PhD Candidates – Machine Learning and Computational Physics for Neutron Tomography Drill Core Imaging

PhD stipend of $34,000 pa for 3.5 years (tax-free)

Tuition fees waived.

Two fully-funded PhD scholarships are available in machine learning and computational physics to advance neutron tomography technology for drill core imaging. The big-picture goal is to develop a drill core scanner based on neutron tomography, a non-destructive technique that provides 3D visualization and characterization of the internal structures of materials, with a special focus on geological and mineral samples.

It is anticipated that one PhD student will predominantly focus on the machine learning algorithms and software aspects of the project, while the other will predominantly focus on computational physics, but this can be adjusted based on the successful applicants’ interests to achieve the overall project goals.

At a high level, the aims are to:

-         Develop and refine physics-based computational models to improve the accuracy and efficiency of neutron tomography scans of drill cores, and

-         Develop machine learning-based software pipelines to create 3D maps of mineral structures from neutron tomography beamline data.

These will aid the mining and resources sector in more efficient and accurate mineral analysis.

The broader context

This PhD project is part of a multidisciplinary collaboration between James Cook University (JCU), the Australian Nuclear Science and Technology Organisation (ANSTO), and three industry partners (Terra Search, Ravenswood Gold, Sunshine Metals) to develop new tools for the mining and resource sectors. There are two PhD positions and a Postdoctoral Research Fellow who will work closely with academic and industry stakeholders.

About the PhD(s)

The successful applicant will contribute to the development and refinement of software and computational tools for the interpretation and modelling of neutron tomography. Key topics to be addressed include:

-         Key Topic 1: Machine learning to interpret neutron tomography data and predict 3D mineral composition, using a combination of supervised and unsupervised learning.

-         Key Topic 2: Physics-based simulations (e.g. monte carlo and continuum models) to describe neutron transport in complex geological objects.

-         Key Topic 3: Machine learning techniques to accelerate simulations.

This research will be directly informed by experimental characterizations of drill core minerology which allows a unique opportunity for the verification of the computational models.

What we offer

The successful applicants will receive a tax-free PhD stipend for 3.5 years, operational funding to support the project, and travel funding for collaboration and to attend relevant conferences. The projects are embedded within the Resources Technology and Critical Minerals Trailblazer, providing a rewarding research environment with strong connections to industry.

Eligibility requirements

Applicants must meet JCU’s entry requirements for admission to a PhD . Award of the scholarship is conditional on the university accepting your enrolment. The successful applicant will be guided through the process of applying for admission.

Additionally, applicants must meet ANSTO security requirements to participate in on-site work at the Dingo neutron imaging facility.

The ideal candidates will have a background in some combination of machine learning , programming, and/or computational physics. Applicants from the following disciplines are encouraged to apply:

• Physics
• Engineering
• Computer science
• Mathematics

This position is based at James Cook University in Townsville Australia, with an immediate start.

How to apply

To express your interest in this scholarship and PhD research opportunity, please prepare the following:

• A brief CV, including qualifications, academic achievements, list of publications (if applicable), work history, and references.
• A cover letter describing your research interests, why you would like to undertake a PhD, and what attracted you to this PhD project in particular.
• A copy of your academic transcript(s).

Please submit your application via email to [email protected] and [email protected]

The scholarship may be filled before the end of the advertising period, so you are encouraged to apply as soon as possible.