A Human-Inspired Method for Mobile Robot Navigation

A new method for real-time navigation of mobile robots in complex and mostly unstructured environment is presented. This novel human-inspired method (HIM) uses distance-based sensory data from a laser range finder for real-time navigation of a wheeled mobile robot in unknown and cluttered settings. The approach requires no prior knowledge from the environment and is easy to be implemented for real-time navigation of mobile robots. HIM endows the robot a human-like ability for reasoning about the situations to reach a predefined goal point while avoiding static and moving or unforeseen obstacles; this makes the proposed strategy efficient and effective. Results indicate that HIM is capable of creating smooth (no oscillations) paths for safely navigating the mobile robot, and coping with fluctuating and imprecise sensory data from uncertain environment. HIM specifies the best path ahead, according to the situation of encountered obstacles, preventing the robot to get trapped in deadlock and impassable conditions. This deadlock detection and avoidance is a significant ability of HIM. Also, this algorithm is designed to analyze the environment for detecting both negative and positive obstacles in off-road terrain. The simulation and experimental results of HIM is compared with a fuzzy logic based (FLB) approach.