🚀 Robots on Sui: • Rent & crypto-tip robots • Remote control across...

Uploaded: 2025-09-22T02:14:55.000Z
Duration: PT104.163S
Channel: Kostas Kryptos

1:00:38

E155: Mysten and Sui – Stablecoins, Privacy, and Quantum Safety - All on One Chain Kostas Kryptos is the Co-Founder and Chief Cryptographer at Mysten Labs. Today, we discuss quantum-resistant blockchain, AI-powered robots, encrypted WhatsApp payments, and stablecoin-powered superapps. This conversation maps out the real future of Web3, privacy, and programmable finance. Timestamps: 0:00 Intro 1:31 Please Subscribe 1:56 Time Difference In Dubai vs U.S 3:47 Why You Don't Sleep 6:18 Changing This 8 Year Habit 6:52 How Did You Realize That This Wasn't Healthy 7:23 The Biggest Algorithm In Your Life 8:17 Payment Methods With Crypto Coming Faster Than We Think 10:35 What Is A Cryptographer Explained To Your Mom 11:48 Partnerships: Jupiter KAST 12:52 How Secure Are WhatsApp Messages From Governments POVs 18:18 Can You Debunk The Bitcoin Worry Of Quantum Computing 24:43 Are You Aware Of Bitcoin Or Ethereum Developers Working On This 26:08 Most Companies Lack Strong Cryptography Teams 27:52 Do You See This Happening With Mysten, Growing Too Big To Make Decisions Quickly 30:48 Why You Dedicated Your Life To Mysten & Building The SUI Network 32:59 Partnerships: Paradex Zashi is now Zodl Mantle 34:03 Something Others Don't Yet Understand About The SUI Network 38:28 Something SUI Is Uniquely Bad At 40:23 Where Is The One Big Breakthrough For SUI Located 47:12 What Is SUI Doing At The Intersection Of AI, Crypto, And Robots 50:03 Why SUI And No One Else With Robots 50:32 Partnerships: Trezor Sui 51:05 SUI's Technical Advantages For Robots - Larger Blocks & Walrus Storage 52:38 Programmable Tunnels For Off-Chain Robot Control 53:42 Crypto Fights, Betting, And Racing With Robots 54:28 Seal Protocol For Data Encryption & Monetization 55:18 Why SUI Is Better Suited For Complex Technologies Than Other Blockchains 56:35 Closing Remarks - The Most Passionate Co-Founder

MR SHIFT 🦁

46,120 views • 5 months ago

0:56

The whole Korea is talking Sui + Walrus + Robots. What made this possible? ↕ Sui's max transaction size > than most other chains, fitting 1024 programmable txns in one call. Unlikely you can do what you see with most other cryptos! ᯓ ✈︎ Sui's speed at < 500 msecs for the best, almost instant UX ⚡Sui's lightning tunnels, for real time remote control... rent and forget gas 🗄️ Walrus's data availability framework that stores telemetry, camera footage and AI models 🔳 microCPUs supporting Sui's SDK; the community played an important role here 📦 Sui's "everything is an object", by far better programming experience than other chains. Composability at its best! thanks to the 🧠 of Irvin John Naulty George Digkas

Kostas Kryptos

19,604 views • 9 months ago

0:51

Sui achieved real-time decentralized AI-enabled Robot Coordination. It is the fastest and most modern decentralized network for financial transactions, but some of its unique qualities extended that to something magical. Think about it: subsecond finality, Seal for privacy & time capsules, dynamic objects, storing AI models in Walrus, batching multiple commands in one transaction via PTB, Native Randomness, Arduino and ESP32 apis, (when required) zero latency via Sui Tunnels == Robot’s favorite playground! Sui is clearly ahead of any competition… I’m getting back to bed with a smile to dream the future. I know many claim “ah that other chain can also do something similar… theoretically”, Come on… “guys everything is possible, it’s modern programmability, cool features and a vision for practice Vs just theory that makes Sui unique and an easier platform to innovate from: don’t just talk, DO and SURPRISE YOURSELF is the motto”. 👀 Today’s Video Recording from Sui’s Hacker team. It’s real.

