Searching for extraterrestrial artefacts on the moon and in the solar system: detection strategies and techniques

Pinault, Lewis James (2024) Searching for extraterrestrial artefacts on the moon and in the solar system: detection strategies and techniques. PhD thesis, Birkbeck, University of London.

Abstract

Building on 1980s hypotheses developed by Alexey Arkhipov, this Thesis presents new Machine-Learning based strategies, methods and techniques for detecting extraterrestrial technological artefacts. The focus is on ways to find evidence of submicron to micron debris that could have travelled across interstellar space via natural forces to the Solar System. It also applies these techniques to the object detection of spacecraft hardware remnants >0.5m on the Moon: exploring the author’s hypothesis that such dust-sized artefacts could include pre-programmed material designed to construct exploratory probes from local resources, much as submicron- scale DNA carries instructions for one of the most complex constructions we know so far in the Galaxy, humans. Imminent robotic and human activities on the Moon and other planetary bodies would benefit from advanced in situ Computer Vision and Machine Learning capabilities to identify and quantify microparticle terrestrial contaminants, lunar regolith disturbances, the flux of interplanetary dust particles, possible interstellar dust, β-meteoroids, and secondary impact ejecta. The YOLO (You-Only-Look-Once- ExtraTerrestrial) algorithm fine-tunes Tiny-YOLO to specifically address these challenges as well. Designed for coreML model transference to mobile devices, the algorithm facilitates edge computing in space environment conditions. In collaboration with JAXA, training on images from the Tanpopo aerogel panels returned from the International Space Station, YOLO-ET demonstrates a 90% detection rate for surface contaminant microparticles, and demonstrates promising early results for detection of both microparticle contaminants on the Moon and for evaluating asteroid return samples. YOLO-ET demonstrates an 80% detection rate for Apollo lunar landing modules, correctly identifying a known Luna 16 as a landing module. YOLO-ET also detects two potential candidates for Luna 9, with confidence levels of 61% and 43%, a spacecraft whose exact location has thus far remained undetermined. The light computing resource demands of YOLO-ET suggest that is well suited to continuous video object detection over the Moon’s surface.

Metadata

Statistics

DownloadsShow export options

Export as XMLJSONCSV

Activity Overview

6 month trend

35Downloads

6 month trend

54Hits

Additional statistics are available via IRStats2.

Archive Staff Only (login required)

Edit/View Item