Control with Probabilistic Signal Temporal Logic
This work addresses the problem of complex task specification and control for autonomous systems like UAVs in surveillance, though it appears incremental as an extension of existing STL methods.
The paper tackles the challenge of controller synthesis for autonomous agents in uncertain environments by proposing a probabilistic extension to signal temporal logic (STL) for tasks over continuous belief spaces, and presents an efficient synthesis algorithm validated through UAV simulations.
Autonomous agents often operate in uncertain environments where their decisions are made based on beliefs over states of targets. We are interested in controller synthesis for complex tasks defined over belief spaces. Designing such controllers is challenging due to computational complexity and the lack of expressivity of existing specification languages. In this paper, we propose a probabilistic extension to signal temporal logic (STL) that expresses tasks over continuous belief spaces. We present an efficient synthesis algorithm to find a control input that maximises the probability of satisfying a given task. We validate our algorithm through simulations of an unmanned aerial vehicle deployed for surveillance and search missions.