Video yükleniyor...

Video Yüklenemedi

Ana Sayfaya Dön

1k placements a day—long-horizon tasks—using vision-only—learned behaviors for very precise <1cm sheet metal insertion. Requires intense engineering across multiple disciplines.

TheHumanoidHub's profile picture

The Humanoid Hub

109,936 subscribers

136,670 görüntüleme • 1 yıl önce •via X (Twitter)

Sağlık & İyilikEğitimBilim & Teknoloji
LIVE

Anya Rossi• Live Now

Private livecam show

7 Yorum

Brett Adcock profil fotoğrafı
Brett Adcock1 yıl önce

This use case is very challenging!

Lieutenant Generalist🇺🇸 profil fotoğrafı
Lieutenant Generalist🇺🇸1 yıl önce

Crazy how fast they're advancing!

vikramvi profil fotoğrafı
vikramvi1 yıl önce

would love to see @Tesla_Optimus tasks being done in Gigafactories $TSLA

🌩️ Sky Nuts and ⚡ Bolts 🦄 profil fotoğrafı
🌩️ Sky Nuts and ⚡ Bolts 🦄1 yıl önce

'trained programming' better terminology than "learned behavior"

Mr Yuk profil fotoğrafı
Mr Yuk1 yıl önce

So you replaced a low skilled worker at half the speed. to be replaced with a highly skilled robot fixer.

CV profil fotoğrafı
CV1 yıl önce

Sieht nicht gut aus für Sie, lieber @chorherr! 😜

Evgeny Muryshkin profil fotoğrafı
Evgeny Muryshkin1 yıl önce

Having a dedicated pick and place machine will increase speed 1000x fold

Benzer Videolar

ViTacFormer is a unified visuo-tactile framework for dexterous robot manipulation. It fuses high-res visual+tactile data using cross-attention and predicts future tactile signals via an autoregressive head, enabling multi-fingered hands to perform precise, long-horizon tasks.
2:18

ViTacFormer is a unified visuo-tactile framework for dexterous robot manipulation. It fuses high-res visual+tactile data using cross-attention and predicts future tactile signals via an autoregressive head, enabling multi-fingered hands to perform precise, long-horizon tasks.

The Humanoid Hub

10,816 görüntüleme • 1 yıl önce

🚀🚀🚀 Ever wondered what it takes for robots to handle real-world household tasks? long-horizon execution, deformable object dexterity, and unseen object generalization — meet GR-3, ByteDance Seed’s new Vision-Language-Action (VLA) model! GR-3 is a generalizable Vision-Language-Action (VLA) model with strong capabilities in complex long-horizon tasks. It understands unseen abstract concepts, manipulates deformable objects robustly, and adapts to novel settings with minimal human data. ✨ Generalization: Generalizes well to unseen objects, environments, and even instructions with abstract concepts. ✨ Long-Horizon Manipulation: Completes long-horizon tasks with strong instruction-following capabilities. ✨ Deformable Object Manipulation: Manipulate deformable objects robustly. Project Page: Arxiv: #ByteDance #ByteDanceSeed #GR3 #VLA #Robotics #FoundationModels
4:10

🚀🚀🚀 Ever wondered what it takes for robots to handle real-world household tasks? long-horizon execution, deformable object dexterity, and unseen object generalization — meet GR-3, ByteDance Seed’s new Vision-Language-Action (VLA) model! GR-3 is a generalizable Vision-Language-Action (VLA) model with strong capabilities in complex long-horizon tasks. It understands unseen abstract concepts, manipulates deformable objects robustly, and adapts to novel settings with minimal human data. ✨ Generalization: Generalizes well to unseen objects, environments, and even instructions with abstract concepts. ✨ Long-Horizon Manipulation: Completes long-horizon tasks with strong instruction-following capabilities. ✨ Deformable Object Manipulation: Manipulate deformable objects robustly. Project Page: Arxiv: #ByteDance #ByteDanceSeed #GR3 #VLA #Robotics #FoundationModels

Xiao Ma

46,323 görüntüleme • 11 ay önce

What are the largest software engineering tasks AI can perform? To answer this, we built MirrorCode, our long-horizon SWE benchmark that lets AI code autonomously for days at a time. The best models complete some tasks we estimate would take human engineers several weeks.
1:11

What are the largest software engineering tasks AI can perform? To answer this, we built MirrorCode, our long-horizon SWE benchmark that lets AI code autonomously for days at a time. The best models complete some tasks we estimate would take human engineers several weeks.

