Abstract
The present study investigates possible evidence for an underground cavity on the western flank of Elysium Mons, initially identified in the Mars Cave Catalog, which is interpreted as a potential subsurface lava tube skylight. Using a multi-sensing approach, we analyze a Potential Cave Candidate (PCC) that emerges as a promising target for future missions, offering insights into Martian subsurface processes, resource potential, and astrobiological prospects. In particular, CTX and HiRISE imagery from the Mars Reconnaissance Orbiter, captured under varying solar illumination angles, reveal an elliptical morphology with persistent shadowed regions and roof collapse features, indicating depth and potential subterranean connectivity. HiRISE imagery further identifies a possible cave entrance suitable for future robotic exploration. Thermal data from the Mars Odyssey's THEMIS instrument indicate nocturnal thermal anomalies, supporting subsurface continuity. Topographic profiles derived from the Mars Global Surveyor’s MOLA data provide precise elevation measurements, reinforcing the hypothesis of an underground cavity. Additionally, TES mineralogical analysis identifies K-feldspar, plagioclase, amphibole, sulfates, carbonates, sheet silicates, high-silicon glass, and surface dust, indicative of volcanic processes and past water activity. These findings inform the proposed location of a landing site and an optimized robotic exploration path, emphasizing the potential use of AI-driven robotic dogs. These advanced systems combine AI-driven mobility, LiDAR-based mapping, and autonomous decision-making to navigate challenging terrains, including crater walls and potential cave entrances. Integrating such technology into planetary in situ exploration could overcome the limitations of traditional rovers, enhancing in situ analysis and advancing the search for habitable environments and biosignatures on Mars. This research has critical implications for the first extraterrestrial cave exploration, China's first Mars cave mission, and the broader quest for liquid water and potential extraterrestrial life, representing a transformative milestone in planetary exploration.
Introduction
The Martian surface is a hostile environment characterized by extreme radiation, significant temperature fluctuations, and global dust storms. For astrobiologists, the "holy grail" lies subsurface, specifically within lava tubes—natural tunnels formed by ancient volcanic activity. These caves offer stable microenvironments and protection from the harsh surface elements. However, traditional wheeled rovers are ill-suited for the steep, rugged "skylights" (collapsed roofs) that serve as cave entrances. This paper addresses the dual challenge of identifying a viable cave target through orbital remote sensing and designing a robotic system capable of navigating these subterranean frontiers.
Methodology and Site Identification
Researchers employed a multi-sensing orbital approach to characterize a specific Potential Cave Candidate (PCC) on the western flank of Elysium Mons. This involved morphological analysis using high-resolution imagery (CTX and HiRISE) to identify elliptical shapes and persistent shadows indicating depth. Topographic mapping was conducted using MOLA data to create precise elevation profiles of the pit and surrounding terrain.
Thermal inertia analysis using THEMIS data was critical for identifying nocturnal thermal anomalies. Because caves act as thermal reservoirs, they stay warmer at night than the surrounding surface. Mineralogical fingerprinting via TES and GRS data further identified minerals like sulfates and carbonates, suggesting a history of volcanic and aqueous activity.
Key Insights: The Thermal and Mineralogical Signature
The selected PCC exhibited a persistent thermal excess, remaining approximately 8.8K warmer than the surrounding surface during the Martian night. This nocturnal anomaly is a classic signature of a deep, insulating void.
Furthermore, the detection of K-feldspar, sulfates, and sheet silicates near the site suggests a complex history of volcanism and water interaction, making it a high-value target for biosignature searches. The candidate is classified as an "Atypical Pit Crater" (APC Type I) with a depth of approximately 140 meters and an elliptical surface opening (~291m x 222m).
Proposed Mission Architecture: AI-Driven Robotic Dogs
To overcome the limitations of wheeled rovers, the study proposes a dual-quadruped mission architecture based on the Unitree B2 robotic platform. Unlike rovers, quadrupeds can handle slopes >45°, climb 40cm stairs, and jump over 1-meter ditches.
To solve the problem of signal loss and power depletion inside caves, the authors propose a "buddy system":
- Surface Robot: Remains at the surface to harness solar power and maintain orbital communication.
- Explorer Robot: Descends into the cave via a retractable tether that provides a constant power and data link.
The researchers identified a safe landing zone just 500 meters from the cave entrance, featuring a gentle slope (~1.1°) and sufficient thermal inertia to ensure a stable landing and easy traversal for the robotic team.
Limitations and Future Work
The primary limitation is that current quadrupedal platforms are terrestrial. Future work must focus on space-hardening these systems to survive the vacuum, extreme cold (-125°C), and high radiation of Mars. Additionally, autonomous SLAM (Simultaneous Localization and Mapping) in completely dark, featureless basaltic tubes remains a significant hurdle, requiring the integration of advanced LiDAR and stereo vision to navigate GPS-denied subterranean frontiers.
Summary updated