SRTSR0319 Combining Coarse Climate Data with Local Observations for High-Resolution Drought Forecasting

For International or Australian domestic applicants

Supervisors: Prof. Jiuyong Li; A/Prof. Jixue Liu and Dr Yun Chen

Drought poses a significant threat to agricultural production and farmers’ livelihoods. Existing drought forecasting systems are often too coarse in both spatial and temporal resolution to enable farmers to prepare for drought events. Furthermore, centralized forecasting systems typically cannot incorporate locally collected weather observations to calibrate predictions for specific locations. This limitation is particularly important in areas where local factors, including terrain and water catchments, create microclimates that diher substantially from those of surrounding regions.

Recent advances in AI agents oher new opportunities to develop localized drought forecasting models that integrate local weather observations with global and regional climate data. This PhD project will develop AI methods that combine coarse-resolution meteorological forecasts with local environmental data to produce high-resolution, location-specific drought forecasts enabling more informed decision-making for drought preparedness. The project will foster interdisciplinary collaboration between the university and CSIRO at the intersection of AI and environment science.

If you are keen on working on cutting-edge research within a dynamic research group, please apply via: Apply Now

Feel free to contact me with your CV, transcripts etc. via Jiuyong.Li@adelaide.edu.au.

SRTSR0347 Constellation-Scale Onboard AI for Early Wildfire Detection

For International or Australian domestic applicants

Supervisors: Dr Stefan Peters; Prof Lin Liu; Dr Sha Lu and Prof. Marta Yebra;

Wildfires cause severe loss of life, property and ecosystems. Emissions from major fires accelerate climate change, making rapid detection vital. Small satellites can now detect wildfires with onboard artificial intelligence, sending alerts rather than downloading imagery, as our team has shown on South Australia's Kanyini mission and Loft Orbital's YAM-6. Onboard models for new missions, however, train on limited labelled data and analyse each pass in isolation. This PhD addresses both limitations.

First, it adapts pretrained Earth-observation foundation models to fire detection, using an existing semi-automatically labelled Sentinel-2 dataset for Australian landscapes, then compresses them for onboard use. Second, it develops constellation-scale methods that fuse repeated passes across SmallSat constellations to confirm fire progression, suppress false alarms and prioritise urgent alerts, moving Adelaide University's SmartSat CRC funded research towards operational capability.

If you are keen on working on cutting-edge research within a dynamic research group, please apply via: Apply Now

Feel free to contact me with your CV, transcripts etc. via stefan.peters@adelaide.edu.au.

SRTSR0485 Causal Machine Learning for Personalised Marketing and Customer Decision-Making

For International or Australian domestic applicants

Supervisors: Prof Lin Liu; A/Prof Giang Trinh and Dr. Ziqi Xu

Effective personalised marketing requires understanding customer responses to interventions such as price promotions and loyalty programs. However, estimating their true impact is challenging because intervention effects vary across customers, outcomes such as loyalty may emerge only over long periods, and data are often limited or biased. For new products, historical data are scarce, while large-scale experiments can be costly.

This PhD project will develop novel causal machine learning methods for estimating heterogeneous causal effects in data-scarce and long-horizon settings. To address the challenges, the research will leverage both machine learning and traditional causal inference and investigate techniques such as causal representation learning and foundation models, uncertainty quantification and the integration of observational and experimental data.

The project will advance causal AI and personalised decision-making while helping organisations optimise marketing investments, improve customer acquisition and retention, and support innovation in Australia's retail and digital economy.

If you are keen on working on cutting-edge research within a dynamic research group, please apply via: Apply Now

Feel free to contact me with your CV, transcripts etc. via Lin.Liu@adelaide.edu.au.

SRTSR0330 Next Generation Causal Recommendation Systems for Personalised Medicine

For International or Australian domestic applicants

Supervisors: A/Prof Thuc Le ; Prof. Jiuyong Li; and Dr. Sindy Pinero;

Personalised treatment recommendations are crucial in medicine, aiming to tailor therapies based on individual characteristics. However, most existing computational methods rely heavily on association patterns, failing to capture the underlying causal effects of treatments. This often leads to biased predictions. Meanwhile, causal inference methods, while grounded in theory, often require large datasets limiting their practical use in clinical settings. To overcome these challenges, this project leverages causal reasoning and modern foundation AI models for personalised medicine. These pretrained models offer strong generalisation capabilities, even on small datasets, making them particularly well-suited for biomedical applications. We aim to answer counterfactual questions such as "what would have happened if a different treatment had been given?" The methods support clinical decision-making with models that capture the true impact of treatments, not just statistical associations.

If you are keen on working on cutting-edge research within a dynamic research group, please apply via: Apply Now

Feel free to contact me with your CV, transcripts etc. via Thuc.Le@adelaide.edu.au.