Simultaneous Localization Mapping and Navigation Accuracy in Indoor Service Robots

Abstract

The widespread deployment of indoor service robots relies heavily on their ability to autonomously navigate complex, dynamic, and unstructured environments. Central to this autonomous capability is the Simultaneous Localization and Mapping framework, which provides the spatial awareness necessary for safe path planning and execution. While significant algorithmic advancements have been made in spatial mapping, a critical gap remains in understanding the precise quantitative relationship between mapping accuracy, driven by underlying sensor fidelity, and the ultimate navigation performance of the robotic platform. This paper addresses this gap by investigating the direct linkages between sensor-level validation, mapping degradation, and navigation execution errors. Through a comprehensive experimental methodology utilizing a custom-built differential drive service robot equipped with heterogeneous sensors, including light detection and ranging, visual cameras, and inertial measurement units, we systematically evaluate how sensor noise, calibration errors, and environmental challenges propagate through the spatial mapping pipeline into the navigation stack. The study utilizes highly controlled indoor environments combined with motion capture ground truth to isolate and quantify these error propagations. The findings reveal that even marginal degradation in mapping fidelity due to sensor limitations exponentially increases trajectory deviations during autonomous navigation, particularly in feature-deprived or highly dynamic spaces. By validating the sensor inputs and correlating them with navigation outcomes, this research provides a robust empirical framework for optimizing sensor selection and algorithmic tuning in indoor service robotics, ultimately enhancing their reliability and operational safety.