Robotics Hardware Hackathon Demo Day: Junfan Labs × Abaka AI. Mountain View 10/24 | Junfan Labs #35

Robotics Hardware Hackathon Demo Day: Junfan Labs × Abaka AI. Mountain View 10/24 | Reading Club #35

Build this week. Demo on Saturday.

Junfan Labs Robotics Hackathon #35 is a take-home robotics hardware hackathon co-hosted by Junfan Labs, Abaka AI and Saturday Robotics. It ends in a demo day at Abaka AI in Mountain View. Junfan Labs events continue the Saturday Robotics count, so this is edition #35.

📢 Update: new venue, same format. Demo day is now at Abaka AI in Mountain View, within walking distance of the Mountain View Caltrain station. The format hasn't changed: a take-home build on your own hardware, then a demo day on Saturday from 2:00 to 5:00 pm. On Friday at 2:00 pm we'll send everyone a check-in link. If an earlier email or calendar invite said San Francisco, Robotics Center or 9:00 am–9:00 pm, that was the old plan.

• 🛠️ Build: Tue, Oct 20 → Sat, Oct 24 at 2:00 pm PT, wherever you are, on your own hardware

• ✅ Check-in link: Fri, Oct 23 at 2:00 pm PT (work in progress is fine)

• 🎤 Demo Day: Sat, Oct 24, 2:00–5:00 pm PT: presentations, judging, then awards

• 📍 Venue: Abaka AI, 990 Villa St, Mountain View, CA 94041 (walking distance from Mountain View Caltrain)

Logistics

1️⃣ Tue, Oct 20 · Hacking starts. Team up however you like and build on your own hardware. Joining late is fine.

2️⃣ Fri, Oct 23, 2:00 pm PT · Check-in link goes out. We'll send everyone a link to check in your demo project with your team info and a demo link or a short description. Work in progress is fine. Check-ins tell us how many demos to expect on Saturday, so please check in if you plan to demo.

3️⃣ Sat, Oct 24, 2:00 pm PT · Final deadline. Hacking ends. Update your check-in with your demo video, slides and code. If lots of teams check in, we'll select a shorter lineup. If the numbers are manageable, every team demos, one by one.

4️⃣ Sat, Oct 24, 2:00–5:00 pm · Presentations & judging at Abaka AI in Mountain View, live or by video.

5️⃣ About 5:00 pm · Awards, including the People's Choice vote.

Who it's for

Researchers, founders, engineers and students working on robot learning, VLA policies, world models, teleoperation and data collection, or anyone with a robot side project. Come solo or as a team of up to 5. Looking for teammates? Post in WhatsApp or Discord (links below).

What to build

Open theme: make a real robot do something new. We'd love to see dexterous manipulation, teleoperation and data collection, robot learning and VLA policies, world models and simulation, and creative hardware builds. There are no separate tracks: the category awards below play that role, and the check-in asks which ones fit your project. Working with robot data? The Abaka AI Award goes to the best data-driven project. Projects you've already been working on are welcome; tell the judges what you built or improved this week.

What to bring on Saturday

• Your robot, if it travels, plus chargers, batteries and a power strip

• A laptop with your demo and a backup video

Hardware isn't provided. If you're bringing hardware, we'll email checked-in teams a setup time before Saturday.

🏆 Prizes & Awards

Winners are announced on stage at Demo Day and receive a certificate signed by the program chairs. Every team that presents on Saturday, live or by video, is eligible for every award.

Overall

• 🥇 First Place Award

• 🥈 Second Place Award

• 🥉 Third Place Award

Overall places are scored on technical depth (30%), novelty (25%), real-world impact (25%) and demo quality (20%).

