Integrating Physical and Logical Feedback Control on a Differential Drive Robot

Abstract

This paper presents a two-layered control architecture for integrating logical and physical feedback control in dynamic environments with real-world imperfections. The higher layer uses game-based synthesis to determine strategic reactions to logical context switches, while the lower layer employs task-dependent feedback controllers to actuate these decisions. Unlike conventional methods, the proposed framework provides direct logical feedback to environment-induced context switches and incorporates dynamic barrier certificates to ensure real-time safety. We demonstrate the practical effectiveness of our approach through an end-to-end implementation on a Turtlebot4, validated in an office environment with dynamic context switches.