Robotics & World Models Reading Club 29: Data, Hardware & the Scaling Problem in Robotics. SF 9/19

A high-signal reading group for AI researchers & builders pushing the frontiers of robotic world models, WAMs, and embodied intelligence. In our previous sessions, we brought together researchers and engineers from Boston Dynamics, Dyna, Google DeepMind, NVIDIA, Stanford, UC Berkeley, Physical Intelligence, Tesla, Generalist, Rhoda AI, and leading Bay Area robotics startups.

Hosted by Junfan Zhu & Aurora Feng.

​​​​Reading Club 29's Core Theme — Data, Hardware & the Scaling Problem in Robotics

  • The Best Glove Is Rigid: Why Robot Hands Should Be Built Around Their DataJoe Dong — Founder & CTO, Chestnut Robotics

  • Can Classical Robotics Still Scale? Lessons from a Robot That Bought CoffeeTony Pratkanis — ex-Stanford, Google

Keynote 1: The Best Glove Is Rigid: A Preview of a Hand Built Around Its Data

Joe Dong, Founder & CTO of Chestnut Robotics

He builds dexterous hands together with the data systems that train them. He previously worked on autonomous driving at Waymo and Xpeng Motors. He writes Physical AI Deep Dives at newsletter.joedong.ai. His recent essays argue that data sourcing, not degrees of freedom, decides what a robot hand should be.

Egocentric human video is becoming the substrate for robot foundation models, but fine manipulation is won or lost at the instant of contact, and that instant is the one thing video cannot measure: the contact hides behind the finger, and force never reaches the pixels. This talk argues that the data glove filling that gap should be rigid, and previews the hand and capture system we built on that premise. (1) Hardware. I will argue against the field's default. Soft is the root of every glove problem: calibration that drifts each time you put it on, no fixed geometry for locating contact, tactile sensors mounted on deforming fabric. Rigid functional surfaces on anatomically aligned joints turn contact localization into forward kinematics and give touch a stable mount, at the cost of a dexterity tax I will argue is worth paying. (2) Co-design. Every retargeting pass from a human hand to a robot hand leaks contact precision, so the glove's geometry should be the robot's geometry. The right dexterous hand is the one a human can capture, not the most dexterous one. I will show the co-designed pair, Aero UMI and Aero Hand, and what fell out of reversing the usual order of hand and capture device. (3) From capture to policy. I will cover the engineering that makes human demonstration and robot execution correspond: ergonomic fit as a data-quality problem, three distinct meanings of accuracy and the active vision-based calibration that ties them together, temporal alignment, and geometry-aware inpainting. Then how we validate the resulting data with Diffusion Policy, and where a small model for fine manipulation sits beneath a long-horizon planner.

Keynote 2

Tony Pratkanis (Ex Stanford + Google, Currently Stealth Startup)

In the now-distant year of 2012, my collaborators and I completed, via robotic means, a fully autonomous procurement of a cup of coffee. For this purpose, we utilized a PR2 - a full-sized wheeled humanoid robot of the era. The robot departed from our laboratory, proceeding through numerous doors and an elevator, until it did arrive upon a traditional coffee shop. Therewithin, the robot legally purchased a coffee, navigated a return through those same doors and elevator, and subsequently delivered the beverage to our academic advisor.

This discourse shall present an overview of the methodology of the aforementioned, and the implications it holds for the scalability of classical robotics algorithms such as SLAM, classical planners, and hand-crafted scripts. Furthermore, I shall also present my subsequent research concerning the interleaving of digital robot application stores with task-level classical planning algorithms, thereby scaling robotics by crowd-sourcing hand-crafted robot programs globally. In conclusion, we shall examine contemporary instances of these algorithms as they are presently deployed within the active commercial robots of the current day, and the challenges which these methods have yet to overcome.