Category awards, decided by the judging panel

• ⚙️ Best Technical Achievement: the most technically ambitious, rigorously engineered project

• 💡 Most Novel Idea: the most original idea, one that makes us see robots differently

• 🚀 Best Real-World Deployment: the project closest to working reliably outside the lab

• 🌐 Best World Model Project: the most compelling use of world models, simulation or learned dynamics on a robot

• 🔧 Best Hardware Hack: the most creative build, modification or integration on real robot hardware

Audience award

• ❤️ People's Choice: voted by everyone at Demo Day

Co-host award

• 🤖 Abaka AI Award, presented by our co-host Abaka AI: the best data-driven robotics project, from data collection and curation to evaluation

Judges & Program Chairs

• Junfan Zhu, Founder & CEO, Junfan Labs & Saturday Robotics; robotics and embodied-AI researcher; reviewer for NeurIPS 2026, EMNLP 2026 and AAAI 2027; program committee, IROS 2026 workshop on Physical World Models for Scaling Embodied AI; organizer, ACM SIGGRAPH 2026 Birds of a Feather on World Models for Robotics; co-author, IEDD (Scientific Data, Nature Portfolio), Agents' Last Exam (NeurIPS 2026); judge, Embodied Metal Hackathon 2026.

• Anjiang Wei, Research Scientist at Google DeepMind. PhD in Computer Science from Stanford University. His research interests include large language models, coding agents, reinforcement learning, and robotics foundation models.

• Pinxin Liu, Member of Technical Staff, Rhoda AI (video generative models for robotics); first author, GestureLSM (ICCV 2025).

• Huaijing Hong, Member of Technical Staff, MaxInsights.

• Fanyu You, Abaka AI.

FAQ

1️⃣ Is this a take-home hackathon? Where is demo day? Yes. Build from Tuesday, Oct 20 until 2:00 pm on Saturday, Oct 24, wherever you like. Demo day is Saturday, 2:00–5:00 pm, at Abaka AI, 990 Villa St, Mountain View, within walking distance of the Mountain View Caltrain station.

2️⃣ Hacking already started. Can I still join? Yes. Start now and check in when the link goes out on Friday, Oct 23 at 2:00 pm.

3️⃣ What can we build? Can it be an existing project? Anything on real robot hardware (see What to build). Existing projects are welcome; tell the judges what you built or improved this week.

4️⃣ Do you provide robots or hardware? No. Build on your own robot, arm, kit or custom rig. If it can't travel, demo by video.

5️⃣ How do teams work? Form your own: solo or up to 5 people. One check-in per team; the team lead checks in for everyone.

6️⃣ How and when do we submit? Check in through the link we send on Friday, Oct 23 at 2:00 pm. Then update your entry with your video, slides and code by the final deadline, Saturday, Oct 24 at 2:00 pm.

7️⃣ Do we need a finished demo by Friday? No. A short description or a rough video link is enough. The check-in helps us plan Saturday's lineup.

8️⃣ Will every team present? If the numbers are manageable, yes: every team demos, one by one. If lots of teams check in, we'll select a shorter lineup from the check-ins and let you know on Saturday before demos start. A short demo video helps.

9️⃣ Live demo or video? What if we can't come in person? Either works, and every team that presents, live or by video, is eligible for every award. Can't make it to Mountain View? Submit a 1–3 minute video and say so when you check in.

🔟 Who judges, and who owns what we build? The program chairs above judge. Your team keeps full ownership of everything you build.

Hosts

Junfan Labs: Junfan Labs is an independent research and technology initiative focused on robotics, world models, and Physical AI. We advance research in AI evaluation and intelligent systems, while providing technical advisory and collaborating with industry to translate frontier research into real-world applications. Our mission is to help build, evaluate, and deploy AI systems that can operate reliably in the physical world.

Abaka AI (co-host): Abaka AI is a full-lifecycle data partner for AI teams — data collection, curation, annotation, and model evaluation across text, image, audio, video, 3D, and agent/RL data. We work with 1,000+ AI labs, partners, and research groups across autonomous driving, generative AI, and embodied AI. Follow us on LinkedIn and X.

Join us

• Junfan Labs WhatsApp: https://chat.whatsapp.com/F9OowndExwyIyqMQ5c6qpm

• Saturday Robotics WhatsApp: https://chat.whatsapp.com/HYxhjyaBbBuFZUlrV0AIaV

• Discord community: https://discord.gg/WH7DrTHRXK

• Follow Junfan Labs: https://x.com/junfanlabs · https://junfanlabs.com

• Follow Abaka AI: https://x.com/Abaka_ai · https://www.linkedin.com/company/abaka-ai

• Follow Saturday Robotics: https://x.com/saturdayrobotic · YouTube https://www.youtube.com/@saturdayrobotic

• Subscribe to the Luma calendar: https://luma.com/saturdayrobotic