Introducing FoodforThought: Gamified Robot Data Labeling for Everyone

Today we are excited to announce a major milestone for Kindly Robotics: the launch of FoodforThought, our community-powered platform for labeling robot demonstration data. This is not just another annotation tool—it is a gamified experience where contributors earn XP, climb leaderboards, and help train the next generation of intelligent robots.

🎮 Gamified Data Labeling

We believe that contributing to robotics research should be rewarding. That is why we have built a complete gamification system into FoodforThought:

XP and Leveling: Every task you complete earns XP based on difficulty. Progress through ranks from Novice to Grandmaster.
Streak Bonuses: Label tasks on consecutive days to build your streak. Maintain a 7+ day streak and earn up to 2x XP multipliers.
Leaderboards: Compete with the community on weekly, monthly, and all-time leaderboards.
Achievements: Unlock special achievements like First Label, Streak Master, Accuracy Expert, and more.

🤖 Real Robot Demonstration Videos

Unlike other labeling platforms that use synthetic or staged data, FoodforThought uses real robot demonstration videos from world-class research datasets like Bridge V2 and Open X-Embodiment. When you label a video on FoodforThought, you are working with the same data that top robotics labs at Berkeley, Stanford, and Google use to train their robots.

🔄 Automated Weekly Ingestion

Our data pipeline runs automatically every week, pulling fresh robot demonstration episodes from research datasets and converting them into labeling tasks. This means there is always fresh content to label!

🎯 Multiple Labeling Types

FoodforThought supports several types of annotation tasks: Action Segmentation (10-30 XP), Bounding Boxes (15-25 XP), Q&A Annotation (10-20 XP), and Polygon Segmentation (20-50 XP).

📊 Skill Library and Data Export

All the labels you create are aggregated into our Skill Library—a growing collection of labeled robot demonstrations available for download in JSON, RLDS, and LeRobot HDF5 formats.

🚀 Get Started Today

Ready to contribute to the future of robotics? Start labeling today!