How can we rigorously quantify the uncertainty of pretrained... perception models to provide end-to-end statistical safety assurances for robot navigation?Perceive With Confidence: Statistical Safety Assurances for Navigation with Learning-Based Perception 🧵show more

From seeing → to understanding → now to following.... In the previous video, we showed how the robot sees and navigates the world. This time, we’ve upgraded SR Agentic — it can now identify a designated person and autonomously follow them in real-time. Powered by perception + reasoning + adaptive navigation, the robot doesn’t just react - it tracks intent and moves with purpose.show more

F.03 can now walk up/down stairs purely using it's... onboard camera perception Our robots now walk from manufacturing when built to HQ This is trained end-to-end with reinforcement learning in simulationshow more

BEL, in Collaboration with WESEE, has successfully developed the... Advanced Autonomous Navigation & Control Software for unmanned Surface Vessels. 🇮🇳 A2NCS will enable USVs to navigate & operate independently with high precision, enhanced mission efficiency & safety.show more

How can we more effectively leverage robot data from... different embodiments for skill transfer? Excited to share that our new work, RoVi-Aug, has been accepted to Conference on Robot Learning as an oral paper！ WIth RoVi-Aug, you can augment an existing robot dataset into a different robot and different viewpoints. A policy trained on the augmented dataset can zero-shot deploy on the unseen target robot with significantly different camera angles! 🧵👇 🔗 Check out our paper:show more

Tennessee State University along with other HBCU’s are celebrating... after police had to escort us off campus for our safety. To be clear, the cops didn’t ask us to leave due to being there without permission. They decided after 10 minutes that for our safety, it’s best that we end the event. Due to a hostile mob, the cops were forced to escort us to our cars.show more

🤖 One predictive backbone, three distinct tasks, consistent gains:... a strong signal that investing in reusable world models is the right abstraction for robot learning. Toyota Research Institute (TRI) just ran the same idea three different ways—and it worked each time. Using our NVIDIA Cosmos Predict 2–style world models, they hit SOTA with dynamic view synthesis, teleop data augmentation (even up against MimicGen as an upper bound), and navigation world models. 🙌 📚 dynamic view synthesis 📚 teleop data augmentation (also compare to MimicGen as an upper bound) 📚 navigation world models 📸 fromshow more

The video is proof that Indian drivers have :... ☠️ No Hazard Perception ☠️ No Driver Education ☠️ No Safety Knowledge No amount of training can fix this as they're ready to face ☠️ with a smiling face. 🔥The bus can rear end them. 🔥The container truck can brake suddenly and they can have an underride crash 🔥 They could be sandwiched between the container and oil tanker. N number of possibilities to have a crash. The behaviour on roads is just like a child who has no fear of fire,and learns the lesson the hard way by burning fingers. We don't have an adult to teach them unfortunately. Avoid excessive risk taking on roads, everything can end in a matter of seconds. #HazardPerception #DriverEducation #IndianRoads Road Safety Guy @SafetyOverSpeed Team Road Squad🚦🚴‍♀️ Naresh R-Threeshow more

It will take 137 years to end extreme poverty... for women. The world cannot wait. With the right decisions and investments, we can accelerate change. See how we can shape the future: #CSW69show more

Women around the UK will seek out Refuge’s support... services when they’re ready to leave an abusive partner. We can only support them with your help. Can you donate to Refuge and help us to provide a warm bed, a cup of tea and pyjamas for a woman that has fled to safety?show more

Devin for Terminal is the first CLI agent with... its own dedicated virtual machine. You can start a project locally and hand it off to Devin to test it end-to-end. Devin for Terminal is super performant, and it works with all frontier models. Try it out today.show more

Devin for Terminal is the first CLI agent with... its own dedicated virtual machine. You can start a project locally and hand it off to Devin to test it end-to-end. Devin for Terminal is super performant, and it works with all frontier models. Try it out today.show more

"How can it possibly be that we don't have... a good theoretical understanding of the most consequential technology of our age"--Surya Ganguli. This episode of The Information Bottleneck podcast with Surya Ganguli is a treat (esp. for those 2+ hrs one is stuck in Bangalore traffic!): machine learning, neuroscience, evolutionary biology, math, statistical physics, and more, all seamlessly discussed, with hosts Ravid Shwartz Ziv and Allen Roush. Full Episode here:show more

A Tesla on Full Self-Driving blows through a stop... sign at 35mph and nearly collides with two cars. The kicker is that this was during a livestream debate-demo-drive between FSD fan Ross Gerber and FSD skeptic Dan O'Dowd. This should go without saying, but for an automated system to be at the safety level of a human, this cannot happen. The fact that this occurred yesterday during THIS drive (along with other safety-critical disengagements) should serve as statistical evidence of how frequent this is occurring. IMO, this is the biggest nail in the Tesla FSD coffin.show more