Epoch AI

64,877 görüntüleme • 11 gün önce

This trick normally requires a very precise setup but you can just Be Good
0:24

This trick normally requires a very precise setup but you can just Be Good

ZFG

77,534 görüntüleme • 7 gün önce

🎉 Day-0 support for in vLLM, available today in v0.23.0! Congrats to Z.ai on GLM-5.2, a flagship model built for long-horizon coding agents. ✨ 1M-token context, built to hold project-scale engineering work in a single run ✨ Tuned for long-horizon coding: large-scale implementation, automated research, and performance optimization ✨ One task can carry a full dev workflow, from requirements to a deployable product across platforms ✨ Client-side and mobile engineering, including an on-device debugging loop Try it out running it on vLLM today: 🔗
0:30

🎉 Day-0 support for in vLLM, available today in v0.23.0! Congrats to Z.ai on GLM-5.2, a flagship model built for long-horizon coding agents. ✨ 1M-token context, built to hold project-scale engineering work in a single run ✨ Tuned for long-horizon coding: large-scale implementation, automated research, and performance optimization ✨ One task can carry a full dev workflow, from requirements to a deployable product across platforms ✨ Client-side and mobile engineering, including an on-device debugging loop Try it out running it on vLLM today: 🔗

vLLM

35,720 görüntüleme • 21 gün önce

OSWorld2.0 Benchmarking Computer Use Agents on Long-Horizon Real-World Tasks
0:20

OSWorld2.0 Benchmarking Computer Use Agents on Long-Horizon Real-World Tasks

AK

28,625 görüntüleme • 7 gün önce

Our humanoid now learns loco-manip skills that generalize across space (Stanford) & time (day & night), using egocentric vision, trained only in simulation
1:03

Our humanoid now learns loco-manip skills that generalize across space (Stanford) & time (day & night), using egocentric vision, trained only in simulation

Yanjie Ze

93,352 görüntüleme • 9 ay önce

Another day, another humanoid robot from china AGIBOT introduces GO-1, a generalist foundation model that integrates a vision-language model with a latent planner for enhanced long-horizon and dexterous manipulation.
3:46

Another day, another humanoid robot from china AGIBOT introduces GO-1, a generalist foundation model that integrates a vision-language model with a latent planner for enhanced long-horizon and dexterous manipulation.

Chubby♨️

23,513 görüntüleme • 1 yıl önce

Need help in the real world? RoboVQA can guide robots and humans through long-horizon tasks on a phone via Google Meet. We release a dataset of 800k (video, question/answer) with robots & humans doing various long-horizon tasks. Data: Google DeepMind
0:49

Need help in the real world? RoboVQA can guide robots and humans through long-horizon tasks on a phone via Google Meet. We release a dataset of 800k (video, question/answer) with robots & humans doing various long-horizon tasks. Data: Google DeepMind

Pierre Sermanet

64,369 görüntüleme • 2 yıl önce

A few weeks ago, we shared our progress on articulated objects and long-horizon tasks. Here are two representative examples: - We've been steadily expanding our asset library to cover more articulated objects. Articulated objects have always been a challenging asset class to handle in simulation. Interacting with them requires robots to master atomic skills such as pushing, pulling, opening, and closing, and to understand part structure, interaction constraints, and how the object moves. - Long-horizon tasks can now be generated at scale. Long-horizon tasks are the other hard category: they require chaining multiple sub-goals in sequence. A failure early in the task can cascade and make the rest unrecoverable. Axis is scaling along three dimensions at once: data volume, data quality, and task difficulty.
1:07

A few weeks ago, we shared our progress on articulated objects and long-horizon tasks. Here are two representative examples: - We've been steadily expanding our asset library to cover more articulated objects. Articulated objects have always been a challenging asset class to handle in simulation. Interacting with them requires robots to master atomic skills such as pushing, pulling, opening, and closing, and to understand part structure, interaction constraints, and how the object moves. - Long-horizon tasks can now be generated at scale. Long-horizon tasks are the other hard category: they require chaining multiple sub-goals in sequence. A failure early in the task can cascade and make the rest unrecoverable. Axis is scaling along three dimensions at once: data volume, data quality, and task difficulty.

Axis Robotics

12,654 görüntüleme • 10 gün önce

“Scotland is best served when we collaborate, when we build consensus and work together across sectors, across disciplines and across cultures.” Today, First Minister John Swinney set out his vision for Scotland. Watch now.
1:48

“Scotland is best served when we collaborate, when we build consensus and work together across sectors, across disciplines and across cultures.” Today, First Minister John Swinney set out his vision for Scotland. Watch now.

