Dynamic replanning for a mobile robot based on internal sensing

Abstract

Schema-based navigational techniques are used to introduce dynamic path replanning for a mobile robot, based on internal sensor information. This mode of operation, termed homeostatic control, forms an integral part of the autonomous robot architecture. A model that is analogous to the mammalian endocrine system serves as the basis for this mode. Simulation results verify the viability of this concept. It is found that dynamic replanning can be carried out using internal monitoring of the robot's state, and not solely based on environmental perception. The robot can be shown to be responsive to changes in fuel, temperature, and other conditions as it navigates through the world.<>