Enhanced Geodetic Ground-Truth Methodology and Experimental Validation for Mobile Robots: Insights from the VRI4WD UFPR-MAP Dataset

This article introduces an enhanced geodetic ground-truth methodology tailored for mobile robot localization and mapping in hybrid indoor and outdoor environments. The work details the development of a comprehensive data acquisition system integrated into the VRI4WD wheeled robotic platform, enabling synchronized capture of diverse sensor modalities. A rigorous geodetic ground-truth methodology is presented, leveraging precise topographic surveying with GNSS and total station measurements referenced to standardized geodetic systems, ensuring highly accurate and globally consistent positioning data. A survey and comparison of leading multi-sensor datasets commonly employed as benchmarks is conducted to contextualize the UFPR-MAP dataset's contributions. The Sensor Blend Sets (SBS) method is adopted as a principled approach for selecting and grouping complementary sensors for distinct robotic tasks within the dataset. Validation experiments demonstrate the sensor fusion using wheel encoder odometry and laser scanner data, employing the GMapping SLAM algorithm to produce consistent occupancy grid maps across varying trajectories and environmental conditions. This integrated framework advances the state of multi-sensor datasets by providing an accessible, geodetically referenced resource for benchmarking autonomous navigation algorithms in complex real-world scenarios.