Video yükleniyor...

Video Yüklenemedi

Ana Sayfaya Dön

Boston Dynamics’ robot-behavior team lead highlights three core initiatives aimed at advancing Atlas’s dexterity: ⦿ Reinforcement learning in simulation ⦿ Whole-body teleoperation to collect data for imitation learning ⦿ Tactile sensing grippers

TheHumanoidHub's profile picture

The Humanoid Hub

60,420 subscribers

18,029 görüntüleme • 1 yıl önce •via X (Twitter)

Bilim & TeknolojiHaberler & PolitikaEğitim
LIVE

Anya Rossi• Live Now

Private livecam show

7 Yorum

The Humanoid Hub profil fotoğrafı
The Humanoid Hub1 yıl önce

Source:

Rainmaker profil fotoğrafı
Rainmaker2 yıl önce

💡 Learn how Reinforcement Learning can boost your trading performance! In this free Substack article I share full code of a trading algorithm based on Reinforcement Learning that beats other Machine Learning models as well as simply buying and holding the stock.

Phil Trubey profil fotoğrafı
Phil Trubey1 yıl önce

The biggest thing from vid was that automatic retry when initial attempt fails is an emergent behavior that comes with sim training at scale. Very very interesting. Bolsters Tesla approach building out their huge compute clusters and manufacturing a fleet for teleop training.

Judge DredD profil fotoğrafı
Judge DredD1 yıl önce

Bad optimus copy

The Robot Services Exchange profil fotoğrafı
The Robot Services Exchange1 yıl önce

solid

VentureMind AI profil fotoğrafı
VentureMind AI1 yıl önce

Working together to build the future 🤝

MatRat21 profil fotoğrafı
MatRat211 yıl önce

Hell yeah Atlas!

Benzer Videolar

TWIST: a real-time teleoperation system for humanoid robots to mimic whole-body motions. Reference motion data is generated by retargeting human motion-capture data to the robot. Then, the controller is trained in simulation using reinforcement learning and behavior cloning.
0:58

TWIST: a real-time teleoperation system for humanoid robots to mimic whole-body motions. Reference motion data is generated by retargeting human motion-capture data to the robot. Then, the controller is trained in simulation using reinforcement learning and behavior cloning.

The Humanoid Hub

46,715 görüntüleme • 1 yıl önce

Reinforcement learning is used to speed the production of behavior for the Boston Dynamics Atlas humanoid robot. At the heart of the learning process is a physics-based simulator that generates training data for a variety of maneuvers.
1:10

Reinforcement learning is used to speed the production of behavior for the Boston Dynamics Atlas humanoid robot. At the heart of the learning process is a physics-based simulator that generates training data for a variety of maneuvers.

RAI Institute

76,564 görüntüleme • 1 yıl önce

Using reinforcement learning, we trained policies for Boston Dynamics Spot that allow the robot to achieve record running speeds of 11.5 mph (5.2 m/s) — over three times faster than Spot's default max speed.
0:49

Using reinforcement learning, we trained policies for Boston Dynamics Spot that allow the robot to achieve record running speeds of 11.5 mph (5.2 m/s) — over three times faster than Spot's default max speed.

RAI Institute

98,538 görüntüleme • 1 yıl önce

Happy to announce mimic-one: a Scalable Model Recipe for General Purpose Robot Dexterity, the culmination of years of research work in dexterous manipulation with imitation learning.
0:29

Happy to announce mimic-one: a Scalable Model Recipe for General Purpose Robot Dexterity, the culmination of years of research work in dexterous manipulation with imitation learning.

Elvis Nava

20,238 görüntüleme • 1 yıl önce

Boston Dynamics' Atlas showcases reinforcement learning policies trained using a motion capture suit, developed in partnership with RAI Institute. Both Boston Dynamics and RAI Institute are owned by Hyundai Group.
0:56

Boston Dynamics' Atlas showcases reinforcement learning policies trained using a motion capture suit, developed in partnership with RAI Institute. Both Boston Dynamics and RAI Institute are owned by Hyundai Group.