Kostas Kryptos

23,643 views • 11 months ago

0:56

Brain-controlled exoskeletons to train humanoid robots! 🧠 Fourier just presented human tele-operators using brain control interfaces and exoskeletal arms to train humanoid robots on home tasks. The brain control interface is the interesting part. Instead of using a controller or joystick to teleoperate, the operator's movements and intentions are captured more naturally through the exoskeleton and BCI. This means the demonstrations are more fluid, more human-like, and better suited for training robots to perform delicate home tasks. Multiple tele-operators are simultaneously generating training data across multiple robots. This is how you build the dataset needed for eventual full autonomy, without waiting years for it to arrive. This might be the bridge between "robots that work in controlled environments" and "robots that work in homes." Not full autonomy right away, but trusted human intelligence operating through a robot body, getting better with every task completed. ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →

Lukas Ziegler

20,898 views • 5 months ago

3:35

SOMEONE JUST BUILT AN OPERATING SYSTEM FOR ROBOTS not a framework... not a library an actual OS dimos is an open-source platform that lets you build, run, and deploy software on any robot humanoids, quadrupeds, drones, wheeled bots you write python modules .... they talk to each other via typed streams...just like microservices, but for physical hardware the wild part..? you can give your AI editor (cursor, claude code) MCP access to send real physical commands "move forward 1 meter" → robot moves. it ships with a full navigation stack SLAM, collision avoidance, route planning. plus a dashboard to debug, visualize, and control your robot in real-time via waypoints, VR, or keyboard. supported hardware already includes unitree G1 humanoids, dji drones, and multiple robot arms. programming physical robots is finally as simple as programming software

Vaishnavi

19,745 views • 3 months ago

1:41

SUI Atomic Agentic transactions demo’ed to Google Sui’s Key Innovation Highlighted: Programmable Transaction Blocks (PTBs) Sui’s architecture enables this atomic multi-transaction execution through Programmable Transaction Blocks (PTBs) a core feature of the Sui blockchain: • What PTBs do - They allow developers (or AI agents) to bundle multiple operations such as swaps, transfers, staking, payments to other agents, or settlements into one atomic transaction. If any step fails, the entire block reverts, preventing partial executions or inconsistent states. • Why this matters for AI agents Traditional blockchains often require sequential transactions (with risks of front-running, failures midway, or high gas costs for coordination). Sui’s PTBs enable agents to perform complex workflows in a single, fast operation (~400ms finality on Sui). • Agent-to-agent payments example - An AI shopping agent could coordinate with multiple merchant agents, initiate several purchases, handle payments (e.g., via stablecoins), and settle everything atomically. Or a DeFi agent could monitor opportunities, execute trades across protocols, stake proceeds, and transfer yields all in one go without intermediate risks. This is described as a first (or at least a leading implementation) among public blockchains for seamless, atomic agent-to-agent value transfer at scale. It’s particularly powerful in the AP2 context, where agents need to operate autonomously but reliably, with user-defined guardrails (e.g., spending limits, consent verification). This positions Sui as specialized infrastructure for the intersection of AI autonomy and on-chain finance, with the atomic execution capability as a standout differentiator demonstrated in real collaboration with Google.

MartyParty

20,189 views • 4 months ago

1:28

Teaching robots to work smoothly with humans! 👀 Soon enough, a robot will be standing in the garage holding a flashlight for your dad, and doing a better job of it than you ever did. 🔦 Carnegie Mellon University and The University of Texas at Arlington researchers introduced HALO, a new method for robots to collaborate better with humans. When robots work with humans, they face unpredictable behavior. Every person moves differently, makes different decisions, and reacts differently to the robot. Traditional learning approaches struggle because the robot and human are fundamentally different, they think and move in completely different ways. This mismatch causes the robot's learning to become unstable, like trying to balance on a wobbly chair. Mentioned research solves this by mathematically guaranteeing the robot's learning stays stable. Pushing objects together in sync, transporting items through tight spaces, and handling super-long objects that require coordination. Real humanoid-robot experiments show HALO handles tricky collaborative situations much better. How it works? It basically has 3 layers. Vision layer understands what needs to happen by looking at the scene. Coordination layer figures out how the robot and human should work together tactically. Control layer executes the movements with fast, stable whole-body control. 📜 Paper website: Congrats Ding Zhao + all co-authors! ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →

Lukas Ziegler

28,192 views • 2 months ago

2:31

The hardest problems in AI aren't research problems anymore. They're deployment problems. It’s how we actually deliver real value, today, to build the future people want. That’s why, after 20 years in AI, my next step was inevitable: make robots do useful work for and alongside people, right now. Today, I am delighted to announce the launch of Walden Robotics to tackle just that. We started this year and are coming out of stealth today with a $300M seed round backed by some of the most serious companies and investors in the world. They have seen firsthand our general-purpose robots being useful in production on day one, and getting better every day after. You can see a glimpse of what we've been building in the video below. Physical AI has gone through a rapid phase transition, in part thanks to pioneering research from my friends and co-founders Russ Tedrake , Ben Burchfiel , Siyuan Feng, Rareș Ambruș , and many others at Walden. But from our long experience working together with co-founders Kerri Fetzer-Borelli and Dave Johnson, we learned how hard it is to deploy cutting-edge AI in a real, live, incredibly sophisticated production environment with an intricate ballet of automation and human ingenuity. That’s why we deliberately created Walden Robotics as a full-stack, human-centric, customer-focused robotics company from the start: we seeded the company with a world-class team across hardware, software, AI, deployment, operations, product, and business talent, so we could continuously optimize our whole system end-to-end, deeply and purposefully, from real-world experience with real customers. The efficacy of this strategy speaks for itself: since February, our general-purpose robots have been doing useful work in production at a Toyota plant in North America, moving from first pilot to real work in under two months. Not a lab. Not a demo. Not a future promise. Real work on a real line, today, at one of the best large-scale manufacturers in the world, with general-purpose robots that get better every day. And this is just the beginning. Two ways to find us: If you run a manufacturing or logistics business and want robots that are widely useful now, not someday, let's talk. We own “ for a reason! And if you want to build them: we're hiring across the company, from software, to hardware, AI, ops, product, business, and more. In particular, as the Chief Strategy Officer at Walden, I am recruiting for three incredibly impactful founding roles to fuel our agent-native go-to-market engine. Check out Let’s build together!

Adrien Gaidon

59,664 views • 2 days ago

4:55

For the Sui Overflow 2026 hackathon, I built Talise. A gasless dollar account that sends money by name, settles in under a second, and keeps the amount private on-chain. This is what "money that moves like a message" actually looks like when you build it on the right rails. The core experience is simple by design. Sign in with Google/Apple, get a self-custodial wallet via zkLogin, claim your handle like sele@talise, and start sending. No seed phrase. No gas token. No 0x address to copy and verify. The complexity lives in the infrastructure. The user never sees it. [Kostas Kryptos] Sending is just a name. Type a handle, enter an amount, confirm. It arrives in under a second. Stablecoin transactions on Sui cost nothing, so what you type is exactly what arrives. No spread taken on the send. No fee waiting at the destination. Talise also has a built in privacy feature. The amount is hidden on-chain and the connection between sender and recipient is completely broken. Nobody watching the chain can see what moved or who it moved between. Real privacy, live on Sui mainnet today, not a line on a roadmap. Beyond the core send: claim links let you wrap dollars in a URL and drop it in any chat, no address needed to receive. Payment streaming lets you send a salary or a subscription by the second. Idle balances route automatically to on-chain lending so your dollars are never dead weight between payments. And cash-out to your local bank is live now in Nigeria, with more corridors coming. Every transaction is sponsored through a gas station so users never hold or spend a gas token. zkLogin, sponsored transactions, sub-second finality, and programmable transaction blocks are what made this possible. Sui is the only chain where this experience exists without compromise. Over 1,400 handles claimed on-chain. Every feature shipped and working in the live build. Testflight app: Talise. Money that moves like a message. Sui Community Official sarah 😾