​​​​​​​​​​​​​Pre-Readings

​​​​​​​​​​​Location

San Francisco

​​​​​​​​​​​​​​Date & Time

Saturday, September 19, 2026 | 2:00 PM – 5:00 PM

​​​​​​​​​​​​​​​Join Discord Community

https://discord.gg/WH7DrTHRXK

​​​​​​​​​​​​​​Follow Saturday Robotics on X

/saturdayrobotic

​​​​​​​​​​​​​​Agenda

2:00 PM – 2:30 PM Door Opens & Social

  • Food 😋, beverages🧋 and UNLIMITED strawberries 🍓 (our official reading club fruits ☺️😄).

2:30 PM – 4:00 PM Keynotes by Joe Dong and Tony Pratkanis

YouTube Recording: TBD (We are looking for recording volunteers)

4:00 PM – 5:00 PM Q&A, ​open-floor roundtable (10–20 min per topic) on spotlight papers or any paper you’d like to highlight. Feel free to share why the paper matters and its technical details.

​Future events

#iros-reading-club-31-0928: 🍾 IROS 2026 x Saturday Robotics — Robotics Research Night | Reading Club 31. Pittsburgh 9/28

​​​​​​​​​​​​​​​​​​​​Past events

#reading-club-28-0912: Booster T2: The Next Frontier of Open Humanoid Robotics

#reading-club-27-0905: Rethinking Robot Development: Co-Designing Morphology, Sensing, and Learning

Chenyang Ma (Applied Intuition, UNC)

#reading-club-26-0829: Sunnyvale 8/29

#reading-club-25-0822: Contact-Rich Robot Learning from Human Videos and Tactile. SF 8/22

Kelin Yu (Maryland, Amazon FAR)

#reading-club-24-0820: Saturday Robotics x EAI @ WRC Beijing | PHYSICAL AI UNPLUGGED —— 与 Physical AI 研究者的一场非正式夜谈

• 李晓帆 — Head of World Models, X Square Robot (自变量机器人)

• Tongzhou Mu — Research Scientist, Rhoda AI

• Haoyi Niu — University of California, Berkeley

• Yufei Wang — Member of Technical Staff, Genesis AI

• Shibo Zhao — Research Scientist, Amazon FAR

• 郑思鹏 — Technical Partner, Being Beyond

• Tianle Zhang — Embodied AI Researcher, JD Explore Academy

• Yihang Li — Embodied AI Researcher, JD Explore Academy

#reading-club-23-0815: Engineering Robotic Simulators for Evaluation and Beyond. SF 8/15

Kaifeng Zhang (Columbia, World Labs)

#reading-club-22-0808: ODEWorld: A Continuous Predictive Architecture via Physical-Time Flow. SF 8/8

Haoyi Niu (UC Berkeley)

#reading-club-21-0801: Vision-Language-Kinematics Supervision for Perception-Based Humanoid Loco-Manipulation — SF 8/1

Yen-Jen Wang (UC Berkeley, Amazon FAR)

#reading-club-20-0725: Agentic Robotics Models (ENPIRE and Cap-X), Mountain View 07/25

Haoru Xue (UC Berkeley)

Kris Hauser (Samsung Research America, Robot Intelligence Lab)

#SIGGRAPH-reading-club-19-0722: SIGGRAPH x Saturday Robotics — World Models for Robotics: Bridging Graphics, Simulation & Physical Intelligence | Reading Club 19, LA 07/22

#reading-club-18-0718: Causal World Models For Real-World Intelligence. SF 07/18

Guanming Wang & Bill (General Instinct, YC P26)

Feng Fan (UCSD, Aether AI)

#reading-club-17-0711: Soft Tactile-Centric Multimodal Intelligence Toward Safe and Dexterous Manipulation. SF 07/11

Quan Luu, Purdue.

#private-lunch-icml-0709: Saturday Robotics x ICML Private Lunch (Seoul)

#reading-club-16-0704: The Embodied AI Hardware Stack — Supply Chain, Sensors, and the Data Flywheel — SF 07/04

Jerry Huang, Robotics Center of Silicon Valley.

