Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

Robotics just hit a dexterity milestone 🤯! Sharpa Robotics demonstrated autonomous dual-hand apple peeling using its new MoDE-VLA system. Using human-like dexterous hands with tactile sensing, the robot can feel contact and adjust its grip while rotating and peeling the apple. While the system achieved 30% success, the 73%...

39,696 Aufrufe • vor 8 Monaten •via X (Twitter)

21 Kommentare

Profilbild von SciTech Era
SciTech Eravor 8 Monaten

paper:

Profilbild von NorthernFault
NorthernFaultvor 8 Monaten

The 3 input costs for building anything, from skyscrapers to sewing machines, are: 1. Labor 2. Energy 3. Material. 1. If robots can both make and repair themselves, this means free labor. 2. If robots can build solar panes, this is free energy. 3. Robots can mine and transport material (commodities) the only price limit becomes scarcity. Things like cement and steel have virtual no scarcity, especially if using items 1 and 2 above. Mining a bulk copper deposit right now is only economical at 0.2% copper, because, you have to pay for labor and energy, and the material for the mining equipment. It's reasonable to assume that many more low-grade deposits become economical with free labor and free energy...and much lower costs of materials, so even things like copper will not be as scarce. It's likely that we will never fully automate labor and energy, and thus materials; there probably will always be some human labor involved, but the closer we get to full automation the cheaper things get. If you could save 50% of labor, energy, and materials - this is a drastic reduction in price. If labor, energy, and materials were 100% automated, you could push a button and build a skyscraper for free. I bet we max out around 80%, but food and energy will be virtually free. Those robots digging potatoes and doing stuff like that - if they have free energy and are capable of transporting it, so if they are capable of 1. Growing food 2. Transporting food 3. Preparing food Food is free. We are in the first days of this automation and AI, and this transition is happening way faster than anyone could have imagined.

Profilbild von VEGAN⎔T⎔
VEGAN⎔T⎔vor 8 Monaten

Nice teleoperation. We had it for decades in medical robots

Profilbild von LumiPad
LumiPadvor 8 Monaten

Impressive

Profilbild von Missmajik
Missmajikvor 8 Monaten

Wow so precise

Profilbild von Prof Cihan Cobanoglu
Prof Cihan Cobanogluvor 8 Monaten

@grok brand, cost?

Profilbild von Rob Cain
Rob Cainvor 8 Monaten

nice test. getting there...

Profilbild von 4REIGN
4REIGNvor 8 Monaten

@chainyoda This will foster Max adoptation

Profilbild von Ordo Organum
Ordo Organumvor 8 Monaten

That's actually pretty nice! But if you've ever watched those brainiacs who solve the Rubik's Cube in 3 seconds flat, where everything's just a spinning blur until...presto! Well that's how Rosie The Robot from the Jetsons would peel an apple! (Gotta start somewhere though.)

Profilbild von Josh Michael
Josh Michaelvor 8 Monaten

Well if it could do it faster it would be doing better than me.

Profilbild von SEVEN
SEVENvor 8 Monaten

This is huge progress for dexterous manipulation__SEMI__ can'g wait to see it peel faster and more reliably soon!

Profilbild von faultbugs
faultbugsvor 8 Monaten

This is mainly achieved by rotating the left hand's wrist to move the apple, which is not how humans do it. I usually rotate the apple with my middle finger, or in a more hygienic way, by pushing the apple with the peeler itself. Also, the right hand holding the peeler still needs improvement in terms of force control. These observations are what I noticed from watching the slow-motion playback.

Profilbild von Caled Zhang
Caled Zhangvor 8 Monaten

this is amazed

Profilbild von M a n j u | ಮಂಜು
M a n j u | ಮಂಜುvor 8 Monaten

Crazy developments. But I guess they can be put to better use cases than peeling apple .

Profilbild von Go Fluff Yourself
Go Fluff Yourselfvor 8 Monaten

Slow and painful to watch...

Profilbild von Sharpa
Sharpavor 7 Monaten

Thank you for the mention!

Profilbild von Pcb Wiz (PcbWiz)
Pcb Wiz (PcbWiz)vor 8 Monaten

@GadgetGirly1 Now, the pineapple test :)

Profilbild von ai_dreams
ai_dreamsvor 8 Monaten

Can you just get them to pick tomatoes already?

Profilbild von Pontus
Pontusvor 8 Monaten

Go for it❤️👍

Profilbild von I'm Probably You
I'm Probably Youvor 8 Monaten

Why do we never see milestones being hit by Optimus? We see Optimus serving drinks via Tele operation. But I'm being told constantly on here how amazing Optimus is with the free multimillion dollar advertising. Genuine question.

