sits wrong, does this adapt or just fail?

Eva Al Soofi2 months ago

👍😊

Lucas Gray-Joy2 months ago

I like the focus on production environments here. That's where the real learning loop gets built, especially for touch, manipulation, and sim-to-real transfer.

Arithmancy2 months ago

Physical AI is becoming more practical with every deployment.

GarethTheDreamer2 months ago

Sim-to-real transfer on those hydraulic hands is legit! Embodied AI finally moving past the demo phase. 🤖

SEAR2 months ago

Industrial welding is perfect

EgoScale2 months ago

Interesting demo. Curious how the system handles part setup variation or small fixture drift. Does it adapt in real time, or is the workspace still tightly controlled?

Related Videos

Robots Doing Stuff #61 | Bimanual Nut Pick and Threading This is a real customer task: one step in a longer multi-part sequence we’re training in our labs. Threading requires precise alignment, force control, and fine manipulation. These are foundational capabilities for industrial dexterity applications that are currently inaccessible with traditional automation. We teleoperate these tasks to collect the data that trains our Physical AI. Next stop: autonomous. Visit to learn more about our Physical AI-enabled automation solutions. #Dexterity #DexterousHand #RobotHand #Robotics #RobotsDoingStuff #SanctuaryAI

Robots Doing Stuff #61 | Bimanual Nut Pick and Threading This is a real customer task: one step in a longer multi-part sequence we’re training in our labs. Threading requires precise alignment, force control, and fine manipulation. These are foundational capabilities for industrial dexterity applications that are currently inaccessible with traditional automation. We teleoperate these tasks to collect the data that trains our Physical AI. Next stop: autonomous. Visit to learn more about our Physical AI-enabled automation solutions. #Dexterity #DexterousHand #RobotHand #Robotics #RobotsDoingStuff #SanctuaryAI

Sanctuary AI

38,370 views • 2 months ago

So many robotics companies trying to solve everything at once… Last year I talked to Nima Gard, Director of AI at Path Robotics about this… Doing one thing really really well. They went after welding. Today they’re introducing Rove™, which embodies Obsidian™ (Path's physical AI model for manufacturing) on a legged mobile platform. The result is the intelligence to autonomously weld with the mobility to bring the worker to the work. Path Robotics (Columbus, Ohio) has raised more than $300 million in venture capital to build autonomous AI-powered welding cells. I think the tech is pretty impressive and video really cool. Congrats guys! Credit: Path Robotics

So many robotics companies trying to solve everything at once… Last year I talked to Nima Gard, Director of AI at Path Robotics about this… Doing one thing really really well. They went after welding. Today they’re introducing Rove™, which embodies Obsidian™ (Path's physical AI model for manufacturing) on a legged mobile platform. The result is the intelligence to autonomously weld with the mobility to bring the worker to the work. Path Robotics (Columbus, Ohio) has raised more than $300 million in venture capital to build autonomous AI-powered welding cells. I think the tech is pretty impressive and video really cool. Congrats guys! Credit: Path Robotics

Ilir Aliu

23,536 views • 5 months ago

Physical AI is here—the next wave of AI and the foundation of the next industrial revolution. NVIDIA founder and CEO Jensen Huang outlines the full stack making it possible: Cosmos and Omniverse to develop physical AI in virtual worlds, Isaac and Newton where robots learn skills in physics-simulated gyms, and Jetson robotics computers where their intelligence runs. Together with Japan’s robotics and manufacturing leaders, NVIDIA is creating a new era of industrial automation.

Physical AI is here—the next wave of AI and the foundation of the next industrial revolution. NVIDIA founder and CEO Jensen Huang outlines the full stack making it possible: Cosmos and Omniverse to develop physical AI in virtual worlds, Isaac and Newton where robots learn skills in physics-simulated gyms, and Jetson robotics computers where their intelligence runs. Together with Japan’s robotics and manufacturing leaders, NVIDIA is creating a new era of industrial automation.

NVIDIA

134,471 views • 1 month ago

“Blue-collar jobs are safe from AI.” Not for long. This welding robot climbs vertical steel walls like Spider-Man. While everyone is watching AI replace coders and office work… industrial robots are quietly entering the physical world too. Welding. Grinding. Inspection. Rust removal. The scary part? Software automation was the easy phase. Now machines are learning to work in the real world.

“Blue-collar jobs are safe from AI.” Not for long. This welding robot climbs vertical steel walls like Spider-Man. While everyone is watching AI replace coders and office work… industrial robots are quietly entering the physical world too. Welding. Grinding. Inspection. Rust removal. The scary part? Software automation was the easy phase. Now machines are learning to work in the real world.