#reading-club-15-0627: Scaling Touch: Flexible Tactile Skin for Dexterous Manipulation

Binghao Huang, Columbia, Amazon FAR.

#deep-tech-week-14-0625: Deep Tech Week Research Night

SPEAR: A Simulator for Photorealistic Embodied AI Research. (ECCV 2026 accepted) by Mike Roberts, Senior Research Scientist, Adobe Research.

Bogdan Cristei, Venture Partner at SHACK15 Ventures.

Simone Totaro, CTO at Saturn Dynamics.

Shumo Chu, CEO at General Intelligence Labs.

Margaret Zhang, CEO at ThirdBrain Labs.

#reading-club-13-0620: HumanEgo: Train Robot Policy from 30 min Egocentric Videos — SF 0620

#reading-club-12-0613: Origami Robotics (YC W26) on Dexterity

#cvpr-denver-11-0606: 🤖🥘 Saturday Robotics x Manycore Tech x Neural Motion | CVPR 2026 Denver Research Night | Robotics & World Models Reading Club 11

Junfan Zhu & Aurora Feng, Founders of Saturday Robotics

Anthony Zhao, Head of North America at Manycore Tech SpacialVerse

Aurora Feng, Founder at Neural Motion. NM-GenET.

Max Zhaoshuo Li, Robotics and World Model Tech Lead at NVIDIA Cosmos. Cosmos 3.

Xiaofan Li, World Model Tech Lead at X Square Robot. WALL-WM.

Zesen Zhao, University of Michigan. Test-Time Scaling for World Action Models via Zero-Shot Geometric Verification.

Pengyi Liao. VGGT-Ω: From 3D Reconstruction to Scalable Spatial Representation.

Jie Wang, University of Pennsylvania, GRASP Lab. Toward a Robotics MMLU: Lessons from Sim & Real Evaluations of Generalist Policies.

Gordon Qian, Senior AI Researcher at Snap. Diffusion-DRF: Free, Rich, and Differentiable Reward for Video Diffusion Fine-Tuning.

#reading-club-10-0530: Bringing Robots to Life — Learning Humanoid Instincts from the Body Up | San Francisco 0530

Haochen Shi (Stanford, co-advised by Karen Liu & Shuran Song)

#private-dinner-01-0529: Robo Plov x Saturday Robotics

#reading-club-09-0523: CVPR Warm-up & Founders Spotlight — DeltaWorld + VisuoTactile Dexterous Hands

Tommie Kerssies (Amazon Frontier AI & Robotics)

Arjun Subramaniam (Factory Intelligence)

#reading-club-08-0516: Embodied Human Data as the “Internet of Motion and Behavior”

Ryan Punamiya (NVIDIA Gear, Georgia Tech)

#reading-club-07-0509: Learning to Dream: World Models, Imagination, Path to Foundation Models for Control

Ahmet Şemi ASARKAYA (Agility Robotics)

#reading-club-06-0502: Evolution of Video World Models for Robotics

Tongzhou Mu (Rhoda AI)

#reading-club-05-0425: World Models for Physical Intelligence: From Predictive Brains to Embodied Robots

Daniel Dugas & Sergio Arnaud (Meta FAIR)

#reading-club-04-0418: Abstractions of the Physical World for Decision-Making

Siming He (UC Berkeley)

#reading-club-03-0411: Robotic Policy Adaptation

Haoyi Niu (UC Berkeley)

#reading-club-02-0404: JEPA Zoo

Julian Saks (/JulianSaks)

#reading-club-01-0328

Join Discord Community

https://discord.gg/WH7DrTHRXK

Follow Saturday Robotics

/saturdayrobotic

Follow YouTube

/saturdayrobotic

Subscribe to Luma Calendar

https://luma.com/saturdayrobotic