Junfan Labs Forum #36: Building Maker Arm for Dream Lab's Dream-1. SF 10/31

Name: Junfan Labs Forum #36: Building Maker Arm for Dream Lab's Dream-1. SF 10/31
Start: 2026-10-31T14:00:00-07:00
End: 2026-10-31T17:00:00-07:00
Location: San Francisco, CA

📅 October 31, 2026

🕐 2:00 PM PT

📍 San Francisco

👥 2 attending

🎟️ $20 to $32.8 · Approval required

What we're seeing

Junfan Zhu often sells out · 3 of their last 8 events did
Approval required · venue shared on acceptance

About This Event

Junfan Labs Forum #36: Building Maker Arm for Dream Lab's Dream-1. SF 10/31

Junfan Labs is an independent research and technology initiative focused on robotics, world models, and Physical AI. Our mission is to help build, evaluate, and deploy AI systems that can operate reliably in the physical world.

Physical AI advances when two loops stay connected. The technical loop takes ideas from papers to prototypes, simulation, real-world testing, evaluation and failure analysis. The ecosystem loop carries them from researchers to builders, startups, industry problems and deployment feedback, which become the next research questions. Junfan Labs works where the two loops meet: research, real-world evaluation, technical advisory and industry partnerships, plus Demo Days and research nights that put new ideas on real hardware.

Hosted by Junfan Zhu.

Reading Club 36's Core Theme: Maker Arm, developing open-source manipulator for Dream-1

Keynote: Ryan Chan - Co-founder, CEO @ MakerMods

World action models are opening new possibilities for language-guided manipulation, but turning a compelling demonstration into an accessible research platform requires more than a capable model. This talk shares the engineering perspective behind Maker Arm, an affordable, 3D-printed, quasi-direct-drive robot arm, and MakerMods’ work integrating Dream Labs’ Dream-1 with Maker Arm and Metal Arm. I’ll discuss mechanical design choices, teleoperation, camera setup, calibration, and the software needed to connect model outputs to physical actions. Through examples of language-guided manipulation, I’ll distinguish demonstrated capabilities from questions that require systematic evaluation, including task generalization, repeatability, and sensitivity to the physical setup. The central question is practical: how can we lower the cost and integration burden enough for more students and researchers to experiment with robot learning?

Pre-Readings

Location

San Francisco

Date & Time

Saturday, October 31, 2026 | 2:00 PM – 5:00 PM

Agenda

2:00 PM – 2:30 PM Door Opens & Social

2:30 PM – 3:30 PM Keynote by Ryan Chan

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

3:30 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.