Manisha Mishra

221,351 views • 4 months ago

🚀 🔥 AgiBot deploys Real-World Reinforcement Learning (RW-RL) in industrial robotics with Longcheer Technology. Robots now learn new skills in tens of MINUTES (not weeks), adapt to variations autonomously, and reconfigure flexibly, solving rigid automation pain points in precision manufacturing. A giant leap of intelligent automation for precision manufacturing! #AgiBot #ReinforcementLearning #RealWorldRL #Robotics #AI #IndustrialRobotics

🚀 🔥 AgiBot deploys Real-World Reinforcement Learning (RW-RL) in industrial robotics with Longcheer Technology. Robots now learn new skills in tens of MINUTES (not weeks), adapt to variations autonomously, and reconfigure flexibly, solving rigid automation pain points in precision manufacturing. A giant leap of intelligent automation for precision manufacturing! #AgiBot #ReinforcementLearning #RealWorldRL #Robotics #AI #IndustrialRobotics

AGIBOT

131,860 views • 10 months ago

Robot dogs are now welders! 👨🏼‍🏭 Path Robotics launches a mobile welding robot on quadruped platform, called Rove. This quadruped robotic welding system is paired with their Obsidian physical AI model. Obsidian already delivers autonomous, adaptive welding inside fixed cells. Rove brings intelligent welding to the part rather than moving the part to a fixed workstation. Shipbuilding, heavy construction, and large-scale fabrication share a constraint, massive assemblies and workpieces that cannot be moved or fixtured inside welding cells. These conditions have limited welding automation in environments that need it most, just as skilled welder shortages reach critical levels. Legged robots have long been considered too unstable for precision welding. Rove changes that. Obsidian provides the perception and adaptability needed to operate in high-variability environments. Saronic Technologies, a leader in autonomous maritime vessels, is one of the first early adopters, integrating Rove into its physical AI-powered shipbuilding operations in Franklin, Louisiana. 🇺🇸 Welding requires millimeter-level precision on a moving quadruped platform, considered impossible until AI-based perception and control made it feasible. 🔥 Andrew Lonsberry gotta be honest, but a welding robot dog wasn't on bingo card for 2026. ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →

Robot dogs are now welders! 👨🏼‍🏭 Path Robotics launches a mobile welding robot on quadruped platform, called Rove. This quadruped robotic welding system is paired with their Obsidian physical AI model. Obsidian already delivers autonomous, adaptive welding inside fixed cells. Rove brings intelligent welding to the part rather than moving the part to a fixed workstation. Shipbuilding, heavy construction, and large-scale fabrication share a constraint, massive assemblies and workpieces that cannot be moved or fixtured inside welding cells. These conditions have limited welding automation in environments that need it most, just as skilled welder shortages reach critical levels. Legged robots have long been considered too unstable for precision welding. Rove changes that. Obsidian provides the perception and adaptability needed to operate in high-variability environments. Saronic Technologies, a leader in autonomous maritime vessels, is one of the first early adopters, integrating Rove into its physical AI-powered shipbuilding operations in Franklin, Louisiana. 🇺🇸 Welding requires millimeter-level precision on a moving quadruped platform, considered impossible until AI-based perception and control made it feasible. 🔥 Andrew Lonsberry gotta be honest, but a welding robot dog wasn't on bingo card for 2026. ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →

Lukas Ziegler

10,227 views • 5 months ago

🤖 WILD: THIS HUMANOID ROBOT CAN WELD LIKE A HUMAN Persona AI's Gen 1 humanoid performed hot-work welding in a live factory, taking on one of the most dangerous jobs in manufacturing. Humanoid robots are getting closer to doing what only humans could do.

🤖 WILD: THIS HUMANOID ROBOT CAN WELD LIKE A HUMAN Persona AI's Gen 1 humanoid performed hot-work welding in a live factory, taking on one of the most dangerous jobs in manufacturing. Humanoid robots are getting closer to doing what only humans could do.

