Towards more Realistic Logic-based Robot Controllers in the GOLOG Framework.

Abstract

High-level robot control languages should not only be expressive enough for realistic domains but also support reasoning about actions, in particular, the projection of robot plans, which is useful for the robot when choosing among different courses of action as well as the designer of robot controllers, since projections allow for qualitative simulations. GOLOG, a language based on the situation calculus, was specifically proposed for this purpose. While it comes equipped with a powerful projection mechanism, however, it lacks expressiveness. In particular, it cannot deal with continuous change, event-driven behavior, and probabilistic effects of actions, all of which are important in the domain of mobile robotics. In this paper, we show how these issues can be dealt with in the GOLOG framework by proposing appropriate extensions of the language. Introduction In the last five years, substantial progress has been made in building mobile robots which can navigate safely ...