By
–
All predictions are now live on Flood Hub for free: sites.research.google/floods… I break down stories like this every day in my free newsletter. Keep up with the latest in AI/Robotics in 5 min a day: therundown.ai/subscribe
→ View original post on X — @rowancheung, 2026-04-06 15:13 UTC
By
–
Google's new AI can predict flash floods 24 hours before they strike.
— Rowan Cheung (@rowancheung) 6 avril 2026
How it works:
> Uses Gemini to extract confirmed flood locations and times from global news
> Builds a dataset of past events that never formally existed.
> That dataset feeds a neural network
> The neural… pic.twitter.com/wR0d43Qpsc
Google's new AI can predict flash floods 24 hours before they strike. How it works: > Uses Gemini to extract confirmed flood locations and times from global news > Builds a dataset of past events that never formally existed. > That dataset feeds a neural network > The neural network combines real-time weather forecasts with local terrain, soil absorption rates, and urban density > Can flag at-risk areas with results that match the accuracy of America's National Weather Service > Can do this in countries that have almost no flood monitoring infrastructure at all Flash floods kill more than 5,000 people every year, but most victims never see them coming. A 12-hour warning alone can reduce flash flood damage by 60%. …For billions of people without any early warning system, this could be the difference between life and death. [Translated from EN to English]
→ View original post on X — @rowancheung, 2026-04-06 15:13 UTC
By
–
Source: zzk273.github.io/LATENT/ I break down stories like this every day in my free newsletter. Keep up with the latest in AI/Robotics in 5 min a day: therundown.ai/subscribe
→ View original post on X — @rowancheung, 2026-04-03 15:15 UTC
By
–
Researchers just taught a robot to play tennis.
— Rowan Cheung (@rowancheung) 3 avril 2026
From just clips of a few amateur players performing basic forehands, backhands, and shuffles…
…a robot learned one of the fastest, most coordinated physical skills there is.
Insane! pic.twitter.com/dDrEhqL4RA
Researchers just taught a robot to play tennis. From just clips of a few amateur players performing basic forehands, backhands, and shuffles… …a robot learned one of the fastest, most coordinated physical skills there is. Insane!
→ View original post on X — @rowancheung, 2026-04-03 15:15 UTC
By
–
Source: technologyreview.com/2026/03/10/1134099/how-pokemon-go-is-helping-robots-deliver-pizza-on-time I break down stories like this every day in my free newsletter. Keep up with the latest in AI/Robotics in 5 min a day: therundown.ai/subscribe [Translated from EN to English]
→ View original post on X — @rowancheung, 2026-03-20 16:15 UTC
By
–
Every Pokémon Go player unknowingly trained future robots.
— Rowan Cheung (@rowancheung) 20 mars 2026
Over the years, players photographed 30 billion images from cities around the world.
And now that data is teaching delivery robots to navigate streets.
Niantic (company behind Pokémon Go), spun out an AI company called… pic.twitter.com/q8WGpuU0vT
Every Pokémon Go player unknowingly trained future robots. Over the years, players photographed 30 billion images from cities around the world. And now that data is teaching delivery robots to navigate streets. Niantic (company behind Pokémon Go), spun out an AI company called Niantic Spatial that's turning years of player data into a visual positioning system for machines. When 500 million people installed the game and pointed their phones at buildings, landmarks, and street corners, they were also capturing detailed location data that normal GPS can't match: > GPS signals bounce off skyscrapers and lose accuracy by up to 50 meters in dense cities. > That's a dealbreaker for sidewalk delivery robots that need to stop at exactly the right point. > Niantic Spatial trained a model on those 30 billion images to pinpoint a robot's exact location based on what its cameras see, not satellite signals. > Delivery startup Coco Robotics is the first partner, using the system to guide thousands of robots across LA, Chicago, and Miami. Don't think anyone thought they were training robots when they were chasing Pikachu during COVID.
→ View original post on X — @rowancheung, 2026-03-20 16:15 UTC
By
–
More info @orcahand if you want to learn more, or even get one for yourself 🙂 orcahand.com/hardware P.S. I break down stories like this every day in my free newsletter. Keep up with the latest in AI/Robotics in 5 min a day: therundown.ai/subscribe
→ View original post on X — @rowancheung, 2026-03-19 16:12 UTC
By
–
This robotic hand can be 3D printed by anyone and assembled in under 8 hours.
— Rowan Cheung (@rowancheung) 19 mars 2026
Researchers at ETH Zurich created the Orca hand, fully open-sourced with artificial bones and tendons.
For context, advanced robotic hands cost over $100,000 and require constant maintenance…
Orca… pic.twitter.com/LEUkM8hVDy
This robotic hand can be 3D printed by anyone and assembled in under 8 hours. Researchers at ETH Zurich created the Orca hand, fully open-sourced with artificial bones and tendons. For context, advanced robotic hands cost over $100,000 and require constant maintenance… Orca costs under $2,000. 50x less (!) A self-calibration system maps every motor to every joint, eliminating the manual tuning that tendon-driven hands usually need. Each fingertip has built-in tactile sensors covered by silicone skin. The hand can actually feel when it touches something, giving it feedback to grip objects without crushing them or letting them slip. It can hold over 20 lbs, learn tasks by watching human demonstrations, and transfer skills trained in simulation directly to the real world. The team proved its durability by having it pick up and place a cube over 2,000 times across 7 hours with no human intervention. The full design files and source code are open source, so any robotics lab in the world can start building one today.
→ View original post on X — @rowancheung, 2026-03-19 16:12 UTC
By
–
I break down stories like every day this in my free newsletter. Keep up with the latest in AI/Robotics in 5 min a day: therundown.ai/subscribe Source: eu-startups.com/2026/02/with…
→ View original post on X — @rowancheung, 2026-03-17 16:13 UTC
By
–
This robotic hand can weave itself together in minutes.
— Rowan Cheung (@rowancheung) 17 mars 2026
Allonic developed a process that "braids" robot bodies around a 3D-printed skeleton in a single automated step.
The tech draws from the textile industry, using braided fibers instead of traditional mechanical joints and… pic.twitter.com/VRh0AUQ2YE
This robotic hand can weave itself together in minutes. Allonic developed a process that "braids" robot bodies around a 3D-printed skeleton in a single automated step. The tech draws from the textile industry, using braided fibers instead of traditional mechanical joints and bearings. Their braiding system grows tendon-driven structures directly onto skeletal cores, creating flexible yet durable parts without screws or assembly. One finger only needs 4 skeletal elements. Then thousands of braided fibers wrap around them to handle force and articulation. The software converts designs directly into machine code that runs the braiding process. But what matters is that robotic hands like this can cost up to $30,000, ~20-30% of a humanoid robot's total price. Allonic's braiding process could crush those costs while enabling custom designs at scale.
→ View original post on X — @rowancheung, 2026-03-17 16:13 UTC