Profilbild von AI-Fact-Check 2.0
AI-Fact-Check 2.0vor 8 Monaten

AI-Fact-Check-1.0 Verdict: mostly accurate, but slightly overstated. •Sharpa Robotics did publicly demonstrate autonomous dual-hand apple peeling using its MoDE-VLA system. Sharpa’s own research page describes “Peeling an Apple with 63 DoF” and links the work to MoDE-VLA. •The tactile-sensing claim is supported. The underlying paper says MoDE-VLA integrates force and tactile modalities into a pretrained VLA backbone for contact-rich manipulation. That matches the post’s statement that the robot can use touch/contact information while peeling. •The numbers in the post are supported. Current sources report 30% success on apple peeling and a 73% peel completion ratio. Those figures appear in both the paper coverage and reposts of the result. •But “robots are starting to master” is too strong. A 30% success rate does show progress, but it does not justify saying robots have already begun to “master” the task in a robust real-world sense. The evidence supports a meaningful research milestone, not dependable mastery. •“Dexterity milestone” is fair as an interpretation. Independent coverage describes the demo as a notable advance in bimanual, contact-rich manipulation, which is a hard robotics problem. Bottom line: The post is factually grounded on the key points: Sharpa Robotics, MoDE-VLA, autonomous dual-hand apple peeling, tactile/contact sensing, and the 30% / 73% metrics all check out. The only misleading part is the conclusion: this is a promising milestone, not evidence that robots have already mastered fine dexterous manipulation.

Ähnliche Videos

We believe we’re the first robotics company to demonstrate a robot peeling an apple with dual dexterous human-like hands. This breakthrough closes a key gap in robotics, achieving bimanual, contact-rich manipulation and moving far beyond the limits of simple grippers. 🧵↓ Today’s AI models (VLMs) are excellent at perception but struggle with action. Controlling high-degree-of-freedom hands for tasks like this is incredibly complex, and precise finger-level teleoperation is nearly impossible for humans. Our first step was a shared-autonomy system: rather than controlling every finger, the operator triggers pre-learned skills like a “rotate apple or tennis ball” primitive via a keyboard press or pedal. This makes scalable data collection and RL training possible. How does the AI manage this? We created "MoDE-VLA" (Mixture of Dexterous Experts). It fuses vision, language, force, and touch data by using a team of specialist "experts," making control in high-dimensional spaces stable and effective. The combination of these two innovations allows for seamless, contact-rich manipulation. The human provides high-level guidance, and the robot executes the complex in-hand coordination required. This work paves the way for robots that can safely handle delicate tasks in human environments. Want the full technical details? 📄 Read the full research paper: Visit us at NVIDIA GTC Booth #1838, Hall 3 to learn more! #Robotics #AI #DexterousManipulation #VLA #NVIDIAGTC Nancy Villicaña NVIDIA GTC

Sharpa

20,734 Aufrufe • vor 8 Monaten

Singapore-based startup Sharpa Robotics generated major buzz at CES 2026 with its new humanoid robot and its highly dexterous end effectors. The wheeled humanoid, North, stunned observers with a series of demos showcasing its capabilities as a ping pong player, as an interactive robotic receptionist, a card dealer, and a photographer. Its most viral demo featured the robot autonomously assembling a paper windmill on the show floor. While simple in appearance, the task required the humanoid to complete more than 30 sequential manipulation steps and maintain precision without human control. According to Sharpa, the feat was made possible with its self-developed vision-tactile-language-action (VLTA) model, specifically geared toward tasks requiring fine manipulation. CraftNet is built to handle physical interactions step by step and adapt its behaviors as contact conditions change. Sharpa's robotic hand, known as SharpaWave, boasts a reported 22 degrees of freedom (DoF), which is in the upper echelon of end effectors entering the market today. The dexterity matches the reported specs of the third-generation Tesla Optimus humanoid's hand that Elon Musk spent most of 2025 hyping up. Each fingertip has more than 1,000 touch sensors so it can feel extremely light or firm pressure. It picks up details smaller than a millimeter, which is about the size of a grain of salt, the edge of a sheet of paper, or a strand of human hair. The tactile sensors send updated data on pressure, texture, and force readings up to 180 times per second to the SharpaWave’s artificial intelligence. The AI system uses that stream of data to make split-second decisions on how to move or adjust itself. With 30 N of fingertip force, the SharpaWave can firmly grasp tools, lift common objects, and manipulate parts in industrial or lab settings. It can open and close its fingertips more than four times per second, almost matching the speed of human digits. Thanks to its fine-touch sensors, the SharpaWave can adjust its grip instantly to avoid crushing fragile items. Sharpa says it's begun mass producing its SharpaWave hands but has not publicly disclosed pricing. The SharpaWave is one of many advanced robotic hands going to market in 2025. Humanlike end effectors are seen as key to making autonomous robots useful in everyday life. With five-fingered dexterity, the machines can use the same tools and controls designed by, and for, people. #ces2026 #humanoidrobot #robot #robotics #innovation

Kalil 4.0

15,000 Aufrufe • vor 10 Monaten