Loading...

PODCAST

Embodied AI 101

Embodied AI 101 tracks the latest research at the intersection of AI and physical intelligence. Robotics, manipulation, world models, and the path from digital intelligence to embodied agents — each episode breaks down recent papers and advances in the field. We examine the methodology, real-world implications, and path to deployment. Covering robot learning, perception, control, and the fundamental challenge of grounding intelligence in the physical world. New episodes daily.

technology

All Episodes

29:16

HiDream-O1-Embodied: A World Model Must Earn Its Actions

Embodied AI 101 ·

2026/09/13

HiDream-O1-Embodied is a unified native architecture that ingests text, images, and video and directly outputs physical actions, positioning itself as a step beyond passive world modeling toward active embodied interaction. The architecture aims to close the loop between perception and action within a single model framework.

en

13:33

Axis's Real2Sim2Real Bet: Make the Data Prove Itself

Embodied AI 101 ·

2026/09/13

Axis introduced a Real2Sim2Real pipeline that converts real objects and scenes into high-fidelity physics-accurate simulations and validates learned policies back on hardware, paired with a performance-driven active trajectory curation engine. This data flywheel approach targets improved robot learning policy quality by systematically closing the sim-to-real gap.

en

30:02

A Rubik's Cube, Two Robot Hands, and One Category Error

Embodied AI 101 ·

2026/09/12

GPT-6 Astra is demonstrated solving a Rubik's Cube using dexterous robot hands in a real physical setup, showcasing a significant leap in vision-language-action integration for fine-grained dexterous manipulation. This demo highlights the frontier of embodied AI combining large foundation models with real-world dexterity.

en

33:14

UnifoLM: Predict the Interaction, Not the Whole World

Embodied AI 101 ·

2026/09/12

Unitree's UnifoLM is a 6B-parameter world model trained on 2,500 hours of real robot data that predicts only dynamic pixels and optical flow rather than full frames, enabling efficient robot action prediction. This interaction-centric design reduces computational overhead while maintaining high fidelity for embodied control.

en

32:00

Stop Making the Video Model Remember the World

Embodied AI 101 ·

2026/09/11

A new world model architecture that decouples world-state evolution from rendering: an agent writes executable rules, a persistent engine tracks entities, and a video model renders the output. This separation enables more controllable and interpretable world simulation for embodied agents.

en

13:12

Atlas Turns Phone Scans Into Robot Worlds—But Not Yet...

Embodied AI 101 ·

2026/09/11

Atlas is a unified spatial intelligence model that reconstructs real scenes from 2–3 images, generates 1440p video along camera paths, and supports real-to-sim robotics workflows using Gaussian Splats to create photorealistic 3D digital twins from phone scans. This represents a significant step toward scalable spatial understanding for physical AI and robot training.

en

33:02

W²G-Net: Carrying Instance Identity from Shiny Scrap...

Embodied AI 101 ·

2026/09/10

Purpose This study aims to develop a robust vision-guided framework for instance-level recognition and grasp-based classification of metal waste in dense, multi-scale and highly reflective industrial environments, addressing the challenges...

en

33:45

When Avoidance Is Not Enough: T-CARE Gives UAV Swarms...

Embodied AI 101 ·

2026/09/10

Multi-agent UAV navigation in cluttered environments is challenged by deadlock, starvation-like persistent yielding, oscillation, and unsafe congestion, particularly in narrow passages and dynamically emerging bottlenecks. Existing approaches often rely on...

en

31:53

Teaching Robots What "Better" Means

Embodied AI 101 ·

2026/09/09

A method for preference-based learning applied to dexterous robotic manipulation, enabling robots to learn from human preferences in freeform settings. Discussed publicly around September 8, 2026.

en

29:34

TANGO: Stop Treating a Humanoid Like a Disk on the Floor

Embodied AI 101 ·

2026/09/09

TANGO directly predicts 29-DoF joint actions from RGB observations for zero-shot real-world humanoid navigation, enabling coordinated arms, torso, and gait control in cluttered 3D spaces. Presented at CoRL 2026.

en

31:22

Stop Teaching the VLA Your Camera: Robot-Centric...

Embodied AI 101 ·

2026/09/08

Vision-language-action (VLA) models predict robot actions from visual observations and language instructions. These actions are defined in the robot's own 3D coordinate frame, yet most VLAs observe the scene in 2D. This paper proposes robot-centric pointmaps to bridge this gap for vision-language-action models.

en

34:32

One Policy, Many Bodies: Qwen-VLA's Bid to Unify...

Embodied AI 101 ·

2026/09/08

Embodied intelligence is often studied through specialized models for individual tasks such as manipulation or navigation, resulting in fragmented capabilities and limited generalization across tasks, environments, and robot embodiments.

en

30:37

When the Simulator Lies About Friction: Physics as a...

Embodied AI 101 ·

2026/09/07

Physics-Regularized Learning for Robust Sim-to-Real Tactile Friction Estimation

en

33:17

Seeing Around Corners, Acting at the Edge: REACT as a...

Embodied AI 101 ·

2026/09/07

Edge-based multimodal sensor data fusion with Vision-Language-Action (VLA) model for real-time autonomous vehicle accident avoidance

en

30:56

RoboTok: Turning the Internet Into a Motion Search...

Embodied AI 101 ·

2026/09/06

RoboTok retrieves human hand-pose trajectories from internet videos to automatically generate scalable robot training data for dexterous manipulation tasks. This approach enables large-scale data collection without manual annotation or teleoperation.

en

34:56

VLAct: The VLA Backbone Is Not Just Plumbing

Embodied AI 101 ·

2026/09/06

VLAct proposes a representation-centric continued pre-training approach for Vision-Language-Action models that converts limited robot data into transferable action knowledge, achieving 92.5% success on RoboTwin with only 20% of full training data. The method and weights are fully open-sourced.

en

33:23

Two Models, One Gradient: Making World-Model RL Work...

Embodied AI 101 ·

2026/09/05

Breaks intractable full world-model RL into a heavy global trajectory model combined with a lightweight latent local dynamics approximator, enabling scalable RL for contact-rich humanoid skills without backpropagating through the full model.

en

34:14

GigaBrain-0.7: Scaling Robot Intelligence by...

Embodied AI 101 ·

2026/09/05

Introduces a three-system architecture for scaling embodied foundation models, representing a new approach to large-scale embodied AI.

en

34:25

Geometry Is the Contact Plan: Bringing Surface...

Embodied AI 101 ·

2026/09/04

Geometry-Informed Contact Optimization and Model Predictive Control for Efficient Contact-Rich Manipulation

en

30:12

A Swarm That Knows Its Margin: Safe Multi-Drone...

Embodied AI 101 ·

2026/09/04

Safe and Scalable Multi-Drone Payload Transport via CBF-Based Reinforcement Learning With Zero-Shot Sim-to-Real Transfer

en

379 results

Similar Podcasts

Lead With AI

Tamara Nall

Engineering the Intelligent Enterprise

PRASAD BHONDE

Double Tap

Steven Scott, Shaun Preece, Steven...

The Good Tech Companies

HackerNoon

BlockHash

BlockHash LLC