The Humanoid Hub

455,189 görüntüleme • 1 yıl önce

The AMO (Adaptive Motion Optimization) project, developed by researchers at UC San Diego, introduces a novel framework for enhancing the dexterity and adaptability of humanoid robots. This system integrates simulation-to-real reinforcement learning with trajectory optimization to achieve real-time, adaptive whole-body control Video Credit: USD #engineering #technology #robotics #robots
1:39

The AMO (Adaptive Motion Optimization) project, developed by researchers at UC San Diego, introduces a novel framework for enhancing the dexterity and adaptability of humanoid robots. This system integrates simulation-to-real reinforcement learning with trajectory optimization to achieve real-time, adaptive whole-body control Video Credit: USD #engineering #technology #robotics #robots

Wevolver

13,632 görüntüleme • 1 yıl önce

Very impressed by the new Tesla Optimus end2end skill learning video! Our TRILL work ( spills some secret sauce: 1. VR teleoperation, 2. deep imitation learning, 3. real-time whole-body control. It's all open-source! Dive in if you're into humanoids! 👾
0:46

Very impressed by the new Tesla Optimus end2end skill learning video! Our TRILL work ( spills some secret sauce: 1. VR teleoperation, 2. deep imitation learning, 3. real-time whole-body control. It's all open-source! Dive in if you're into humanoids! 👾

Yuke Zhu

50,666 görüntüleme • 2 yıl önce

Boston Dynamics recently showcased Atlas performing a powerful backflip with remarkable precision. The robot landed cleanly on its toes and immediately launched into a second backflip. This advanced capability is powered by a whole-body learning system developed by the RAI Institute and deployed by Boston Dynamics.
0:24

Boston Dynamics recently showcased Atlas performing a powerful backflip with remarkable precision. The robot landed cleanly on its toes and immediately launched into a second backflip. This advanced capability is powered by a whole-body learning system developed by the RAI Institute and deployed by Boston Dynamics.

Space and Technology

78,087 görüntüleme • 14 gün önce

Using our brain simulator, we’ve trained a reinforcement learning agent to maximize bits per second. Here is the RL policy converting brain data to cursor control in simulation:
0:29

Using our brain simulator, we’ve trained a reinforcement learning agent to maximize bits per second. Here is the RL policy converting brain data to cursor control in simulation:

Neuralink

18,410 görüntüleme • 1 yıl önce

🎉Excited to share a fun little hardware project we’ve been working on. GELLO is an intuitive and low cost teleoperation device for robot arms that costs less than $300. We've seen the importance of data quality in imitation learning. Our goal is to make this more accessible 1/n
0:50

🎉Excited to share a fun little hardware project we’ve been working on. GELLO is an intuitive and low cost teleoperation device for robot arms that costs less than $300. We've seen the importance of data quality in imitation learning. Our goal is to make this more accessible 1/n

Philipp Wu

162,934 görüntüleme • 2 yıl önce

Today, we're joined by Nikita Rudin, co-founder and CEO of Flexion Robotics to discuss the gap between current robotic capabilities and what’s required to deploy fully autonomous robots in the real world. Nikita explains how reinforcement learning and simulation have driven rapid progress in robot locomotion—and why locomotion is still far from “solved.” We dig into the sim2real gap, and how adding visual inputs introduces noise and significantly complicates sim-to-real transfer. We also explore the debate between end-to-end models and modular approaches, and why separating locomotion, planning, and semantics remains a pragmatic approach today. Nikita also introduces the concept of "real-to-sim", which uses real-world data to refine simulation parameters for higher fidelity training, discusses how reinforcement learning, imitation learning, and teleoperation data are combined to train robust policies for both quadruped and humanoid robots, and introduces Flexion's hierarchical approach that utilizes pre-trained Vision-Language Models (VLMs) for high-level task orchestration with Vision-Language-Action (VLA) models and low-level whole-body trackers. Finally, Nikita shares the behind-the-scenes in humanoid robot demos, his take on reinforcement learning in simulation versus the real world, the nuances of reward tuning, and offers practical advice for researchers and practitioners looking to get started in robotics today. 🗒️ For the full list of resources for this episode, visit the show notes page: 📖 CHAPTERS =============================== 00:00 - Introduction 04:07 - Is robot locomotion solved? 06:04 - Sim-to-real gap 08:58 - Adding semantics to policies 09:42 - Modular vs end-to-end architectures 10:29 - Planner model 12:21 - Adapting RL techniques from quadrupeds to humanoids 15:39 - Behind robot demos 18:09 - Humanoid robots in home environments 22:03 - Training approach 23:56 - VLA models 27:59 - Closing the sim-to-real gap 32:55 - Task orchestration using VLMs 36:38 - Tool use 38:10 - Model hierarchy 43:37 - Simulator versus simulation environment 44:57 - Combining imitation learning and reinforcement learning 46:42 - RL in real world versus RL in simulation 52:58 - Reward tuning and value functions in robotics 56:38 - Predictions 1:00:10 - Humanoids, quadropeds, and wheeled platforms 1:02:45 - Advice, recommended robot kits, and community pla
1:06:07