Sele

13,211 views • 25 days ago

0:25

A robot that works with 1 to 6 legs? This reconfigurable robot adapts like a spider, switching between different movement styles on the fly. SnapBot isn’t just a legged robot: it’s a shape-shifting machine. With up to 700 configurations, it can walk, crawl, or row its way through any terrain. Why this is exciting: ✅ Changes its setup from one to six legs for different tasks ✅ Moves in multiple ways, including walking, rowing, and crawling ✅ Adapts to its environment like nature’s best climbers and crawlers ✅ A step toward robots that adjust to real-world challenges instantly Nature-inspired robotics is unlocking new ways to move. SnapBot proves that flexibility is key to the future of locomotion.

Ilir Aliu

78,006 views • 1 year ago

2:17

Most $SUI holders know one thing about the token: total supply is capped at 10 billion. They have never read the mechanic that makes that cap matter. It is called the Storage Fund. And it is the most important thing in the $SUI tokenomics docs that nobody is talking about. Here is exactly how it works: Every time a transaction adds data to the Sui blockchain, the user pays a storage fee. That fee does not go to validators directly. It goes into the Storage Fund, a pool of SUI that never fully depletes. Here is where it gets interesting. The Storage Fund has its own stake in the network. It earns staking rewards the same way every other stakeholder does. Those rewards are then distributed to validators to compensate them for storing historical data. This solves a problem every other blockchain ignores: When a new validator joins Sui, they have to store all the historical data from transactions that happened before they existed. Why would a new validator pay to store someone else's old data? The Storage Fund pays them for it. Past users who created the storage requirements in the first place funded the pool. Future validators get compensated from that pool indefinitely. The fund pays out only the returns on its capital, never the principal. It cannot be drained. It is designed to survive forever. Now here is the part that directly connects to $SUI token value. The Sui docs state this explicitly: Deflation is a feature of Sui, not a bug. Here is why: Total supply is capped at 10 billion SUI. As network activity increases, more transactions are processed. More transactions mean more storage fees flowing into the Storage Fund. As the Storage Fund grows, it holds more SUI. More SUI held in the fund means less SUI in active circulation. Less circulating supply against the same or growing demand means the value of each SUI token increases. Network growth directly reduces circulating supply. That is not speculation. That is the economic model built into the protocol at the architecture level. One more detail worth knowing: If you delete data you stored on chain, you receive a partial refund of your original storage fees. The system charges for storage, rewards deletion, and compounds the fund's stake indefinitely. Most people holding $SUI today are pricing the speed narrative: The parallel transaction processing. The sub-second finality. The Move language safety. They have not started pricing the storage fund deflation mechanic. That gap between what the tokenomics actually does and what the market currently understands is where the long-term thesis lives. The people who read the docs always buy before the people who read the price.

2xnmore

47,515 views • 2 months ago

2:21

Today, we’re thrilled to announce our $200M Series C funding round at a $1B valuation, led by RoboStrategy and existing investors including General Catalyst. Standard Bots is now America’s largest manufacturer of AI-native industrial robots. Our customers include Sunoco, Lockheed Martin, NASA, and the US Army along with hundreds of other manufacturers across the country. We’re proud to say that we’re on track to deploy 10% of all U.S. industrial robots by next year. We are expanding our Glen Cove, New York facility to 70,000 square feet to scale our vertically integrated production process. We currently design almost all our own parts, including our own actuators, and we assemble every final product in-house. By 2027, we’ll manufacture everything — from metal in to robots out — right here in America. We believe AI-native robots are the essential power tool of the 21st century — the tool that will grow American manufacturing and help every American worker to be a force at work. You just show your robot how it’s done, and it learns through demonstration. No coding, no consultants, just unbox and deploy faster than anything else on the market. Right now it’s possible for the United States to revitalize our manufacturing base if we become the worldwide leader in this transformative technology. We must build American robots, and put them to work in American factories. It’s a national imperative, and it’s our central mission. This fundraise gets us one step closer to the goal. The future of American manufacturing is bright! Join Standard Bots, and show your robot how it’s done — we’re just getting started.