Scottish Government

37,579 görüntüleme • 1 yıl önce

Introducing DINOv3: a state-of-the-art computer vision model trained with self-supervised learning (SSL) that produces powerful, high-resolution image features. For the first time, a single frozen vision backbone outperforms specialized solutions on multiple long-standing dense prediction tasks. Learn more about DINOv3 here:
1:25

Introducing DINOv3: a state-of-the-art computer vision model trained with self-supervised learning (SSL) that produces powerful, high-resolution image features. For the first time, a single frozen vision backbone outperforms specialized solutions on multiple long-standing dense prediction tasks. Learn more about DINOv3 here:

AI at Meta

899,855 görüntüleme • 10 ay önce

How well do today’s frontier models handle long-horizon, multi-step web agent tasks, such as identifying the top 25 U.S. CS PhD programs with ML/AI faculty likely accepting students and compiling the results into a structured sheet? Check out our new work on Odysseys: Benchmarking Web Agents on Realistic Long Horizon Tasks Paper: Leaderboard: We introduce Odysseys, a benchmark of 200 long-horizon tasks derived from real browsing sessions and evaluated on the live Internet. We show that binary pass/fail is inadequate in this setting and propose rubric-based evaluation, which better aligns with human judgment and provides more informative signals. Across leading models, the best achieves only 44.5% success, leaving substantial headroom. We further introduce a Trajectory Efficiency metric (rubric score per step) and find efficiency remains extremely low (1.15%), highlighting a key bottleneck. Odysseys provides a realistic benchmark for measuring progress toward web agents capable of sustained, efficient, real-world operation. See a more detailed thread by Jing Yu Koh.
1:12

How well do today’s frontier models handle long-horizon, multi-step web agent tasks, such as identifying the top 25 U.S. CS PhD programs with ML/AI faculty likely accepting students and compiling the results into a structured sheet? Check out our new work on Odysseys: Benchmarking Web Agents on Realistic Long Horizon Tasks Paper: Leaderboard: We introduce Odysseys, a benchmark of 200 long-horizon tasks derived from real browsing sessions and evaluated on the live Internet. We show that binary pass/fail is inadequate in this setting and propose rubric-based evaluation, which better aligns with human judgment and provides more informative signals. Across leading models, the best achieves only 44.5% success, leaving substantial headroom. We further introduce a Trajectory Efficiency metric (rubric score per step) and find efficiency remains extremely low (1.15%), highlighting a key bottleneck. Odysseys provides a realistic benchmark for measuring progress toward web agents capable of sustained, efficient, real-world operation. See a more detailed thread by Jing Yu Koh.

Russ Salakhutdinov

22,518 görüntüleme • 2 ay önce

🚨🚨New paper: if you want robots to do bimanual long-horizon tasks well, try RaC: a human in-the-loop data collection protocol that naturally amplifies Recovery & Correction behaviors + trains on them. 📈 data efficiency 10x vs prior results (+many nice properties). 🧵⬇️
0:50

🚨🚨New paper: if you want robots to do bimanual long-horizon tasks well, try RaC: a human in-the-loop data collection protocol that naturally amplifies Recovery & Correction behaviors + trains on them. 📈 data efficiency 10x vs prior results (+many nice properties). 🧵⬇️

Aviral Kumar

25,884 görüntüleme • 10 ay önce

I’ve been a university student coming up on 7 years across multiple disciplines, and this is the most viscerally accurate commentary I’ve seen.
5:02

I’ve been a university student coming up on 7 years across multiple disciplines, and this is the most viscerally accurate commentary I’ve seen.

Bugman Hegel

61,329 görüntüleme • 3 yıl önce

The robotic craftsman from Machina Labs are redefining metal forming! 🤖⚡️ Using AI-driven robots, they shape sheet metal without molds or dies—faster, cheaper, and more flexible than traditional methods. A game-changer for manufacturing!
6:25

The robotic craftsman from Machina Labs are redefining metal forming! 🤖⚡️ Using AI-driven robots, they shape sheet metal without molds or dies—faster, cheaper, and more flexible than traditional methods. A game-changer for manufacturing!

