Affgan @treck658

震撼一幕！索尼 AI 乒乓球机器人 Ace，在正式乒联规则下，击败了世界排名前 26 的日本名将木原美悠。

@xaydenroy Very nice 👍

Trying out the robot arm from pulsar at ICRA 2026. Felt very smooth and cool to try backdriving it at different velocities to feel the damping effect. It also has some programmed stops near pinch points. #icra2026

Wow. This is crazy. A developer trained an AI agent in simulation and deployed it onto a real robotic air hockey table using reinforcement learning. This robot can track the puck with millimeter-level accuracy and react in roughly 20 milliseconds, fast enough to challenge even skilled human players. We’re moving from robots that follow programmed rules to machines that learn strategies in simulation and execute them in the physical world.

@bayraktar_1love The road looks in better condition than any in England.

@100kDiary $CNDX All the way

@Formula1Wimbo If you win the Nürburgring 24, you’re driver of the day in my book. 😉

@Formula1Wimbo Yes, they have radio, but endurance isn’t like F1 with live corner-by-corner coaching. Telemetry/timing goes to the box, engineers call strategy, gaps, car issues, weather, Code 60s etc. But traffic management is mostly on the driver, especially at the Nordschleife.

@SalStax Tell me you don’t understand the Nürburgring 24 without telling me. That grid is not “a bunch of amateurs”. It’s packed with factory GT drivers, endurance specialists, ex-F1 drivers and Nordschleife experts. Max being quick there is still seriously impressive.

AI talking about the future of AI…

The Wuji Hands in action.

Are you kidding me??? It grasps multiple objects with different ways, all at once with… a single hand??? No pauses. 1x speed. GENE-26.5 is @gs_ai_’s robotics-native multimodal foundation model. It’s trained on 200,000+ hours of real human hand data (motion, force, touch) and runs on a 54-DoF bimanual system: Scaling that human data 4x lifted real-robot success rates from 16.6% to 65.6% on long-horizon dexterous tasks!! Same model weights, zero fine-tuning for this exact sequence. You know those tiny coordinated movements you do without thinking…? Robots couldn’t reliably do that before. Now they can. Today. This is the video you’ll send to my friends, outside of our bubble, when they say “robots are still just demos.” Congrats to the entire team around @zhou_xian_! Credit: Seen at Zu Wang (@zuwang95) Genesis official announcement for the full story + longer demo: genesis.ai/blog/gene-26-5…) ——— Weekly robotics and AI insights. Subscribe free: 22astronauts.com

High accuracy with limited claw end effectors.

Tesla is building a completely new production line for a product the world has never seen Optimus Gen 3 is going to be absolute insanity I don't think people realize how crazy this transition is going to be

Automating recycling with AI-driven sorting! ♻️ AMP Robotics developed the AMP Cortex system, an AI-driven garbage-sorting robot that processes images of recyclable materials, classifying them into hundreds of categories by material type, color, form, and other characteristics. The recycling industry faces severe challenges. Global recycling contamination rates run 20-25%, meaning a quarter of collected recyclables end up in landfills due to sorting errors. Manual sorting is labor-intensive, dangerous (broken glass, sharp metals, biohazards), and struggles with throughput demands. AMP Cortex uses computer vision and robotic arms to identify and pick recyclables at high speed. The system learns to recognize materials across varying lighting, dirt, damage, and mixed streams.conditions where traditional optical sorters fail. Facilities process hundreds of tons daily. Every percentage point improvement in recovery rate translates to significant revenue and environmental impact. The economics are compelling. Labor costs for manual sorters are high, turnover is severe, and sorting speed is limited. ~~ ♻️ Join the weekly robotics newsletter, and never miss any news → ziegler.substack.com

Even when caught off guard, a bipedal humanoid #Robot that handles tasks while adapting via @ZappyZappy7 #AI #Robotics #MachineLearning #ArtificialIntelligence #ML #Innovation

It’s a 3D printer and automated assembly station in one. Developed at Meiji University, the Functgraph begins as a standard 3D printer, then transforms into a compact manufacturing system. It can print parts for its own tools, collect them, clean them, and assemble them without human input. Rather than only producing objects, Functgraph can carry out real physical tasks such as folding laundry or slicing vegetables by creating and assembling task-specific tools. Video Credit: YouTube/@YK-dc5tk #3dprinting #engineering #technology

Bio-Inspired #Robot Snails Use Suction to Form Structures and Navigate Complex Terrain by @lukas_m_ziegler #EmergingTech #Technology #Innovation #Tech

Robots that actually WORK in real factories. Without endless retraining. > Just 20 minutes of data. > 99.4% success rate. > 108 motors soldered in 5+ hours straight. Sub-0.6 mm precision on messy, deformable (!) cables. This hybrid “learning-augmented” system adds neural brains + 3D safety monitoring to ordinary cobots… and suddenly complex factory tasks become reliable, human-safe, and stupidly fast. This is for all of you who are into manufacturing. Huge credit to robotics engineer Yunho Kim @awesomericky99 📍Paper: arxiv.org/abs/2604.22235

Gen-1 ties zipties Read more about Gen-1 in our blog posts in the comments below ↓