Empirical Analysis of Sim-and-Real Cotraining of Diffusion Policies for Planar Pushing from Pixels
This work provides incremental insights into sim-and-real cotraining for robotics, specifically for planar pushing tasks, helping inform simulation design and policy training.
The study investigated how combining simulated and real demonstration data improves imitation learning for planar pushing tasks, finding that simulated data boosts performance when real data is limited, with gains scaling up to a plateau and real data increasing the ceiling, and that reducing physical domain gaps is more impactful than visual fidelity.
Cotraining with demonstration data generated both in simulation and on real hardware has emerged as a promising recipe for scaling imitation learning in robotics. This work seeks to elucidate basic principles of this sim-and-real cotraining to inform simulation design, sim-and-real dataset creation, and policy training. Our experiments confirm that cotraining with simulated data can dramatically improve performance, especially when real data is limited. We show that these performance gains scale with additional simulated data up to a plateau; adding more real-world data increases this performance ceiling. The results also suggest that reducing physical domain gaps may be more impactful than visual fidelity for non-prehensile or contact-rich tasks. Perhaps surprisingly, we find that some visual gap can help cotraining -- binary probes reveal that high-performing policies must learn to distinguish simulated domains from real. We conclude by investigating this nuance and mechanisms that facilitate positive transfer between sim-and-real. Focusing narrowly on the canonical task of planar pushing from pixels allows us to be thorough in our study. In total, our experiments span 50+ real-world policies (evaluated on 1000+ trials) and 250 simulated policies (evaluated on 50,000+ trials). Videos and code can be found at https://sim-and-real-cotraining.github.io/.