Standard Bots

435,198 views • 1 month ago

49:15

E161: Sui: Institutions Are Coming - SUI is Ready! evan.sui the Co-founder & CEO of MystenLabs.sui - the company behind SuiNetwork and Walrus 🦭/acc. In this conversation Evan shares a breakdown of why most blockchains were built wrong and what actually needs to happen before institutions go all in. Timestamps: 0:00 Introduction 2:02 Evan’s Move To New York Explained 3:20 So Much Has Shifted In The Last Year 5:03 How Do You Deal With Unpredictability 6:18 What’s Changed The Most Since Our Last Conversation? 8:59 How Evan Feels About These Changes 10:59 Partnerships: Jupiter KAST 11:40 A One Size Fits All Approach Can’t Work For Every Complex Reason 16:40 Evan’s Realization That Everything Needing To Be Built Differently 18:11 Has SUI Officially Proven This Way To Be Correct 19:22 Considering Tokens Dropping In Value - What Has SUI Proven Despite This 23:10 How Do You Define “Product Market Fit” 24:53 Partnerships: Paradex Zodl (fka Zashi) Sumsub 26:05 What Does A Much Bigger Adoption Look Like 28:09 The Difference Between Institutions Understanding But Prolonging & Early Adopters 30:29 The Dynamic Between Institutions & The Chains They Build On 32:06 How Different Is It To Build For An Institutional User vs A Retail User 34:23 Building The “Better” Technology Is Never Enough 37:05 Do All Major Chains Converge Or Specialize 40:00 Partnerships: Trezor @BitwiseInvest Sui 40:57 What Is SUI Doing Differently From Ethereum Or Solana Explained Simply 43:10 An Example Of Assets That Are The Same vs Not The Same 46:11 SUI Didn’t Build Something Limited Like Facebook Libra Or DM, They Built Something Better 48:11 Closing Thoughts

MR SHIFT 🦁

80,505 views • 4 months ago

1:39:56

E114: Walrus 🦭/acc : Decentralizing Data Storage, the perfect complement to the SUI Network George Danezis is the Chief Scientist at MystenLabs.sui , co-creator of the Sui , and one the minds behind Walrus 🦭/acc Timestamps 0:00 Intro 2:00 Partnerships: Jupiter, Bitwise, SUI & Mantle 2:46 Sleeping Habits 4:05 Who is George? 4:23 What is a Chief Scientist? 5:16 Research at Microsoft vs. Now 6:59 What is “Good Research”? 8:49 Corporate Innovation vs. Reality 10:31 Why I Love Teaching Security 11:52 Self Custody with Trezor 12:44 Teaching vs. Real-World Application 16:28 What Makes an Idea Great? 17:51 The Idea That Never Worked Out 19:50 Being Late to Ideas, Not Early 20:56 The Essence of Peer-to-Peer Networks 23:50 Why We Need Peer-to-Peer Systems? 27:51 Why Decentralization Was Hard to Realize 31:09 Bitcoin's Game-Changing Impact 33:10 ETH Aha Moment 34:46 Facebook’s Conviction in Blockchain 37:39 Scaling DIEM/LIBRA to Sui 39:20 Biggest Issue at Facebook 40:20 Why Facebook Bet on Crypto 41:40 Why I Left Facebook 45:07 Balancing Passion and Pay 46:43 What is a Blockchain Good For? 51:42 Understanding Blockchain’s Limitations 53:21 CryptoKitties 54:36 Meme Coins: Culture vs. Scam 57:02 The Role of Norms in Crypto 1:00:32 What is the Sui Network? 1:02:32 Why Do We Need The Sui Network? 1:03:34 Layer 1 vs. Layer 2 1:05:07 Why Build a Layer One? 1:09:21 Top Layer One Competitors 1:10:45 Decentralization vs. Practicality 1:14:06 Minimum Decentralization 1:16:33 Data Storage & Its Biggest Problem 1:20:05 What is Walrus Protocol? 1:22:09 Why Walrus is Possible Now 1:25:35 Why Sui Network for Data Storage? 1:27:37 Vision of Walrus Protocol 1:28:59 Building a Neutral Tech Future 1:30:44 Using Existing Blockchains 1:33:03 Future of Decentralized Production 1:35:51 Prediction For The Next 12 Months 1:39:04 Concluding Remarks