Coin Bureau

61,081 views • 2 months ago

Announcing Axol: a dual-arm robot designed for teams working with physical AI. Made in America. Axol is for builders who believe robots should work, in the real world, not just staged environments, and that the future of physical AI should be open, not closed. #robotics #automation #ai #hardware

Announcing Axol: a dual-arm robot designed for teams working with physical AI. Made in America. Axol is for builders who believe robots should work, in the real world, not just staged environments, and that the future of physical AI should be open, not closed. #robotics #automation #ai #hardware

Almond Robotics

63,842 views • 3 months ago

Welding robots. The original use case for industrial automation—first deployed in the 1960s at GM. 60 years later, they dominate automotive body shops. Hundreds of robots in perfect sync, laying thousands of spot welds per vehicle. But step outside high-volume production and they're rare. Most fab shops still weld by hand. Why? Every new part needs skilled programming. Path planning. Torch angle. Travel speed. Wire feed. Gas flow. All tuned for each weld on each component. The economics only work if you're making thousands of the same thing. High-volume? Solved. High-mix, low-volume? Traditional welding robots fail here... but watch this space a few start ups are starting to fill the gap.

Welding robots. The original use case for industrial automation—first deployed in the 1960s at GM. 60 years later, they dominate automotive body shops. Hundreds of robots in perfect sync, laying thousands of spot welds per vehicle. But step outside high-volume production and they're rare. Most fab shops still weld by hand. Why? Every new part needs skilled programming. Path planning. Torch angle. Travel speed. Wire feed. Gas flow. All tuned for each weld on each component. The economics only work if you're making thousands of the same thing. High-volume? Solved. High-mix, low-volume? Traditional welding robots fail here... but watch this space a few start ups are starting to fill the gap.

Jack 🤖

78,283 views • 9 months ago

Humanoid robots are taking over the largest robotics and AI show in America. Grady Trimble got an inside look at the robot family stealing the show and how they are being used in the real world. "AI is enabling robots to do all kinds of work, stuff like this, and hospitals, in factories, in warehouses... It's not just entertainment," President of the the Association for Advancing Automation Jeff Burnstein says.

Humanoid robots are taking over the largest robotics and AI show in America. Grady Trimble got an inside look at the robot family stealing the show and how they are being used in the real world. "AI is enabling robots to do all kinds of work, stuff like this, and hospitals, in factories, in warehouses... It's not just entertainment," President of the the Association for Advancing Automation Jeff Burnstein says.

FOX Business

15,329 views • 2 months ago

FULL INTERVIEW: travis kalanick joins TBPN to discuss his new company Atoms, physical AI, Uber, and more: 01:18 - Why he's been building in stealth for 8 years 04:32 - Atoms and the future of physical AI 08:10 - Creating a culture of builders 12:05 - Lessons from Uber 24:30 - The vision for physical AI and robotics 31:15 - Why humans will be the main beneficiaries of AI 38:20 - Mining, autonomous robots, automation 47:05 - Why Travis moved to Texas

FULL INTERVIEW: travis kalanick joins TBPN to discuss his new company Atoms, physical AI, Uber, and more: 01:18 - Why he's been building in stealth for 8 years 04:32 - Atoms and the future of physical AI 08:10 - Creating a culture of builders 12:05 - Lessons from Uber 24:30 - The vision for physical AI and robotics 31:15 - Why humans will be the main beneficiaries of AI 38:20 - Mining, autonomous robots, automation 47:05 - Why Travis moved to Texas

TBPN

627,943 views • 6 months ago

🚨 Elon Musk G20 vision: AI could add $20 30 trillion to the global economy annually. By 2027, AI could handle virtually any digital task and reach Stockfish level at software development. Humanoid robots could potentially grow the economy 10× or more especially when robots begin building robots. The age of exponential automation is coming.

🚨 Elon Musk G20 vision: AI could add $20 30 trillion to the global economy annually. By 2027, AI could handle virtually any digital task and reach Stockfish level at software development. Humanoid robots could potentially grow the economy 10× or more especially when robots begin building robots. The age of exponential automation is coming.

X Ethan

22,331 views • 17 days ago

German robotics company NEURA Robotics has created NEURA Gym: a large-scale, physical AI gym and training ground where hundreds of robots, including the humanoid 4NE-1, learn through real-world interactions. The company is collaborating with the Neuraverse ecosystem to create the world's largest repository of physical training data, promoting the application of humanoid robots.

German robotics company NEURA Robotics has created NEURA Gym: a large-scale, physical AI gym and training ground where hundreds of robots, including the humanoid 4NE-1, learn through real-world interactions. The company is collaborating with the Neuraverse ecosystem to create the world's largest repository of physical training data, promoting the application of humanoid robots.

