Compiled implementation performance for online planning

Determine whether an implementation of the DA-ASC/FP-SQP planner in a compiled language can achieve the cycle times required by an online motion planner.

Background

The reported implementation is a Python prototype with a median optimization time of 5.8 seconds per case and a 90th percentile of 26.9 seconds. Although the paper establishes linear per-iteration computational growth with the horizon length, it does not establish that this complexity translates into online-planning cycle times.

The unresolved question concerns the practical real-time performance of a compiled implementation, rather than the theoretical per-iteration complexity of the algorithm.

References

We claim only the linear per-iteration cost of Section~\ref{sec:e-cost}; whether an implementation in a compiled language reaches the cycle times of an online planner has not been tested.

— Dynamics-Aware Adaptive Corridors with Feasibility-Perturbed Trust-Region SQP for Certified Nonholonomic Motion Planning  (2610.05796 - Shi, 5 Oct 2026) in Section 6, Discussion and Limitations, subsection “Solve time”