MR SHIFT 🦁

111,082 views • 1 year ago

1:05

Experiments in progress. The one on the right has been learning for ~3 hours, the one in the middle for ~1 hour, and the one on the left just started a few minutes ago. The initial motivation for making the physical Atari was just to commit ourselves to a subset of algorithms that can make progress in this setup. This commitment rules out algorithms that require billions of samples to learn (or worse, require multiple environments running in parallel). Atari games are simple enough that we should be able to show learning on them in a short amount of time with no prior knowledge. Since then, I've realized that this setup is also a good way to compare different paradigms in robotics in a principled way. These paradigms are sim2real, learning from tele-operated data, and learning directly on the robots. So far, I have observed that getting sim2real to work reliably is hard. It requires tweaks that don't scale. Policies that can play perfectly in simulation fall apart because of latencies and the messiness of the real world. These aspects could be modeled to improve the simulation, but not without sinking significant human engineering hours. I have higher hopes for learning from tele-operated data, but that requires a human to learn the task first. These experiments are on my to-do list. I have to learn to play some of the games well through the robot. I’m half-decent at playing Pong and Ms Pacman now. Learning directly on robots is looking like the most promising approach. This approach takes away pesky distribution shifts and makes it possible to have algorithms that continually improve with more data and time without any human intervention. It feels great to let experiments run overnight and wake up to find improved policies. With learning on robots, I should, in principle, be able to go on a long vacation and come back to find better policies for complex tasks beyond Atari games. Whether that is possible with current learning algorithms is a different question.

Khurram Javed

52,110 views • 7 months ago

0:31

Japan Just Built a HouseBot You Control Without Speaking and It Changes Everything! Donut Robotics has officially unveiled its first bipedal humanoid, Cinnamon 1, and instead of focusing on louder voices or bigger motors, the company went in the opposite direction. Silence. Cinnamon 1 introduces what Donut Robotics calls Silent Gesture Control, a system that allows the humanoid to be guided using simple hand and finger movements rather than spoken commands. This approach feels especially well suited for real world environments where traditional voice control falls apart. Busy factory floors. Construction sites filled with constant noise. Even quiet indoor settings where voice commands feel awkward or intrusive. It also opens the door for far more accessible human robot interaction, particularly for users with impairments. While the current Cinnamon 1 hardware is built on an OEM platform, the intelligence driving it is where Donut Robotics is placing its long term bet. The team is actively developing custom Vision Language Action AI that allows the robot to interpret what it sees, understand intent, and respond with physical action. The goal is not just smarter robots, but robots that feel more natural. Even more ambitious is the company’s plan for full domestic production. Donut Robotics has stated its intention to localize both manufacturing and AI development in Japan, reinforcing the country’s reputation for precision engineering and thoughtful robotics design. If timelines hold, Cinnamon 1 units are expected to begin deployment in factories and construction environments by the end of 2026. That puts this humanoid squarely in the category of near term reality rather than distant concept. The takeaway is simple but important. As humanoid robots move out of labs and into daily work environments, the winners may not be the loudest or flashiest machines. They may be the ones that understand us without a word being spoken.

