Symmetry-Aware Surrogate-Assisted Testing of UAVs
Symmetry-Aware Surrogate-Assisted Testing of UAVs
This research introduces a symmetry-aware surrogate-assisted testing method for UAV autopilots that exploits geometric symmetries to transform obstacle configurations into a shared canonical coordinate frame.
The study demonstrates a symmetry-aware surrogate model that improves critical-test-case hit rates for UAV autopilots in simulation but notes limitations when transferring predictions across missions with differing flight-path characteristics.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- Environment
- SimulationOpen source S3
Reported numbers
simulations
>2,500
View original evidence
-path characteristics, highlighting the boundaries of geometric canonicalization. To support reproducibility and future research, we provide the trained surrogate model and a dataset comprising more than 2,500 high-fidelity simulations.
Open source S5
- paper: not_reported
- code: not_reported
- dataset: Reported pending Open source S5
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
cing the amount of high-fidelity simulation data required for training. The surrogate model is trained to predict the minimum clearance distance between a UAV and surrounding obstacles and is used to guide test generation towards safety-critical scenarios. We evaluate this approach using the Aerialist UAV test bench and benchmark its performance against techniques from the 2024 to 2026 UAV Testing Competitions. Resul
Open source S3
-path characteristics, highlighting the boundaries of geometric canonicalization. To support reproducibility and future research, we provide the trained surrogate model and a dataset comprising more than 2,500 high-fidelity simulations.
Open source S5
ts show that the approach achieves an average critical-test-case hit rate of 42% across the evaluated missions, compared with approximately 10% for the constrained-random baseline used to generate the training data. This approach also achieves high input diversity and competitive benchmark performance. However, our findings reveal limitations in transferring surrogate predictions across missions with differing flight
Open source S4
Source：arXiv Robotics — research abstracts · arxiv.org