Systematic robustness for layered control architectures

Develop a systematic methodology to incorporate model uncertainty and process noise into the synthesis and analysis of layered control architectures, providing robustness guarantees across layers within the optimal-control-decomposition framework used to derive decision-making, trajectory-planning, and real-time feedback subproblems.

Background

In setting up the overall synthesis problem, the paper adopts nominal dynamics without uncertainty to simplify exposition, while acknowledging that real systems invariably face modeling errors and process noise.

The authors discuss promising directions such as Lyapunov-based tracking certificates and assume–guarantee contracts but emphasize that a principled approach starting from a robust formulation and propagating guarantees through the architectural decomposition is missing.