
The Analytics for the Australian Grains Industry (AAGI) initiative is a five-year strategic partnership to enhance the profitability and global competitiveness of the Australian grains sector through advanced analytics. This five-year initiative (2023-2027) is mainly funded by the Grains Research and Development Corporation (GRDC) with a $36 million investment and builds on the previous Statistics for the Australian Grains Industry 3 (SAGI3). Investment. The University of Adelaide, in collaboration with Curtin University and The University of Queensland, is leveraging machine learning, data fusion, and statistics to support grain growers in making data-driven decisions. With a total co-investment of $56 million from the three strategic partners, this project offers HDR students a unique opportunity to contribute to cutting-edge research that has a real-world impact on the agricultural sector.
The Analytics for the Australian Grain Industry (AAGI) Scholarship Program (AU node) is funded by the Division of Research and Innovation as part of the co-investment from the University of Adelaide. This program will support twelve full-time PhD students commencing studies from 2025 to 2027.
In 2025, we are recruiting three PhD students for the following Projects:
Project 1: Unifying on-farm data and crop models to enhance tactical crop decisions
Summary: Despite the increasing availability of on-farm data and advances in process-based crop models such as APSIM, their integration often remains limited. This project proposes to get more out of on-farm data streams and process models through their more formal mathematical integration, with the desired outcome being to increase the water and/or nitrogen use efficiency of Australian cropping systems. We propose to use a range of emerging data science approaches within the fields of uncertainty quantification, data assimilation and optimisation under uncertainty, complementing data-driven approaches such as physics-informed machine learning. We will start by focusing on informing nitrogen management decisions (i.e. how much and when to apply), since they remain the largest contributor to the gap in water-limited production potential. This PhD project will contribute to the broader AAGI project “Harnessing emerging data science to unlock crop model potential and achieve production frontiers”.
Project-specific prerequisites: Strong quantitative skills are essential. Candidates with Masters or Honours degrees in the following disciplines, or with equivalent research or work experience will be favourably considered: Data Science; Applied Mathematics; Agricultural or Environmental Engineering, Agricultural Economics, Management and Information Technology.
Number of scholarships: Two
Contact person: Dr Matthew Knowling
Project 2: Efficient construction and visualization of pangenomes for crops with large genomes
Summary: Pangenomes are highly relevant for grains RD&E pre-breeding research because they capture the full spectrum of genetic diversity within a species, surpassing the limitations of single-reference genomes. By integrating multiple genomes from different individuals or populations, pangenomes provide a more comprehensive understanding of gene presence/absence, structural variations, and evolutionary dynamics. Graph-based pangenomes further enhance this capability by representing genomic variations within a unified framework, allowing for more precise mapping and analysis of complex genetic regions. In addition to graph-based methods, state-of-the-art dynamic programming methods have shown superior accuracy and performance in detecting more continuous and accurate core-genome sequences. However, their potential merits in detecting accessory sequences have not been investigated or benchmarked against graph-based methods.
In this project we aim to develop novel dynamic programming computational methods for pangenome generation of diploid and polyploid crop species and benchmark them against other methods. In addition, we will also develop a practical web-based visualization tool to represent pan genome and structural variations.
Project-specific prerequisites: Strong Java programming skills are essential. Candidates with a Masters or Honours degrees in the following disciplines, or with equivalent research or work experience will be favourably considered: Computer and Data Science; Applied Mathematics and Statistics.
Number of scholarships: One
Contact person: Dr Mario Fruzangohar