Seminar: Adrian Ball, 5th Nov, 3p

When: Thurs 5th of Nov, 3p AEST

Where: This seminar will be presented online, RSVP here.

Speaker: Dr Adrian Ball

Title: Boundary Fusion: An Ore Body Modelling Technique

Abstract: Ore body modelling is focused on the estimation of ores and their transition boundaries. Typically, exploration holes are used to acquire ore data. Due to the cost of drilling, typically this data set is dense in the vertical dimension, and sparse in the horizontal plane. Accurate modelling from such a sparse data set is challenging, therefore, probabilistic models are appealing as they inherently have an uncertainty measure. In this seminar, I introduce a boundary fusion model, which generates global boundary transition estimates from a collection of local sub-region estimates. A few challenges of the application of a boundary modelling process are also addressed, with the objective of providing a flavour of the work being undertaken.

Bio: Adrian completed both his Masters and PhD at ACFR, focusing on machine learning and statistical modelling in the Human-Robot Interaction. Since then, he joined RTCMA as a post-doctoral researcher, working on the probabilistic modelling of ore boundaries, with a brief stint as an external contractor designing and building model architecture for the automatic generation of robot gestures. Since late 2019, Adrian returned to RTCMA and has been concentrating on the application of the boundary fusion modelling process to bauxite deposits.

Contacts

Sydney Institute for Robotics and Intelligent Systems
info@acfr.usyd.edu.au