PhD scholarships in developing machine learning and data mining techniques for fuel loss detection at service stations

Develop effective machine learning and data mining techniques for identifying fuel loss detection at service stations.

This project aims to develop effective techniques to identify the sources of fuel losses, such as leaks and calibration errors in underground storage tanks at service stations. Monitoring fuel losses at service stations is influenced by many external factors which can be difficult to predict. The project expects to use machine learning to develop the techniques and test them with live data at service stations. The expected outcomes are a set of tailor-made machine learning techniques for effective fuel loss detection and a software suite that can be easily incorporated into the normal operation of service stations. This should reduce the costs to the petroleum industry from wasteful leaks and the environmental damage caused by these leaks. The project aims to achieve the following specific objectives:

  • Separate out the sources of errors/variance in the fuel storage system at service stations, such as calibration errors, meter errors and volume changes due to temperature variation. The data collected by sensors deployed in fuel tanks is often very noisy and can have a significant impact on the leak detection model
  • Develop effective machine learning techniques to detect and identify unexpected changes in the fuel storage system under normal operation. These techniques should be more accurate, robust to noise, and capable of capturing the dynamics of the business operation at a service station
  • Perform fuel loss detection in real-time while minimising misclassification using a fast and efficient online learning technique. Fast responses are critical in helping operators reduce loss and environmental impact
  • Design a next-generation software suite incorporating machine learning techniques with better algorithmic transparency. This tool should be of practical value to the wider industry sector and regulatory authorities

This is a project funded by an ARC Linkage Grant (LP190100991) over three years from 2020 to 2022.

Two PhD scholarships (each equivalent to an APA scholarship) are available. The scholarship consists of a $31,260 (tax free) stipend per year for the duration of three years, plus a possible 6 month extension (depending on excellent progress).

Candidates with backgrounds in machine learning and data mining are encouraged to apply.

To be eligible for this scholarship you must:

  • Have first-class Honours or 2A Honours or equivalent or a Masters degree (with a minor thesis component) in a relevant discipline of computer science/information technology/data science
  • Be an Australian citizen, Australian permanent resident or an international student meeting the minimum English language requirements
  • Provide evidence of good oral and written communication skills
  • Demonstrate the ability to work independently as well as within a team
  • Meet RMIT’s entry requirements for the Doctor of Philosophy.

It will be highly desirable if you have a GPA of 3.5 or above, and solid experience in machine learning and data mining applications. 

Candidates should contact Professor Xiaodong Li at xiaodong.li@rmit.edu.au. Prospective candidates should provide CV, academic transcripts and a written expression of interest before lodging any application with SGR.

Applications are open now.

Applications will close when the positions are filled.

We are particularly interested in applicants with solid experience in developing machine learning and data mining solutions to real-world problems.

This scholarship will be governed by RMIT University's Research Scholarship Terms and Conditions.

Professor Xiaodong Li at xiaodong.li@rmit.edu.au

aboriginal flag
torres strait flag

Acknowledgement of Country

RMIT University acknowledges the people of the Woi wurrung and Boon wurrung language groups of the eastern Kulin Nation on whose unceded lands we conduct the business of the University. RMIT University respectfully acknowledges their Ancestors and Elders, past and present. RMIT also acknowledges the Traditional Custodians and their Ancestors of the lands and waters across Australia where we conduct our business - Artwork 'Luwaytini' by Mark Cleaver, Palawa.