- Dr Petra Kuhnert (CSIRO)
As agriculture meets the digital age, we are faced with challenging decisions about when to sow, when and how much to fertilise and when to irrigate. With the aid of high performance computing, models that represent complex agricultural processes can be used to simulate a wide range of farming scenarios in space and through time. Coupled with other sources of information (e.g. measured data and expert information) the challenge becomes being able to quantify the uncertainties and interpret the outputs in such a way that can facilitate on-ground decision-making.
- How do we make spatio-temporal decisions from a collective suite of information (data, modelled output, expert information) in the face of uncertainty?
- How can we better communicate the uncertainties from agricultural models to better inform farm management?
- What new visualisations can be developed to help convey uncertainty and lead to better risk based decision-making and how can we assess the usability and decision-making ability of these visuals?
This PhD project will develop cutting-edge statistical machine-learning methods, as well as innovative visualisations and interactive experiences to facilitate spatio-temporal decision-making in agriculture.
With the recent advancements and interest in the communication of uncertainties to guide decision-making, this research is timely. It will complement the current suite of activities within Digiscape, a future science platform within CSIRO that is helping to bridge the gap between agricultural science and digital technology.
This CSIRO-QUT Digital Agriculture Scholarship is worth $32k annually, plus $5k per year for development, training, and/or travel.
This project will have direct input into the Digiscape platform, led by Dr Petra Kuhnert. Digiscape is bringing to bear:
- cutting-edge climate science
- new sources of locally and remotely sensed data
- informatics for agro-ecosystems
- rigorous analysis of uncertainties
- innovation in both the ICT and social dimensions of systems integration.
You will be integral in the development of visualisation methodologies for the interpretation of modelled output and uncertainties to facilitate decision-making.
A Digiscape “uncertainty toolbox” will house the new analytics and algorithms for improved prediction and forecasting in agricultural systems to assist in spatio-temporal decision-making.
Skills and experience
Applications are open to both domestic and international applicants.
Required skills and experience:
- A high quality Honours or Masters degree in a quantitative discipline such as Data Science, Mathematics, Statistics, or Computer Science.
- A strong mathematics and statistics background.
- Experience computing in a software environment such as R, Python and/or C++.
- Good oral and written communication skills.
- Highly motivated, proactive, curious, and enthusiastic about scientific research.
Desirable skills and experience:
- Knowledge of machine learning methods such as neural networks and random forests.
- Knowledge of agricultural science or related discipline.
For further information regarding admission and scholarship application process please contact:
Higher Degree Research Scholarships Officer
QUT, Faculty of Science and Engineering
Int. phone: +61 7 3138 4783
For further information or to discuss this research project, please contact Erin Peterson.