The National Industry PhD Program is an Australian Government initiative to enhance workforce mobility among graduate researchers, and to promote knowledge transfer between academia and industries across all areas. PhD candidates under this program are connected with academic supervisors and industry-based researchers, to co-design innovative, applied research projects. Through their doctoral candidature, students will experience research in both university and industry settings, and undertake specialised training in research translation and commercialisation.

The University of Adelaide and Lockheed Martin Australia invite applications for a project under this program, exploring the development of Physics Informed Neural Networks (PINNs) for efficient signal modelling in areas such as weather prediction, signal tracking, fluid dynamics, and space exploration.

Advancing Signal Modelling with Physics-Informed Neural Networks

This project aims to develop Physics Informed Neural Networks (PINNs) for efficient signal modelling in areas such as weather prediction, signal tracking, fluid dynamics, and space exploration. By incorporating physical laws into neural networks, we seek to improve accuracy, reduce data requirements, and lower costs for large-scale modelling tasks. PINNs enhance predictive capabilities and efficiency by combining data-driven methods with physical principles. Unlike traditional finite element methods, which demand extensive data and are costly, PINNs embed governing physical laws directly into the learning process. This allows effective management of limited and noisy data, resulting in faster and more cost-efficient signal modelling. However, embedding physical constraints within the network architecture remains a significant challenge. Current research often integrates these constraints into the training objective, complicating model training. This project aims to leverage advancements in computer vision, particularly in implicit neural representations, to embed priors in neural models through specific activation functions.

This project will be undertaken in collaboration with Dr Hemanth Saratchandran and Prof Simon Lucey of the Australian Institute for Machine Learning, and Advanced Systems & Technologies, Lockheed Martin Australia. The student will co-located at the Australian Institute for Machine Learning at the University of Adelaide and at Lockheed Martin offices located in Adelaide and Melbourne.

 Expected outcomes

The Finite Element Method (FEM) is the current dominant approach for modelling real-world signals but requires substantial, uniformly distributed data. Real-world data is often sparse, non-uniform, and noisy, making FEM challenging. This project aims to develop Physics Informed Neural Networks (PINNs) to address these issues by incorporating physical priors into neural networks. PINNs can model real-world signals with sparse, non-uniform, and noisy data. A key question is determining the optimal method for integrating physical priors into neural networks. This project will explore recent advancements in implicit neural representations, which have demonstrated effective neural network activations for computer vision. Our goal is to design new PINN architectures with specific activation functions tailored to the problem and physical phenomena, providing a robust method for incorporating physical priors. These designs will also reduce parameters needed for signal modelling, enabling more cost-efficient training algorithms.

Program overview

The successful candidate will receive:

• Admission to a PhD program at the University of Adelaide;
• A four-year scholarship package totalling $52,268 per annum (2025 rate, indexed annually), and a tuition fee waiver; 
• Supervision from research specialists at the University of Adelaide and Lockheed Martin; and
• A specialised training program in research translation and communication.

Upon successful examination of their thesis, the student will be awarded a PhD from the University of Adelaide.

 Eligibility Requirements 

This opportunity is open to Australian citizens who can meet the requirements for PhD admission at the University of Adelaide, and who can demonstrate suitable experience in Mathematics, Physics and/or Computer Science (through a high-quality Honours or Masters degree). The successful candidate must be able to enrol as a full-time PhD student at the University in the year of the offer. They must remain based in Adelaide, South Australia for the duration of the award.

Students that have previously completed a PhD program are, unfortunately, ineligible for the National Industry PhD Program.

The ideal candidate will also possess the following skills and qualifications:

• Undergraduate degree (Bachelor’s) in mathematics, physics, computer science, or a closely related field.
• A relevant Master’s degree (e.g., MSc, MPhil, or equivalent)
• Strong academic track record, with exceptional grades in advanced mathematics, theoretical physics, or computer science courses.
• Strong understanding of linear algebra, calculus, differential equations, and numerical methods. 
• Advanced programming skills in languages such as Python, C++, MATLAB, or R.
• Strong academic curiosity and enthusiasm for the chosen research area.

Application Process 

To apply, please email the following documents to principal supervisor Dr Hemanth Saratchandran (Hemanth.saratchandran@adelaide.edu.au) and Jessica Cortazzo, Manager, Projects and Strategic Partnerships, (jessica.cortazzo@adelaide.edu.au) with the subject line ‘National Industry PhD Program Application’:

• CV
• Cover Letter (of not more than 2 pages) outlining your interest in the PhD project and describing how your background and research area align with the project
• Degree certificates and relevant academic transcripts, with translations of non-English documentation

The due date for applications is 30 June 2025. Please note that applications will be shortlisted on a rolling basis, and the scholarship advertisement may be withdrawn early if a suitable candidate is identified.

Enquiries

Dr Hemanth Saratchandran

Email: Hemanth.saratchandran@adelaide.edu.au 

General Enquiries:

Email: hdrindustryenq@adelaide.edu.au