BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//hackathonmaxxing//EN
CALSCALE:GREGORIAN
X-WR-CALNAME:Hackathons in London
BEGIN:VEVENT
UID:luma:evt-jQGx9vYnoPCoW1m@hackathonmaxxing
DTSTAMP:20261122T174714Z
DTSTART:20261128T090000Z
DTEND:20261129T190000Z
SUMMARY:Reality Check - Robotics Hackathon
URL:https://lu.ma/aiengine-r88j?utm_source=hackathonmaxxing&utm_medium=referral&utm_campaign=london
LOCATION:London\, United Kingdom
DESCRIPTION:Reality Check\, a weekend of robot learning\, from simulation to real robot arms.\n\nMore details: https://www.aiengine.build/reality-check-hack\n\n🎯The Challenge\n\nRobots that can reliably manipulate objects in the real world are one of the hardest open problems in AI. Models that look great in training often fall apart when the scene changes\, or when they meet real hardware for the first time.\n\nThe question we’re putting to you: how far can you push a small pre-trained model on real manipulation tasks in one weekend?\n\nAnd for a reality check\, guest models will take one shot at the same challenge at the finals\, so you’ll see how your fine-tuned model stacks up.\n\nThe challenge is built by active robotics researchers\, the winners of our February hack\, so expect it to be hard.\n\n🔨What to Build\n\nYour submission is a fine-tuned or post-trained manipulation model\, tested across multiple manipulation tasks on one robot embodiment.\n\nYou’ll start from pre-trained models and a small tele-op dataset for the robot. Make it go further however you like\, including generating synthetic data for training.\n\n⚙️How It Works\n\nTeams of up to 4\, building from Saturday morning to Sunday afternoon.\n\nEarly rounds are scored on a live leaderboard in a shared simulation environment\n\nEvaluation runs offline on local hardware with no internet access\, so all models must run locally\n\nOvernight training runs on cloud GPUs from Saturday into Sunday\n\nThe top teams go to the final and deploy on real robot arms\n\nExpect speakers\, robot demonstrations and technical briefings. Prizes\, bounties and partners to be announced.\n\nBlack Forest Labs is the frontier AI research lab for visual intelligence. builds foundation models for image generation\, including the FLUX family\, which has seen rapid adoption across open-source and enterprise platforms.Cognition is the applied AI lab behind Devin\, an autonomous software engineer. Engineering teams use Devin to take on coding tasks end to end in their own codebase. Modal i https://lu.ma/aiengine-r88j?utm_source=hackathonmaxxing&utm_medium=referral&utm_campaign=london
END:VEVENT
END:VCALENDAR
