Squeezing Data from a Rock: Machine Learning for Martian Science

Abstract

Data analysis methods have scarcely kept pace with the rapid increase in Earth observations, spurring the development of novel algorithms, storage methods, and computational techniques. For scientists interested in Mars, the problem is always the same: there is simultaneously never enough of the right data and an overwhelming amount of data in total. Finding sufficient data needles in a haystack to test a hypothesis requires hours of manual data screening, and more needles and hay are added constantly. To date, the vast majority of Martian research has been focused on either one-off local/regional studies or on hugely time-consuming manual global studies. Machine learning in its numerous forms can be helpful for future such work. Machine learning has the potential to help map and classify a large variety of both features and properties on the surface of Mars and to aid in the planning and execution of future missions. Here, we outline the current extent of machine learning as applied to Mars, summarize why machine learning should be an important tool for planetary geomorphology in particular, and suggest numerous research avenues and funding priorities for future efforts. We conclude that: (1) moving toward methods that require less human input (i.e., self- or semi-supervised) is an important paradigm shift for Martian applications, (2) new robust methods using generative adversarial networks to generate synthetic high-resolution digital terrain models represent an exciting new avenue for Martian geomorphologists, (3) more effort and money must be directed toward developing standardized datasets and benchmark tests, and (4) the community needs a large-scale, generalized, and programmatically accessible geographic information system (GIS).

Thermal Imagery for Rover Soil Assessment Using a Multipurpose Environmental Chamber Under Simulated Mars Conditions

“…This is especially critical in view of the rapid development of machine learning methods, which demand large amounts of training measurements [33], [34], [35]. This problem has been addressed by creating synthetic Martian data using generative neural networks [36]. Particularly, synthetic IR Martian images for thermal inertia estimation were generated from a multi-modal combination of Earth data, including RGB imagery [32].…”

Architectural Framework and Feasibility of Internet of Things-Driven Mars Exploration via Satellite Constellations

“…This assessment examines the characteristics of the communication links between Earth, represented by NASA's Deep Space Network (DSN) [31], and a series of hypothetical Mars orbiters. The DSN consists of three ground stations in Goldstone, California, USA; Madrid, Spain; and Canberra, Australia.…”

“…Scientific applications: These are applications used by scientists or companies to analyze and study data collected by IoT nodes or other instruments on Mars, such as Java Mission-planning for Analysis and Remote Sensing (JMARS) for atmospheric data analysis [31]. Section 5 of this work details the main applications that can be developed on Mars with this technology, considering the lines of research currently underway or planned to be carried out in the future.…”

Recognition and Classification of Martian Chaos Terrains Using Imagery Machine Learning: A Global Distribution of Chaos Linked to Groundwater Circulation, Catastrophic Flooding, and Magmatism on Mars

“…Machine learning is increasingly used as an application in industry. It has the potential to reduce the time and effort needed for planetary surface studies when a great quantity of data is available, such as in the case of Mars (e.g., [28,[36][37][38]). Such studies using machine learning include automated landform detections on Mars and comparison with Earth's [25,28].…”