Abstract. This paper considers a local navigation method for a mobile robot operating in extreme conditions, i.e., the absence of global navigation and centralized control means, as well as limited-capability sensors. Local navigation is based on the potential field method, and the motion programs developed with this method are transferred to the level of elementary actions. Complex behavior is formed by meta finite-state machines (FSMs) implementing a control hierarchy. The robot’s basic task is to move to a given area or region of space. Within this task, an elementary procedure––obstacle avoidance––is considered and implemented based on the potential field method as well. The results of experiments are provided for three different levels, namely, computations, simulations (the physical level), and real robots, with the same computational model and FSMs used in all experiments. As shown, the successful solution of the local navigation task is determined by the size of the robot’s field of view, but the latter can be naturally expanded due to the universality of control mechanisms.
Keywords: potential field method, finite-state machine control, meta finite-state machine, movement models, mobile robot, local navigation, obstacle avoidance, control system hierarchy.
Acknowledgments. This work was carried out within the state assignment of NRC Kurchatov Institute.
Karpov, V.E., Finite-State Machine Control and the Method of Potential Fields for Mobile Robot Navigation in an Environment with Obstacles. Control Sciences 3, 86–99 (2026).
Finite-State Machine Control and Potential Field Method for Mobile Robot Navigation in an Environment with Obstacles