Munro Live

32,866 görüntüleme • 1 yıl önce

Shane Wighton, from the YouTube channel Stuff Made Here, used 3D-printed tooling to form a sheet metal component as part of a concept validation process. Metal manufacturing is essential for all areas of the economy. Because of their strength, stiffness, and long-term durability, metal components are used in applications from appliances to construction parts and car body panels. Traditional metal manufacturing techniques include forming, casting, molding, joining, and machining. Sheet metal forming involves various processes where force is applied to a piece of sheet metal to plastically deform the material into the desired shape, modifying its geometry rather than removing any material. Sheet metals can be bent or stretched into a variety of complex shapes, permitting the creation of complex structures with great strength and a minimum amount of material. Sheet metal forming is the most cost-effective forming procedure today for manufacturing parts in large quantities. It can be highly automated in factories or, at the other end of the spectrum, manually operated in metal workshops for small series parts. It is a versatile, consistent, and high-quality procedure to create accurate metal parts with limited material waste. From metal cans to protective housing for hardware, parts created by sheet metal forming are found everywhere in our daily lives. In this article, learn the basics of sheet metals, the various sheet metal forming processes, and how to reduce the cost of sheet metal forming with rapid tooling and 3D printed dies. For a detailed overview and the step-by-step method, watch our webinar or download our white paper: Research conducted by Shane Wighton. Check out the fantastic 15-minute video on his YouTube channel 'Stuff Made Here'! its top-notch engineering content.:
0:33

Shane Wighton, from the YouTube channel Stuff Made Here, used 3D-printed tooling to form a sheet metal component as part of a concept validation process. Metal manufacturing is essential for all areas of the economy. Because of their strength, stiffness, and long-term durability, metal components are used in applications from appliances to construction parts and car body panels. Traditional metal manufacturing techniques include forming, casting, molding, joining, and machining. Sheet metal forming involves various processes where force is applied to a piece of sheet metal to plastically deform the material into the desired shape, modifying its geometry rather than removing any material. Sheet metals can be bent or stretched into a variety of complex shapes, permitting the creation of complex structures with great strength and a minimum amount of material. Sheet metal forming is the most cost-effective forming procedure today for manufacturing parts in large quantities. It can be highly automated in factories or, at the other end of the spectrum, manually operated in metal workshops for small series parts. It is a versatile, consistent, and high-quality procedure to create accurate metal parts with limited material waste. From metal cans to protective housing for hardware, parts created by sheet metal forming are found everywhere in our daily lives. In this article, learn the basics of sheet metals, the various sheet metal forming processes, and how to reduce the cost of sheet metal forming with rapid tooling and 3D printed dies. For a detailed overview and the step-by-step method, watch our webinar or download our white paper: Research conducted by Shane Wighton. Check out the fantastic 15-minute video on his YouTube channel 'Stuff Made Here'! its top-notch engineering content.:

Formlabs

39,948 görüntüleme • 1 yıl önce

My submission for a backend engineering role in Japanese startup > Mount any SQLite file and get a read-only SQL interface > <3 MB Docker image, built using bun, ts and websockets
0:24

My submission for a backend engineering role in Japanese startup > Mount any SQLite file and get a read-only SQL interface > <3 MB Docker image, built using bun, ts and websockets

Mrinal

159,667 görüntüleme • 11 ay önce

🤖 Ever dreamed of controlling a humanoid robot to perform complex, long-horizon tasks — using just a single Vision Pro? 🎉 Meet CLONE: a holistic, closed-loop, whole-body teleoperation system for long-horizon humanoid control! 🏃‍♂️🧍 CLONE enables rich and coordinated interactive tasks: 🥊 boxing 🏓 table tennis 🤲 object pickup 📦 room arrangement 🤝 handover … and more! 🌀 Our closed-loop error correction powered by LiDAR odometry ensures precision, while motion-captured demonstrations supercharge policy learning — unlocking the full potential of the G1 robot. 🎥 It’s hard to squeeze the magic into 1 minute — check out the full video demo and project page here: 🔗
3:15

🤖 Ever dreamed of controlling a humanoid robot to perform complex, long-horizon tasks — using just a single Vision Pro? 🎉 Meet CLONE: a holistic, closed-loop, whole-body teleoperation system for long-horizon humanoid control! 🏃‍♂️🧍 CLONE enables rich and coordinated interactive tasks: 🥊 boxing 🏓 table tennis 🤲 object pickup 📦 room arrangement 🤝 handover … and more! 🌀 Our closed-loop error correction powered by LiDAR odometry ensures precision, while motion-captured demonstrations supercharge policy learning — unlocking the full potential of the G1 robot. 🎥 It’s hard to squeeze the magic into 1 minute — check out the full video demo and project page here: 🔗

Siyuan Huang

66,556 görüntüleme • 1 yıl önce