CyberRobo

73,731 views • 11 months ago

This is what Production-Ready Physical AI looks like. Yesterday, we announced our expanded Physical AI strategy, backed by a demonstration of a complex industrial wire plugging task completed with a 99.5%+ success rate and a 2.54-second cycle time. Why is this significant? "Manipulating a flexible wire into a moving target on a live conveyor is exactly the kind of contact-rich dexterity problem that has kept tasks like this out of reach for traditional automation ... Solving it required models built around performance from day one with reliability, cycle time, and safety measured against real production benchmarks. That's the bar Physical AI has to clear to matter in enterprise environments, and it's the bar we've now cleared." — Olivia Norton, CTO at Sanctuary AI. Read more:

This is what Production-Ready Physical AI looks like. Yesterday, we announced our expanded Physical AI strategy, backed by a demonstration of a complex industrial wire plugging task completed with a 99.5%+ success rate and a 2.54-second cycle time. Why is this significant? "Manipulating a flexible wire into a moving target on a live conveyor is exactly the kind of contact-rich dexterity problem that has kept tasks like this out of reach for traditional automation ... Solving it required models built around performance from day one with reliability, cycle time, and safety measured against real production benchmarks. That's the bar Physical AI has to clear to matter in enterprise environments, and it's the bar we've now cleared." — Olivia Norton, CTO at Sanctuary AI. Read more:

Sanctuary AI

30,139 views • 3 months ago

Physical AI is the biggest opportunity most people are sleepwalking past. "Physical AI is a $50 trillion market." - Jensen at The All-In Podcast Robots. Autonomous vehicles. Embodied AI. The infrastructure to power all of it. Nvidia isn't just selling GPUs anymore. They're building the OS for the physical AI era.

Physical AI is the biggest opportunity most people are sleepwalking past. "Physical AI is a $50 trillion market." - Jensen at The All-In Podcast Robots. Autonomous vehicles. Embodied AI. The infrastructure to power all of it. Nvidia isn't just selling GPUs anymore. They're building the OS for the physical AI era.

Shruti

51,545 views • 6 months ago

a16z co-founder Marc Andreessen on why the shift from software AI to physical AI will create the biggest industry in human history: AI today is software. You open ChatGPT on your phone, and there's no manufacturing component, no real generator of middle-class jobs. But Marc Andreessen 🇺🇸 argues that a second, far more consequential turn is coming: the shift to embodied physical AI. "There's another turn on AI that's coming, which is the turn to embodied physical AI, which is robotics." This transition, he points out, has already begun. Drones went from human-piloted to fully autonomous. Cars went from purely physical products to rolling computers that drive themselves. "If you ever have the misfortune of finding yourself in San Francisco, take a Waymo. It's amazing to live in a cyberpunk future reality where you're in literally a state-of-the-art self-driving car that's literally out of the Jetsons." He makes clear: this isn't science fiction anymore. The self-driving car really works, and it's only the beginning. The next frontier is humanoid robots. "The general purpose robotics thing is going to happen. And it's going to happen in the next decade and it's going to happen at giant scale." Andreessen shares Elon Musk's view that robotics could become the single biggest industry in the history of the planet: "There's going to be billions, tens of billions, hundreds of billions of robots of all shapes, sizes, descriptions running around doing all kinds of things." And all of those robots need to be designed and built. Which brings Andreessen to his core argument: the real opportunity isn't in trying to recover what was lost. It's in leaning forward into what's coming. "We don't try to get the old manufacturing jobs back. What we should do is lean hard into the manufacturing jobs of the future."

a16z co-founder Marc Andreessen on why the shift from software AI to physical AI will create the biggest industry in human history: AI today is software. You open ChatGPT on your phone, and there's no manufacturing component, no real generator of middle-class jobs. But Marc Andreessen 🇺🇸 argues that a second, far more consequential turn is coming: the shift to embodied physical AI. "There's another turn on AI that's coming, which is the turn to embodied physical AI, which is robotics." This transition, he points out, has already begun. Drones went from human-piloted to fully autonomous. Cars went from purely physical products to rolling computers that drive themselves. "If you ever have the misfortune of finding yourself in San Francisco, take a Waymo. It's amazing to live in a cyberpunk future reality where you're in literally a state-of-the-art self-driving car that's literally out of the Jetsons." He makes clear: this isn't science fiction anymore. The self-driving car really works, and it's only the beginning. The next frontier is humanoid robots. "The general purpose robotics thing is going to happen. And it's going to happen in the next decade and it's going to happen at giant scale." Andreessen shares Elon Musk's view that robotics could become the single biggest industry in the history of the planet: "There's going to be billions, tens of billions, hundreds of billions of robots of all shapes, sizes, descriptions running around doing all kinds of things." And all of those robots need to be designed and built. Which brings Andreessen to his core argument: the real opportunity isn't in trying to recover what was lost. It's in leaning forward into what's coming. "We don't try to get the old manufacturing jobs back. What we should do is lean hard into the manufacturing jobs of the future."

