Field Report on Ground Penetrating Radar for Localization at the Mars Desert Research Station

2025-04-21

AI Summary

This study addresses the insufficient robustness of pose estimation for Mars rovers operating on rugged terrain. For the first time, it systematically validates the feasibility of ground-penetrating radar (GPR) as an autonomous localization modality within a Mars analog environment (MDRS). Leveraging over 50 real-world trajectory datasets collected across two field campaigns, we propose a trajectory-level GPR signal processing and localization error analysis framework, enabling rigorous evaluation of positioning accuracy and robustness under Mars-like topographic conditions. Results demonstrate that GPR achieves sub-meter pose estimation accuracy in complex, unstructured environments; key limiting factors identified include subsurface dielectric heterogeneity, surface roughness, and antenna ground clearance—critical deployment constraints for operational use. This work establishes a novel application paradigm for GPR, extending its role from geological surveying to navigation-grade localization, and provides a GNSS- and vision-feature-independent autonomous navigation pathway for deep-space exploration.

Technology Category

• Intelligent Robots: Localization, Mapping, and Navigation
• Natural Language Processing: Language Grounding & Multi-modal NLP
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Systems and Infrastructure for Web, Mobile and WoT: Web performance, measurement, and characterization
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

In this field report, we detail the lessons learned from our field expedition to collect Ground Penetrating Radar (GPR) data in a Mars analog environment for the purpose of validating GPR localization techniques in rugged environments. Planetary rovers are already equipped with GPR for geologic subsurface characterization. GPR has been successfully used to localize vehicles on Earth, but it has not yet been explored as another modality for localization on a planetary rover. Leveraging GPR for localization can aid in efficient and robust rover pose estimation. In order to demonstrate localizing GPR in a Mars analog environment, we collected over 50 individual survey trajectories during a two-week period at the Mars Desert Research Station (MDRS). In this report, we discuss our methodology, lessons learned, and opportunities for future work.

Problem

• Validate GPR localization in Mars analog environments
• Explore GPR as rover localization modality
• Improve rover pose estimation using GPR data

Innovation

• Using GPR for rover localization on Mars
• Validating GPR in Mars analog environments
• Collecting 50+ survey trajectories at MDRS