Newton 1.0 is now generally available. 🙌 Take robot... learning to the next level with: 🤖 Stable Articulated & Complex Mechanism Simulation – accurate, reliable machine modeling. 🖐️ High-Fidelity Hydroelastic Contact Modeling – realistic soft contact and touch-based interactions. 🧵 Deformable Body Simulation – simulate cables, cloth, rubber, and other elastic materials with VBD. ⚡ Accelerated Robot Learning at Scale – seamless integration with open simulation and learning frameworks, NVIDIA Isaac Sim and Isaac Lab for scalable workflows. Learn how to integrate this open-source physics engine into your workflow: #NVIDIAGTCshow more

The era of ultra-high-resolution imaging has arrived. Modern image... sensors exceeding 200 MP resolution are common in smartphones, with over 400 MP sensors under development. However, the large number of pixels poses significant challenges for acquisition and processing, especially on edge devices. Which pixels should be acquired, and when, for bandwidth-efficient imaging and perception? We introduce Policy-based Foveated Imaging and Perception, an on-device, real-time, predictive, and task-aware framework that dynamically allocates sensor resolution to prioritize important regions under specific perception objectives. This paper will be presented at #SIGGRAPH2026! [1/6]show more

Trump on Iran: We are taking out all of... their capability for anything having to do with the Hormuz Strait. I think, in the end, we will end up controlling the whole thing.show more

Hey Shri, Congratulations on the end of your short,... meaningless political career. You can now devote some time to learning English. Thanks for the laughs.show more

BREAKING: TRUMP CLAIMS US WILL CONTROL THE STRAIT OF... HORMUZ IN THE END “We are taking out all of their capability for anything having to do with the Hormuz Strait. I think, in the end, we will end up controlling the whole thing.”show more

As a newly appointed 𝗔𝘀𝘀𝗶𝘀𝘁𝗮𝗻𝘁 𝗣𝗿𝗼𝗳𝗲𝘀𝘀𝗼𝗿 at Imperial College... London, I'm thrilled to announce the 𝗦𝗮𝗳𝗲 𝗪𝗵𝗼𝗹𝗲-𝗯𝗼𝗱𝘆 𝗜𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝘁 𝗥𝗼𝗯𝗼𝘁𝗶𝗰𝘀 𝗟𝗮𝗯 (𝗦𝗪𝗜𝗥𝗟) at 𝗜𝗺𝗽𝗲𝗿𝗶𝗮𝗹 𝗖𝗼𝗹𝗹𝗲𝗴𝗲 𝗟𝗼𝗻𝗱𝗼𝗻. 𝗦𝗮𝗳𝗲 𝗪𝗵𝗼𝗹𝗲-𝗯𝗼𝗱𝘆 𝗜𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝘁 𝗥𝗼𝗯𝗼𝘁𝗶𝗰𝘀 𝗟𝗮𝗯 (𝗦𝗪𝗜𝗥𝗟) ( is a new research lab focused on the intersection of safety and intelligence in next-generation robotics. We're hiring exceptional PhD students who are passionate about pushing the boundaries of robot learning. 𝗪𝗵𝗮𝘁 𝗺𝗮𝗸𝗲𝘀 𝗦𝗪𝗜𝗥𝗟 𝘂𝗻𝗶𝗾𝘂𝗲? We operate at the exciting convergence of: • Online & offline reinforcement learning • Imitation learning & human demonstrations • Sample-efficient learning methods • Whole-body and soft robotics systems We're 𝗹𝗼𝗼𝗸𝗶𝗻𝗴 𝗳𝗼𝗿 𝗽𝗿𝗼𝘀𝗽𝗲𝗰𝘁𝗶𝘃𝗲 𝗣𝗵𝗗 𝘀𝘁𝘂𝗱𝗲𝗻𝘁𝘀 interested in: • Developing safe exploration algorithms for robotic systems • Creating sample-efficient learning methods that minimize real-world trials • Building foundation models for robotics with safety guarantees • Advancing soft robotics and compliant human-robot interaction • Bridging theory and practice in embodied AI Why now? As robots become more capable and work closer with humans, we need systems that are both intelligent enough to handle complex tasks 𝗔𝗡𝗗 safe enough for real-world deployment. Traditional approaches treat safety and intelligence as competing priorities, we believe they're synergistic. If you're a motivated researcher who wants to develop the theoretical foundations and practical algorithms for tomorrow's safe, intelligent robots, I'd love to hear from you. Want to join? Apply viashow more

The next time you fly, India's space technology could... help bring your aircraft safely to the ground. For the first time, a commercial jet in India successfully landed using GAGAN, ISRO's indigenous satellite navigation system, marking a major milestone for Indian aviation. Developed by ISRO and the Airports Authority of India, GAGAN guided an IndiGo Airbus A320 during a successful satellite based landing approach at Udaipur Airport. Instead of relying heavily on expensive ground based navigation equipment, GAGAN uses satellite technology to provide highly accurate landing guidance. This could help more airports offer precision landings, improve safety during poor weather, reduce flight diversions, and make India's aviation network more efficient and future ready. ISRO | Airports Authority of India #ISRO #GAGAN #IndianAviation [ISRO, GAGAN, IndiGo, Indian Aviation Technology]show more