[Project Completion Report] Every Crater a Tale: Machine Learning Reveals the Hidden History of the Moon
The Australian Space Data Analysis Facility is delighted to formally and successfully conclude its part in the “A Machine Counts Craters on the Moon: The Progress Towards a Complete Global Impact Crater Dataset” project – a joint endeavour between Curtin University’s Space Science and Technology Centre (SSTC), CSIRO, Pawsey Supercomputing Centre, and ASDAF. This project aimed to unravel the geological history of the Moon by detecting and analysing the millions of craters constellating its surface with a machine learning algorithm, the Crater Detection Algorithm (CDA),which runs on Pawsey supercomputers. To achieve this, the team had to first convert the CDA – which was originally developed by CSIRO and scientists at the Curtin Institute for Computation and the SSTC to analyse Martian images – to analyse lunar crater images. This adaption commenced in late 2021 as one of the first ASDAFsupported projects and reached statistical validation on Pawsey supercomputers by April 2022.
Read the full project completion report below.