Computable Feasible-Horizon Sets
Determine computable lower boundaries of the admissible-horizon set for broad nonlinear systems with actuator magnitude and rate limits, uncertainty, safety constraints, and a common initial set, and characterize the associated deadline-dependent viable initial-state set.
References
The open problem is to make the lower boundary of $\mathcal T_{\mathrm{adm}$ computable for broad nonlinear classes and to quantify the gaps $T_{\mathrm{plant}{\min}(\mathcal X_0)\le T_{\mathcal C}{\min}(\mathcal X_0)\le T_{\kappa,\sup}(\mathcal X_0)\le T{\mathrm{cert}_\kappa(\mathcal X_0), \qquad \kappa\in\mathcal C$, where $T_{\kappa,\sup}(\mathcal X_0)=\sup_{z_0\in\mathcal X_0}T_\kappa(z_0)$ and $T{\mathrm{cert}_\kappa$ is the certified bound, all four for the same initial set, accuracy, uncertainty, and admissible inputs.
Open is a joint feasibility theory for $\text{safety} + \text{transient envelope} + \text{terminal accuracy} + \text{deadline} + \text{input magnitude/rate}$ that distinguishes remaining safe, returning to a safe set, and reaching a target without leaving it, and that returns the induced change of deadline or accuracy as an output of the analysis rather than as an unspecified slack variable.
The open direction is a unified specification and comparison framework in which the probability level, noise intensity, certified initial set, and control authority are explicit.