Feasibility Study of Low-Cost GPS Waypoint Path Planning for Autonomous Robots in Indoor-Outdoor Transitional Environments

Abstract

This paper presents a feasibility study on a lowcost, multi-receiver GPS-based waypoint navigation system designed for Autonomous Mobile Robots (AMR) operating in indoor-outdoor transitional environments. Navigating such areas presents significant challenges due to signal degradation, multipath reflections, and partial sky blockages. To mitigate these issues, the system leverages a multi-receiver configuration utilizing three synchronized u-blox NEO-6M modules. We evaluated two data fusion techniques, namely the simple centroid method and a weighted centroid method, in which the weights were assigned inversely proportional to the Horizontal Dilution of Precision (HDOP) values. Experimental results within a mixed campus environment at Faculty of Electrical Engineering & Technology, Universiti Malaysia Perlis (UniMAP) demonstrate that the HDOP-weighted centroid method enhances positioning accuracy, reducing the average error to 14.44 meters compared to 15.79 meters for the simple centroid. While this error margin is substantial for precise micro-navigation, the generated waypoints successfully formed a continuous and logical path, validating the feasibility of employing low-cost, multi-receiver configurations for approximate macro-level path planning. This work highlights the necessity of integrating robust sensor fusion techniques for reliable autonomous navigation in challenging transitional environments.