Robots that learn through touch. CAP unlocks scalable, contact-aware manipulation with less data and strong zero-shot results. 🤖 learn more: arxiv.org/pdf/2602.09017 and cap-policy.github.io work from: NYU, UC Berkeley, UCLA, Hello Robot Inc., Ai2, University of Waterloo

Great conversation on the ATDev Podcast with our own V Nguyễn on human-centered design, “solutioneering,” and how clinicians help turn robotics research into real-world impact. A must-listen for anyone interested in assistive tech and social good. 🤖 Listen Here: www.youtube.com/watch?v=YKUT...

New research from Peking University introduces LagMemo, a language-grounded 3D memory that helps robots remember and navigate complex spaces while handling multi-goal, open-vocabulary tasks. Big step toward more capable embodied AI in the real world. Learn more: weekgoodday.github.io/lagmemo/

New research: Fast-SmartWay enables robots to follow language instructions and navigate new environments without panoramas or multi-stage pipelines. Great step toward faster, more practical vision-language navigation in embodied AI. 🤖Read More: arxiv.org/pdf/2511.00933

Great new work from Khalifa University on closed-loop plant disease detection and treatment using vision-language perception and AI planning. Excited to see Stretch supporting precision agriculture research! 🌱🤖 Learn More: www.scitepress.org/Papers/2025/...

Our community newsletter features some exciting work! 🥂 Testing socially assistive robots ✨ LagMemo visual navigation system 🥂 Event-Grounding Graph ✨ Fast-SmartWay: zero-shot visual-language navigation 🥂 Detecting and treating crop diseases ...and so much more! 🤩 hello-robot.com/community-up...

So happy to present at the Mobility For Everyone (MOVE) Center at Kennesaw State University Research this week. Huge thanks to Dr. Ayse Tekes for the invitation and for leading impactful work in assistive and inclusive technology. We love showing how Stretch 3 makes a difference in real life!

Using Stretch 3, Reality Capture is an automated indoor scanning system that collects images on its own, even in cluttered spaces, achieving 92% alignment in cluttered areas and 98% in open rooms. High-quality 3D models, less manual work, more possibility. www.youtube.com/watch?v=FXQS...

Clear communication is key for safe, comfortable human-robot interaction. This new work introduces CoRI, a pipeline that turns a robot’s motion plan and visual perception into simple, natural-language explanations of what it’s about to do. Learn more: cori-phri.github.io

Love seeing Stretch in action! Our friends from CNBEST in Beijing took Stretch to the 2025 International Conference on Intelligent Robots and Systems last month to demo it's capabilities. It featured impactful keynote speeches, cutting-edge technical sessions, and dynamic networking opportunities.

While finding missing or misplaced items may be second nature to humans, it proves to be a major challenge for robots. This study tackles that issue with a new open-vocabulary search and retrieval framework. Learn more: autonomousrobots.nl/assets/image...

Happy Holidays from Hello Robot! 🦃 DynaMem - dynamic spatio-semantic memory 🌲 Automated 3D reconstruction system 🦃 Robotics in everyday caregiving 🌲 Remote trick or treating with seniors 🦃 Event-Grounding Graph framework Learn more: hello-robot.com/community-up...

Last year, Northeastern published a fascinating piece on their “robot training lab," where their team deployed our Stretch 3 robot (nicknamed “Marlo”) into a real-world campus environment to learn navigation, human interaction, and task assistance. 🤖 news.northeastern.edu/2024/12/10/r...

What an amazing experience for Hello Robot to be invited to demo Stretch 3 at the TechWomen program orientation at UC Berkeley, that connects emerging women leaders in STEM from Africa, Central & South Asia, Europe, and the Middle East with U.S. mentors and companies.🤖

Researchers at Princeton are pairing VR with robotics to make it happen, blending digital interactions with physical delivery. Imagine selecting a soda with a gesture and watch it appear on your desk, all powered by a robot hidden from view. Learn more: www.wevolver.com/article/eras...

Robot-led rehab is showing promise for people with Parkinson’s. 🤖🧑‍⚕️ A pilot study with exercise specialists found it can boost consistency & engagement, while feedback highlights the need for more natural responses & easier use. #Robotics #Healthcare www.mdpi.com/2227-9032/13...

Back to School and back with new work from the Hello Robot Community! 👩‍🏫 Teaching a robots 📝 Assisting in senior living 👨‍🏫 Invisible robots deliver 📝 Learning from failures 👩‍🏫 Evaluating memory 📝 Magnetic touch 👨‍🏫 Parkinson's rehab 💻 To learn more, check out: hello-robot.com/community-up...

We were excited to welcome the auli.tech team along with Henry and Jane Evans, to Hello Robot HQ! We explored using Cato to operate the Stretch robot as part of Hello Robot’s effort towards greater accessibility with assistive input devices. We are grateful for their invaluable insights!

Honored to be featured with our friend Henry Evans. After a stroke 22 yrs ago, Henry became a powerful advocate for adaptive robotics. Watch how Stretch 3 helps him eat, play, & connect with family, showing the power of human-centered tech to restore independence. www.cbsnews.com/video/a-robo...

How can robots better communicate what they “see” and “do”? This study, introduces RONAR, an LLM-powered system that translates real-world robot experiences into natural, human-readable narratives, boosting transparency, and improving collaboration with humans. 🤖 sites.google.com/view/real-wo...

