Jamal Rostami, Colorado School of Mines
Brad Blair, Moonrise Mining, Inc.

"Automating Lunar Drilling Data Compression Using Artificial Intelligence and Advanced Computational Hardware"

ABSTRACT -- This talk will describe a project that aims to aid in the search for water and other resources on the Moon. That search for water will involve drilling into the lunar regolith (soil) in many different locations. Advanced computing techniques can assist the drilling in at least two ways:

processing data at the drilling site, thus reducing the amount of communications bandwidth needed
controlling the drill in real time, adapting to the various layers that the drill encounters.

A NASA Early Stage Innovation (ESI) study at the Colorado School of Mines trained an Artificial Intelligence (AI) using an instrumented drill and water-bearing simulated lunar regolith. The study used four types of simulated lunar regolith:

low-porosity aqueous-icy
high-porosity aqueous-icy
unfused granular icy
fused granular icy

The project developed a "Lunar Material Characterization while Drilling Algorithm", using drilling data. Blind tests show that the algorithm can detect the various layers that the drill encounters, identifying the forms of water ice, the degrees of porosity, and the positions of the boundaries between them. The algorithm can also identify the drilling state, detect auger choking, calculate drill torque, and (using data from separate laboratory tests) identify the strength of the material being drilled.

Next Week: "The Past, Present, and Future of Glass and Light", Jim Heaney, 301-286-9133

Engineering Colloquium home page: https://ecolloq.gsfc.nasa.gov

Jamal Rostami, Colorado School of Mines Brad Blair, Moonrise Mining, Inc.

"Automating Lunar Drilling Data Compression Using Artificial Intelligence and Advanced Computational Hardware"

Jamal Rostami, Colorado School of Mines
Brad Blair, Moonrise Mining, Inc.