Today, we're joined by Nikita Rudin, co-founder and CEO of Flexion Robotics to discuss the gap between current robotic capabilities and what’s required to deploy fully autonomous robots in the real world. Nikita explains how reinforcement learning and simulation have driven rapid progress in robot locomotion—and why locomotion is still far from “solved.” We dig into the sim2real gap, and how adding visual inputs introduces noise and significantly complicates sim-to-real transfer. We also explore the debate between end-to-end models and modular approaches, and why separating locomotion, planning, and semantics remains a pragmatic approach today. Nikita also introduces the concept of "real-to-sim", which uses real-world data to refine simulation parameters for higher fidelity training, discusses how reinforcement learning, imitation learning, and teleoperation data are combined to train robust policies for both quadruped and humanoid robots, and introduces Flexion's hierarchical approach that utilizes pre-trained Vision-Language Models (VLMs) for high-level task orchestration with Vision-Language-Action (VLA) models and low-level whole-body trackers. Finally, Nikita shares the behind-the-scenes in humanoid robot demos, his take on reinforcement learning in simulation versus the real world, the nuances of reward tuning, and offers practical advice for researchers and practitioners looking to get started in robotics today. 🗒️ For the full list of resources for this episode, visit the show notes page: 📖 CHAPTERS =============================== 00:00 - Introduction 04:07 - Is robot locomotion solved? 06:04 - Sim-to-real gap 08:58 - Adding semantics to policies 09:42 - Modular vs end-to-end architectures 10:29 - Planner model 12:21 - Adapting RL techniques from quadrupeds to humanoids 15:39 - Behind robot demos 18:09 - Humanoid robots in home environments 22:03 - Training approach 23:56 - VLA models 27:59 - Closing the sim-to-real gap 32:55 - Task orchestration using VLMs 36:38 - Tool use 38:10 - Model hierarchy 43:37 - Simulator versus simulation environment 44:57 - Combining imitation learning and reinforcement learning 46:42 - RL in real world versus RL in simulation 52:58 - Reward tuning and value functions in robotics 56:38 - Predictions 1:00:10 - Humanoids, quadropeds, and wheeled platforms 1:02:45 - Advice, recommended robot kits, and community pla

The TWIML AI Podcast

22,264 görüntüleme • 5 ay önce

Boston Dynamics' Handle robot is designed for warehouses, using a rear counterweight and suction grippers to efficiently and smoothly move heavy loads.
0:20

Boston Dynamics' Handle robot is designed for warehouses, using a rear counterweight and suction grippers to efficiently and smoothly move heavy loads.

Science girl

1,433,528 görüntüleme • 1 yıl önce

In the last month, we’ve been building an open-source framework for robot learning and sim-to-real transfer, made for RL whole-body control from simple walking to complex human imitation Check out the details on HN: Get started in 5 minutes ⬇️
0:38

In the last month, we’ve been building an open-source framework for robot learning and sim-to-real transfer, made for RL whole-body control from simple walking to complex human imitation Check out the details on HN: Get started in 5 minutes ⬇️

K-Scale Labs

75,983 görüntüleme • 1 yıl önce

Boston Dynamics' Atlas robot made its first in-person debut at CES 2026, where it showcased autonomy, teleoperation and how the two go together.
4:00

Boston Dynamics' Atlas robot made its first in-person debut at CES 2026, where it showcased autonomy, teleoperation and how the two go together.