How do you build a robot that can open unfamiliar objects in new places? This study put mobile manipulation systems through 100+ real-world tests and found that perception, not precision, is the real challenge.🤖 Learn more: arjung128.github.io/opening-arti... Watch full video: youtu.be/QcbMnEXd70Q

Can swearing make robots more relatable? Turns out yes! A new study from Oregon State found that when robots use light profanity after messing up, people view them as more human and even funny, without losing trust or credibility. 🤖 spectrum.ieee.org/cursing-soci... #Robotics #SocialRobots

Could robots learn just by watching videos? Insert VidBot: a new approach that teaches robots how to perform real-world tasks by learning from everyday human videos. No manual training, no specific demos, just smart, scalable learning from the internet. 🤖https://hanzhic.github.io/vidbot-project/

We had an incredible time showcasing everything Stretch 3 is capable of at the AI for Good Summit! It was a pleasure to be joined by the talented team from the NYU GRAIL lab, who demonstrated their cutting-edge work on Robot Utility Models. learn more at: robotutilitymodels.com #AiforGood #robots

Stretch 3 went on an adventure to the Portland Art Museum! Thanks to Stretch’s moving tablet and mobility, our friend Henry was able to explore the the museum just like he was there in person. Big thanks to the museum’s accessibility team for making it such a welcoming and meaningful day!

The Hello Robot team is ready to go at the AI for Good Global Summit! Excited to connect with innovators from around the world and share how Hello Robot is building useful, inclusive robots that make a real difference. Can’t wait to get underway! #AIforGood #HelloRobot

Kicking off Summer with new work in the Hello Robot Community! 🎆 Embodied question answering 🎆 Robot profanity?! 🎆 End-to-end drawer and cabinet opening 🎆 Estimating fruit ripeness 🎆 Robotic tutor to help train nurses 🎆 New Stretch MuJoCo release! Learn more! hello-robot.com/community-up...

Hello Robot is heading to Geneva for the AI for Good Global Summit! As the leading UN platform for AI, this summit brings together innovators, researchers, and changemakers to explore how AI can tackle some of the world’s biggest challenges. Come visit us or watch online! aiforgood.itu.int/summit25

What will it take to put robots in homes? Hello Robot cofounder and CTO Dr. Charlie Kemp shared some insights with students at Carnegie Mellon, exploring the technical and human-centered challenges of bringing robots into personal spaces. ▶️Watch the full talk: www.youtube.com/watch?v=fQT7...

Soaking up the sun at the Robotics: Science and Systems conference in Los Angeles this weekend? Stop by the Hello Robot booth to say hi and get a hands on look at Stretch! Hope to see you there 😎 roboticsconference.org

What an incredible moment for Stretch! Our Stretch 3 robot joined Mark Russinovich and Scott Hanselman onstage at Microsoft Build 2025 for a live demo, autonomously fetching a drink to highlight the exciting potential of cutting-edge AI models and robotics. www.youtube.com/watch?v=KIFD...

🚀Stretch MuJoCo v0.5 is released! It's a high-fidelity simulation of Stretch 3. Here’s what's new: • ROS2 and Python libraries • RGB-D and Lidar sensors • 100s of kitchen-style environments • Runs on Ubuntu, MacOS, Windows, or online on Google Colab

For the finale, @markrussinovich.bsky.social and I solved a real-world problem by gaffer taping a Surface Laptop to a @hellorobot.bsky.social (thanks for the loner!) and taught it to get me a Diet Coke live on stage youtu.be/KIFDVOXMNDc?...

New work from the Hello Robot community this spring: 🌷 Open-source dynamic semantic mapping 🌷 Servoing with Vision Models 🌷 Zero-shot manipulation from RGB human videos 🌷 Tons of HRI research 🌷 New Software Drop with Firebase networking And more! Check it out: hello-robot.com/community-up...

We’re deeply honored to be recognized with the inaugural Robots for Good award at the 2025 Robotics Summit. Our sincere thanks to the Hello Robot team and to our worldwide user community for helping make this possible!

Thank you to everyone who came out to meet with us at the Human-Robot Interaction Conference last week!! See you next year in Edinburgh!

Software drop for Stretch! - Tutorial for React/ROS2 in the web interface - Smoother control in dex teleop - Easier Stretch AI installation - Better Nav2 mapping

What's new in the Stretch community this month? ❄️ Gazebo Harmonic ❄️ Dynamic semantic maps for open-vocabulary tasks ❄️ Natural-language narration of robot experiences ❄️ Implicit human-robot communication And more! Follow the link below for more details: hello-robot.com/community-up...

Way back in 2023, before multimodal foundation models were a thing, we wanted to apply language agents to visual domains. One idea was to use vision models to extract perceptual features and put them into text templates. But “a picture is worth 1000 words” — a big context! Can be slow, distracting.

We are thrilled to announce a partnership between Hello Robot, the ALS Association (@alsassociation.bsky.social), and University of Pittsburgh researchers to advance the development of Stretch to meet the evolving needs of people living with ALS. Learn more: hello-robot.com/press-release-als-2024

Introducing Stretch AI - a new open-source suite of tools, tutorials, and reference code to explore and build AI-enabled home robot applications with Stretch 3! Check it out at github.com/hello-robot/stretch_ai and learn more from @cpaxton.bsky.social below!