Solving Rubik's Cube with a Robot Hand
This addresses the sim2real transfer challenge for robotics, enabling complex tasks like Rubik's Cube solving with a robot hand, though it is incremental in advancing existing domain randomization methods.
They tackled the problem of transferring models trained in simulation to real-world robots for complex manipulation tasks, achieving the ability to solve a Rubik's Cube with a humanoid robot hand through improved sim2real transfer.
We demonstrate that models trained only in simulation can be used to solve a manipulation problem of unprecedented complexity on a real robot. This is made possible by two key components: a novel algorithm, which we call automatic domain randomization (ADR) and a robot platform built for machine learning. ADR automatically generates a distribution over randomized environments of ever-increasing difficulty. Control policies and vision state estimators trained with ADR exhibit vastly improved sim2real transfer. For control policies, memory-augmented models trained on an ADR-generated distribution of environments show clear signs of emergent meta-learning at test time. The combination of ADR with our custom robot platform allows us to solve a Rubik's cube with a humanoid robot hand, which involves both control and state estimation problems. Videos summarizing our results are available: https://openai.com/blog/solving-rubiks-cube/