Grand Challenge
Earth Rover
GC/01
Human gamer teams compete with AI teams to advance autonomous navigation.
summary
Human gamer teams compete with AI teams in an ongoing urban navigation challenge to advance Embodied AI. In this challenge, gamers and AIs will compete in two phases navigating “in-the-wild” GPS goal-oriented scenarios.
In the first phase, competitors will navigate in a seen environment driving across the Bay Area (from UC Berkeley campus to Stanford University campus). This will be repeated regularly until a winning AI model is finally able to beat human benchmarks on the same task.
In the second phase, similar navigation challenges will be conducted except that this will be done in unseen environments in various cities around the world.
Co-Organizers
Researchers from academia (NUS, NYU, GMU, Princeton, etc.) and Google DeepMind, FrodoBots Lab
Prize
Up to $1M
Up to $100k USD will be available for the AI model that wins the first round that successfully navigates the Bay Area, and $900k for AIs that win by navigating multiple unseen cities.
GC/02
Robotic Origami
Human origami experts compete with AI teams in an ongoing robotic origami challenge.
prize
Up to $1M
APPLICATIONS
Open Now
Co-Organizers
FrodoBots Lab, Nippon Origami Association, Tesollo, Daxo Robotics, Orca Hand
GC/03
Robotic IKEA Furniture Assembly
Human furniture assemblers compete with AI teams in an ongoing furniture assembly challenge.
prize
Up to $1M
APPLICATIONS
Open Now
Co-Organizers
FrodoBots Lab