The AI Robot Guy on X

257,928 views • 5 months ago

2:44

China's humanoid robotics market is on fire. With orders expected to top 30,000 units this year—a tenfold jump from 2024's total of less than 3,000—2025 is officially shaping up to be the "Year of Mass Production." This surge, driven by an expansion into new sectors like industrial manufacturing, logistics, and elder care, is reflected in a wave of new deals across the industry. Here's a look at some of the key commercial progress: Astribot: A 1,000-unit order for industrial and logistics deployment over two years. TianTai Robotics: Signed a major 10,000-unit order for caregiving robots. Noetix Robotics : Received over 2,000 intent orders in one month, valued at over 100 million yuan, with a focus on education and commercial performances. AgiBot: Expects to ship thousands of units this year and tens of thousands in 2026. Unitree Robotics: Has orders for thousands of units and is one of the most visible products in the industry. UBTech: Aims to deliver 500 industrial humanoids in 2025, with educational robot orders already exceeding 300 units. Robot Era: Delivered over 300 units by July 2025 with 500 more on hand. TLIBOT: Has around 1,000 intent orders. Galbot: Secured orders for its supermarket security robot, Galbot, in 100 stores. AI² Robotics: Has nearly 500 orders for its general-purpose robots for industrial and public service scenarios. But here’s the crucial reality check. While the order boom is exciting, it doesn't automatically translate to fulfilled deliveries. Many companies lack the production capacity to keep up. A significant portion of these are "intent orders" or framework agreements, not guaranteed sales. Furthermore, the market is heavily B2B-focused, with consumer demand representing only about 5% of sales. Some orders are even symbolic, for public relations or strategic purposes. This “order frenzy” is a starting point, not the finish line. The true test for China's humanoid robot industry isn't who can secure the biggest order, but who can consistently deliver on it and build a stable market for the future.

RoboHub🤖

199,146 views • 10 months ago

1:12

🚨 BREAKING: Big news in the computer vision world! 🎥 Luxonis | Robotic Vision just dropped its new OAK 4 line, and it’s a big upgrade for edge computer vision. Instead of being “just a stereo camera,” OAK 4 is a fully standalone vision computer with 52 TOPS of on-device AI. Models run locally, depth is computed locally, and no external PC or cloud pipeline is required. This is why robotics teams love it: lower latency, lower cost, fewer failure points in the field. The hardware is built for the real-world. IP67, shock-resistant, wide-FOV RGB + stereo pair, IR projection, IMU, audio, and a patent-pending calibration system that keeps depth accurate even when conditions change. But the real move is the platform. With Luxonis Hub, you can deploy models, grab telemetry, push OTA updates, or collect data when performance drifts, all from a unified interface. It turns a single device into an end-to-end edge CV system. Most customers today in robotics are groups who just want something that works: AMRs, bin-picking systems, trailer-loading robots, and ag-tech. 🤖 And they all say the same thing, the appeal isn’t raw TOPS, it’s the all-in-one simplicity that lets them scale without building custom infrastructure. Feels like the direction edge vision has been waiting for: rugged hardware + high-throughput on-device compute + a real management layer. A next step toward “plug-and-deploy” perception for robots. 🔗 Find out more here: ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →

Lukas Ziegler

41,955 views • 7 months ago

0:58

I don’t know if we live in a Matrix, but I know for sure that robots will spend most of their lives in simulation. Let machines train machines. I’m excited to introduce DexMimicGen, a massive-scale synthetic data generator that enables a humanoid robot to learn complex skills from only a handful of human demonstrations. Yes, as few as 5! DexMimicGen addresses the biggest pain point in robotics: where do we get data? Unlike with LLMs, where vast amounts of texts are readily available, you cannot simply download motor control signals from the internet. So researchers teleoperate the robots to collect motion data via XR headsets. They have to repeat the same skill over and over and over again, because neural nets are data hungry. This is a very slow and uncomfortable process. At NVIDIA, we believe the majority of high-quality tokens for robot foundation models will come from simulation. What DexMimicGen does is to trade GPU compute time for human time. It takes one motion trajectory from human, and multiplies into 1000s of new trajectories. A robot brain trained on this augmented dataset will generalize far better in the real world. Think of DexMimicGen as a learning signal amplifier. It maps a small dataset to a large (de facto infinite) dataset, using physics simulation in the loop. In this way, we free humans from babysitting the bots all day. The future of robot data is generative. The future of the entire robot learning pipeline will also be generative. 🧵

Jim Fan

165,246 views • 1 year ago

Live Cam

Video Failed to Load

Kostas Kryptos

Anya Rossi• Live Now

0 Comments

Related Videos