Big Brain AI

35,113 views • 2 months ago

“Do you know where a sidewalk begins and ends?” — a simple question from Niantic Spatial CEO, Inhi Cho Suh, that gets to the heart of what today’s AI is missing. For robots, this isn’t theoretical. To safely navigate something as everyday as a sidewalk, they need to understand boundaries, context, and change in real time. Learn about how Niantic Spatial is building the missing piece in Physical AI. #NianticSpatial #HumanX #GeospatialAI #Robotics #AI #PhysicalAI

“Do you know where a sidewalk begins and ends?” — a simple question from Niantic Spatial CEO, Inhi Cho Suh, that gets to the heart of what today’s AI is missing. For robots, this isn’t theoretical. To safely navigate something as everyday as a sidewalk, they need to understand boundaries, context, and change in real time. Learn about how Niantic Spatial is building the missing piece in Physical AI. #NianticSpatial #HumanX #GeospatialAI #Robotics #AI #PhysicalAI

Niantic Spatial 🌎

18,656 views • 5 months ago

Physical AI hits production speed on automotive assembly! 👀 Sanctuary AI just achieved 99,5%+ success on a wire plugging task at a Tier 1 automotive supplier, 2.54 seconds per task, validated against live production benchmarks. The system matched the customer's actual production line throughput. The task sounds simple. But if you look closer, it's not. Inserting flexible wires into moving targets on a conveyor requires contact-rich dexterity, force sensing, real-time adaptation. Materials shift. Targets move. This has stayed out of reach for traditional automation. Instead of waiting for humanoid hardware to mature, Sanctuary AI is deploying Physical AI on existing commercial robotic platforms (Universal Robots in this case). So their solution is fully hardware-agnostic. Deploy today on what you already have. Scale to next-gen humanoids later. Olivia Norton (co-founder, CTO): "Physical AI adoption is gated by AI that meets both performance and cycle time requirements. That's what customers are seeking, and that's what we are delivering." Looks like manufacturing, logistics, and other labor-constrained industries don't need to wait. Read more here: ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →

Physical AI hits production speed on automotive assembly! 👀 Sanctuary AI just achieved 99,5%+ success on a wire plugging task at a Tier 1 automotive supplier, 2.54 seconds per task, validated against live production benchmarks. The system matched the customer's actual production line throughput. The task sounds simple. But if you look closer, it's not. Inserting flexible wires into moving targets on a conveyor requires contact-rich dexterity, force sensing, real-time adaptation. Materials shift. Targets move. This has stayed out of reach for traditional automation. Instead of waiting for humanoid hardware to mature, Sanctuary AI is deploying Physical AI on existing commercial robotic platforms (Universal Robots in this case). So their solution is fully hardware-agnostic. Deploy today on what you already have. Scale to next-gen humanoids later. Olivia Norton (co-founder, CTO): "Physical AI adoption is gated by AI that meets both performance and cycle time requirements. That's what customers are seeking, and that's what we are delivering." Looks like manufacturing, logistics, and other labor-constrained industries don't need to wait. Read more here: ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →

Lukas Ziegler

45,883 views • 3 months ago

🚨 Breaking News: Solo Tech won at CES 2026. 🦾🏆 In a 10×10 booth, we showcased the highest density of working robots on the floor proving that the future of AI is physical, performant, and deployable today. Up close with real systems. No slides. No simulations. Just Physical AI in action. Alongside leaders shaping what’s next NVIDIA Robotics , Sharpa , and Boston Dynamics the message was clear: This is the ChatGPT moment for robotics. And it won’t happen behind screens. It will happen in Robotics Gyms spaces around the world where machines train, models adapt, and builders learn by doing. The era of AI that acts has begun. #BreakingNews #CES2026 #SoloTech #PhysicalAI #RoboticsGym #EdgeAI

🚨 Breaking News: Solo Tech won at CES 2026. 🦾🏆 In a 10×10 booth, we showcased the highest density of working robots on the floor proving that the future of AI is physical, performant, and deployable today. Up close with real systems. No slides. No simulations. Just Physical AI in action. Alongside leaders shaping what’s next NVIDIA Robotics , Sharpa , and Boston Dynamics the message was clear: This is the ChatGPT moment for robotics. And it won’t happen behind screens. It will happen in Robotics Gyms spaces around the world where machines train, models adapt, and builders learn by doing. The era of AI that acts has begun. #BreakingNews #CES2026 #SoloTech #PhysicalAI #RoboticsGym #EdgeAI

Dhruv Diddi

14,841 views • 8 months ago