CNET

35,648 görüntüleme • 4 ay önce

In order for robots to be deployed in the real world, performing tasks of real value, they must be reliable. Unfortunately, even more, most robotic demos work maybe 70-80% of the time at best. The way to get better reliability is to do real-world reinforcement learning: having the robot teach itself how to perform the task up to a high level of success. The key to doing this is to start with a core of expert human data, use that to train a policy then iteratively improve it, until finally finishing with on-policy reinforcement learning. Kun Lei talks through a unified framework for imitation and reinforcement learning based on PPO, which enables this improvement process. In this episode, Kun Lei explains the theory behind his reinforcement learning method and how it allowed his robot to run in a shopping mall juicing oranges for seven hours at a time, among experiments on a wide variety of tasks and embodiments. Watch episode 58 of RoboPapers now, hosted by Michael Cho - Rbt/Acc and Chris Paxton!
1:11:51

In order for robots to be deployed in the real world, performing tasks of real value, they must be reliable. Unfortunately, even more, most robotic demos work maybe 70-80% of the time at best. The way to get better reliability is to do real-world reinforcement learning: having the robot teach itself how to perform the task up to a high level of success. The key to doing this is to start with a core of expert human data, use that to train a policy then iteratively improve it, until finally finishing with on-policy reinforcement learning. Kun Lei talks through a unified framework for imitation and reinforcement learning based on PPO, which enables this improvement process. In this episode, Kun Lei explains the theory behind his reinforcement learning method and how it allowed his robot to run in a shopping mall juicing oranges for seven hours at a time, among experiments on a wide variety of tasks and embodiments. Watch episode 58 of RoboPapers now, hosted by Michael Cho - Rbt/Acc and Chris Paxton!

RoboPapers

18,541 görüntüleme • 5 ay önce

LimX Dynamics P1 is a versatile bipedal robot equipped with reinforcement learning capabilities, enabling it to master challenging terrains such as forests and ditches while maintaining its balance.
0:45

LimX Dynamics P1 is a versatile bipedal robot equipped with reinforcement learning capabilities, enabling it to master challenging terrains such as forests and ditches while maintaining its balance.

Science girl

3,295,085 görüntüleme • 2 yıl önce

Mecka is excited to power EgoVerse: a growing ecosystem for robot learning from egocentric human data Proven across 4 leading research labs, EgoVerse data consistently boosts robot performance We release the tools for anyone to collect data, inference/train, and contribute! 🧵
0:56

Mecka is excited to power EgoVerse: a growing ecosystem for robot learning from egocentric human data Proven across 4 leading research labs, EgoVerse data consistently boosts robot performance We release the tools for anyone to collect data, inference/train, and contribute! 🧵

Josh

102,105 görüntüleme • 3 ay önce

TidyBot++ is an open-source mobile manipulator optimized for household tasks. The robot can be teleoperated using a mobile phone interface, enabling data collection for imitation learning.
0:22

TidyBot++ is an open-source mobile manipulator optimized for household tasks. The robot can be teleoperated using a mobile phone interface, enabling data collection for imitation learning.

The Humanoid Hub

29,188 görüntüleme • 1 yıl önce

Learning by looking. Hyundai’s Boston Dynamic Robot.
1:20

Learning by looking. Hyundai’s Boston Dynamic Robot.

Brian Roemmele

14,267 görüntüleme • 26 gün önce

Bernt Børnich says teleoperation data and synthetic data are like a crutch to bootstrap your way to robots learning in the real world.
1:52

Bernt Børnich says teleoperation data and synthetic data are like a crutch to bootstrap your way to robots learning in the real world.

The Humanoid Hub

27,977 görüntüleme • 1 yıl önce