Sim-to-real reliability of learning-based methods

Determine whether learning-based methods developed and evaluated in simulation reliably transfer to real-world robotic systems, and characterize the conditions that affect successful sim-to-real transfer.

Background

Simulation enables scalable evaluation and training, but domain gaps—especially in sensing and contact dynamics—can undermine real-world performance.

The authors note persistent uncertainty about whether methods validated in simulation will work in the real world, underscoring the need to understand and mitigate sim-to-real gaps.

References

While many learning-based methods are developed and evaluated in simulation, it is unclear whether they would work in the real world.

— From Machine Learning to Robotics: Challenges and Opportunities for Embodied Intelligence  (2110.15245 - Roy et al., 2021) in Section 6.2 (Assessing Robot Learning: Performance Evaluation)

Validating GS-VLA on a real robot requires both an external metric-depth source (per the previous point) and the engineering effort to instrument a physical rig, neither of which we were able to put in place within the resource and personnel constraints of this project. We therefore report the simulator results as a strong but ultimately preliminary signal, and treat a real-robot replication as the natural follow-up.

— GS-VLA: Plug-and-Play Viewpoint Canonicalization for Frozen VLA Policies via Gaussian Splatting  (2608.19066 - Park et al., 19 Aug 2026) in Section 5, “Limitation,” subsection “Real-world evaluation”