Project Supervisor
Additional Supervision
Location
Melbourne Burwood Campus
Research topic
In modern agriculture, the fragmented nature of sensor technologies and disparate data platforms pose challenges in obtaining a comprehensive understanding of farm dynamics. As agriculture becomes increasingly data-driven, Artificial Intelligence (AI) has emerged as powerful tools in predictive breeding and farm management. However, despite the proven advantages of AI technologies for management decisions in cropping systems, widespread adoption remains limited due to a lack of standardised and securely integrated farm data and the need for high-compute infrastructures to run these algorithms. Additionally, although AI-based models rank higher on innovation and accuracy, they lack transparency and interpretability, as there is limited physics-based knowledge captured in the models, creating additional adoption bottlenecks.
This candidature period includes a compulsory Industry Placement period of six months. The student is expected to continue to work predominantly on the degree related research/thesis during the placement period. The PhD candidate would be spending at least 6 months with the Agriculture Resources Sciences team at AgriBio (Centre for AgriBioscience, 5 Ring Road, La Trobe University, Bundoora VIC 3083, Australia) incorporating possible trips to the AVR SmartFarms as required.
Project aim
Traditional AI models excel at identifying patterns in large datasets but often fail to incorporate fundamental principles of biology, chemistry, and physics that govern crop growth and environmental interactions. This project aims to bridge that gap by incorporating deterministic, physics-based models/functional relationships - which describe natural processes involved in soil-plant-environment continuum - into machine learning algorithms. Ultimately, this project aims to use AI in conjunction with biophysical models to understand and improve the gene-to-landscape linkage and apply them in predictive breeding and farm decisions.
One of the principal areas is to use machine learning-based AI models on Artificial Intelligence of Things (AIoT) to identify amalgamated artificial swarm intelligence in the soil-gene-plant-environment continuum while incorporating underlying physics functionalities and deterministic modelling. By enabling the real-time fusion of diverse datasets (either existing or newly collected), this dynamic approach will provide up-to-date, context-aware insights for farm management decisions (tactical and strategies) and improve predictive breeding programs. By leveraging decentralised, collective AI models, the proposed swarm-based approach ensures a deeper understanding of the ecosystem interactions that drive plant performance, helping optimise both breeding choices and management strategies. The project will enable achieving more efficient and targeted breeding programs, ultimately contributing to improved food security and sustainability.
Important dates
Applications close 5pm, Friday 31 January 2025
Benefits
This scholarship is available over 3 years.
- Stipend of $41,650 per annum tax exempt
Eligibility criteria
To be eligible you must:
- be a domestic candidate. Domestic includes candidates with Australian Citizenship, Australian Permanent Residency or New Zealand Citizenship.
- meet Deakin's PhD entry requirements
- be enrolling full time and hold an honours degree (first class) or an equivalent standard master's degree with a substantial research component.
- not be in full time employment at time of commencement of scholarship;
- meet the requirements for CSIRO Student affiliate onboarding (e.g. satisfy National Police Check);
- not be subject to an obligation to a third party to provide that third party with rights to any Intellectual Property created in the course of their degree; and
Please refer to the research degree entry pathways page for further information.
Additional desirable criteria include:
- Strong interest in Artificial Intelligence, Machine Learning, and Biophysical Modelling.
- Familiarity with programming languages (e.g., Python, C++) and experience with sensor technologies and data processing is advantageous.
How to apply
Please email a CV and cover letter to Prof Manzur Murshed. The CV should highlight your skills, education, publications and relevant work experience. If you are successful you will then be invited to submit a formal application.
Contact us
For more information about this scholarship, please contact Prof Manzur Murshed or
Prof Manzur Murshed
Email Prof Manzur Murshed
+61 3 924 46858
Dr Arbind Agrahari Baniya
Email Dr Arbind Agrahari Baniya