Optimal Observation Mode Scheduling for Systems under Temporal Constraints

Abstract

Autonomous control systems use various sensors to decrease the amount of uncertainty under which they operate. While providing partial observation of the current state of the system, sensors require resources such as energy, time and communication. We consider discrete systems with non-deterministic transitions and multiple observation modes. The observation modes provide different information about the states of the system and are associated with non-negative costs. We consider two control problems. First, we aim to construct a control and observation mode switching strategy that guarantees satisfaction of a finite-time temporal property given as a formula of syntactically co-safe fragment of LTL (scLTL) and at the same time, minimizes the worst-case cost accumulated until the point of satisfaction. Second, the bounded version of the problem is considered, where the temporal property must be satisfied within given finite time bound. We present correct and optimal solutions to both problems and demonstrate their usability on a case study motivated by robotic applications.