Mars Geological Classification Through An Intelligent Laser Spectroscopy System

Author: James Gormley, Department of Mechanical Engineering

Mentor: Dr. Prasoon Diwakar, Department of Mechanical Engineering

The goal of this project is to develop an in-situ testing and analysis technology for Mars site characterization operations with the capabilities of identifying water-ice, mineralogy, and elemental composition of Mars samples. The proposed tool will use a hybrid system of Laser-Induced Breakdown Spectroscopy (LIBS) and Raman spectroscopy techniques to achieve the stated goal. In addition to the hybrid analytical system, machine learning will be applied to autonomously analyze the large data sets collected. To develop this machine learning application, publicly available data sets from the NASA ChemCam Mars rover mission will be used. ChemCam provides over 8 years of LIBS data, collected on the surface of Mars, and is valuable in understanding the different mineralogy of various planetary regoliths. This presentation will include details of hybrid laser-based analytical techniques along with intelligent machine learning based data analysis, characterization, and classification of Mars geological features.