Can Robotics Hackathon Change Your Life?

Hugging Face · Beginner ·📄 Research Papers Explained ·1mo ago

Key Takeaways

Explores the potential of Robotics Hackathon to launch careers and inspire new research in AI robotics

Full Transcript

One year ago, thousands of people around the world spent one weekend building robots they've never built before. This is what they made and where they are now. >> Hi, I'm Nicodeme from Hugging Face. Exactly one year ago, the Robot Worldwide Hackathon brought together thousands of builders across dozens of cities, [music] all using the same open source robotic stack, all racing the same [music] 48-hour clock. So, I sat down with the winners to ask two simple things: What did you build and what happened next? Turns [music] out for a lot of them, that weekend changed things. For many people, this was their very first robot. >> I always wanted to get into robotics, but it was difficult before the robot. Because the robot has a was already a ready-made stack to start doing everything. >> [music] >> A few months before I saw a YouTube podcast which was talking about our The Robot is Awesome. 30 [music] minutes after I I bought the robot on the in the next weekend I I built it. Little bot was my first hackathon. >> After that, I did nine other hackathons and I won six out of it. So, it gave me really good boost. >> The same library, the same clock, and widely different ideas. Here is what they built. Pranav and his team trained a model to fold t-shirts. >> We created a model, I believe it was the ACTC model, to fold t-shirts and I think it achieved around like a 80% success rate, which luckily worked when the judges came around. [music] >> This team turned multiple arms into a little factory that sorts and places LEGO to draw pixel art. >> We came up with the idea of playing with LEGO. We had also an idea that doing pixel art is quite fun, but then the rest of the story actually got built during the hackathon itself, like to to make sort of a factory where some robots are storing the cube of the [music] right bright color and some others are placing the cubes. >> Zakaria and Nico bolted four arms into a walking centipede, born of course from a joke. [music] >> So, it was like a big joke and then it turned into a thing the day of. >> Darko's team made a robot arm you could control with your phone, and later their work got integrated [music] into the robot library. >> At the end, this teleoperation device I got integrated [music] to the robot, so we are very happy about it. >> Masato's team taught an arm to hand a dish to a person [music] and take one back, all from a single small VLA policy and a text instruction. Juan tried reinforcement learning on real hardware for the first time, teaching an arm to push a cube into [music] a square, just hard enough, but not too hard. Arna's team used Gaussian splatting to make models more robust to small differences in the camera placement. >> From that experience and seeing how hard it could be getting your camera set up, we decided that we really want to play on that data augmentation layer, make it easier to to do something, especially in an environment as volatile as a hackathon co-working space. >> In Japan, this team built a robot that complete electrical circuit by placing the battery and clipping the alligator clip. >> Yeah, everyone was shouting in the room once it started to work. >> Lila spent the weekend solo building a phone [music] app to fully drive the Loqui mobile robot. >> Yeah, I didn't have any idea what I'm going to do during the hackathon, and I just have like 2 days to do something that [music] is fun with Loqui and just try to experiment on the way and change the ideas. >> [music] >> Team Loqui built a graphical user interface for Loqui robot that just recently was announced as the official GUI for Loqui robot, and almost all of them remember the same moment, the one where it finally worked. [music] >> [music] >> Yay! >> [music] [music] >> Yeah! >> [music] >> Okay, and here is the part I really wanted to know. One year later, what did that weekend lead to? >> Actually, after the hackathon, I dived even further into the robotic side of things. [music] I've recently been working on more of agentic orchestration with the robots. >> Have my own YouTube channel where I talk about some technical details about different robotic projects, robotic models, and [music] so on. >> The robot was my first hands-on experience with, you know, VNA's. It just clicked, you know, and I had a habit where I look at the Discord channel everyday. >> Because of winning that hackathon, I got couple of internship, but I learned a lot from there. And that actually turned into a full-time job now. That really helped me. >> Right now, we are actually also using the robots platform to integrate tactile sensing. And all that confidence was actually coming from the hackathon last year. Where I realized it was quite simple to use that library and and to access to all the community around, especially with the robot Discord. >> For me, before the robot, there was no specific formulas. And with the robot, I believe that this is something that just happened now. It's really easy to share data and to share data sets and also share models. And I believe it has been quite a game-changer for me at actually on my own research. >> I am still writing that self-publishing book regarding the robot. >> So, making this little guy, which is supposed to be like a little wallflower essentially. We call him Walnut. >> Since then, we have actually purchased a Unitree G1 robot as well. And a lot of our effort has gone there. So, very happy to see that Unitree also supports the Unitree G1. >> And now you are thinking, that's cool, but I never joined a hackathon. I am nervous and I have no idea what to do. No worries. Most of the participants felt the same way. [music] So, I asked for some advice. >> Just jump into it. If you are really interested, just having, you know, the motivation is enough. >> When I joined, I really did not knew how to use LeoBot at that point of time, but it was pretty straightforward. Don't be afraid, just go ahead and enjoy the hackathon. >> Don't think too much about it. That's a really great experience. >> My favorite things about hackathons is at the end of it, even if I didn't place, uh I always come out of it having learned exponentially more than were I to just sit in front of the computer and go through a course. >> Making the first step is the hardest, but it's so rewarding once you do it. >> I think there's only one way to get good at robotics and it's to actually work with robotics. There's a lot of fields where you can read about it, watch and implement examples, but robotics feels like those things that are still where it's still necessary to actually get your hands dirty, to try, to fail, to try again, uh and then hopefully at some point succeed, which was also a bit the story of our journey through the hackathon. >> And as you can see, 1 year later, almost everyone is still building. Huge thanks to everyone who shared their story. The robots, the data, the community, it's all open and it's all out there waiting for [music] you. So, whenever the next one comes around, maybe I will see you there. If you want to experiment with robots, check out the LeoBot, but be careful. Playing with robots usually cause a lot [music] of excitement. >> Wow! Yeah!

Original Description

One year ago thousands of people worked for 48h to build something cool with LeRobot library: https://huggingface.co/lerobot LeRobot Worldwide Hackathon was a global AI robotics challenge that took place June 14-15, 2025. Discover how one weekend of open-source building launched careers, inspired new research, and turned total beginners into robotics experts. From folding t-shirts to Lego sorting factories, see what is possible when the community gets their hands dirty with open source AI and robotics. 🏆 See all winners from the event: https://huggingface.co/spaces/LeRobot-worldwide-hackathon/all-winners 🌍 Learn more about the hackathon: https://huggingface.co/LeRobot-worldwide-hackathon #robotics #lerobot
Watch on YouTube ↗ (saves to browser)
Sign in to unlock AI tutor explanation · ⚡30

Related Reads

Up next
Welcome to the Next Temperamental Era
Charles Schwab
Watch →