A competition held by Newcrest Mining has resulted in an innovative solution that uses machine learning for the analysis of historical mining core trays.
Held as part of The Newcrest Crowd’s crowdsourcing platform, the Get 2 the Core competition was held for participants to provide an algorithmic solution for how to derive value from historical core tray imagery.
When it comes to modern core tray photography, the manual identification of cropping marks in core trays is relatively straightforward, but historical samples can be much more variable due to crumbling, offset (parabolic) rows, poor lighting and other factors.
Participants were tasked with finding an automated solution that took this variability into account and as a result reduce the necessary man hours required. The eventual winner from 18 teams (and recipient of $10,000) was Jill Adams of Australian entrant ‘Straight off the Couch’.
Members of mining tech company MICROMINE’s Perth office also participated in the event using its Mask R-CNN technology. The company expects to translate its results to its Geobank data management software.
MICROMINE’s Wojciech Slabik said, “We have been working with machine learning techniques to solve mining problems on our Pitram team and we immediately noticed that we could apply these methods also to the problem presented by Newcrest.”