2024 RTP round - Modelling uncertainty using robust adversarial data on concrete corrosion progression.

Status: Closed

Applications open: 7/07/2023
Applications close: 25/08/2023

View printable version [.pdf]
About this scholarship
Description/Applicant information

Digitization has emerged as an important tool for achieving sustainability objectives in industry. A 2020 ARC Discovery Group (ARCDG) research highlighted how digitization is being used in industrial operational safety enhancement.
Developing and deploying integrated Internet of Things (IoT) supported real-time monitoring of concrete corrosion would lead to achieving sustainability objectives through proactive preventive maintenance of infrastructure. However, IoT-supported solutions, by themselves, cannot provide insight into the process of corrosion initiation and progression. In order to understand the gradual progression of corrosion and deterioration of infrastructure, IoT-enabled real-time data needs to be used for designing, testing and deploying digital twins. A major problem in designing reliable digital twins is the lack of longitudinal data available at each stage of the corrosion progression process. 
Generating robust and reliable adversarial data can help in modelling various stages of the corrosion progression process. In order to assess the relevance of adversarial data, uncertainty measurement is conventionally used. Popular measures of uncertainty include predictive entropy and mutual information. This project will focus on developing a hybrid uncertainty measurement approach for generating reliable adversarial data to model various stages of the corrosion progression process. 

This PhD scholarship is being requested to for developing a hybrid uncertainty measurement approach for generating robust, reliable and realistic adversarial data in order to model various stages of the corrosion progression process. 

1. Carry out theoretical research on creating adversarial data for process monitoring.
2. Investigate sources of uncertainty in adversarial data.
3. Develop an exhaustive dataset of existing data on concrete corrosion progression.
4. Use the existing data for generating robust adversarial data on corrosion progression.
5. Investigate the pros and cons of the predictive entropy and mutual information methods of measuring data uncertainty.
6. Develop a hybrid method of uncertainty measurement and use it on the generated data.
7. Map the real concrete progression data on the generated data for data validation. 

This RTP scholarship will allow progressing toward better understanding of the corrosion progression process. The adversarial data will help in modelling various stages of the corrosion process. Ultimately, this work will lead to effective monitoring, diagnosis and maintenance of critical infrastructure for achieving sustainability. The incumbent doctoral students will develop, test and validate adversarial data using the real-time data collected from images, videos, and chemical, temperature and humidity sensor for understanding the corrosion process. 

This project may provide an internship opportunity. 

Student type

• Future Students

Faculty

• Faculty of Science & Engineering
  • Engineering courses

Course type

• Higher Degree by Research

Citizenship

• Australian Citizen
• Australian Permanent Resident
• New Zealand Citizen
• Permanent Humanitarian Visa

Scholarship base

• Merit Based

Value

The annual scholarship package (stipend and tuition fees) is approx. $60,000 - $70,000 p.a.

Successful HDR applicants for admission will receive a 100% fee offset for up to 4 years, stipend scholarships at the 2023 RTP rate valued at $32,250 p.a. for up to a maximum of 3 years, with a possible 6-month completion scholarship. Applicants are determined via a competitive selection process and will be notified of the scholarship outcome in November 2023. 

For detailed information, visit: Research Training Program (RTP) Scholarships | Curtin University, Perth, Australia.

Scholarship Details
Maximum number awarded

1

Eligible courses

All applicable HDR courses

Eligibility criteria

1. An excellent Honours or Master’s degree in the field of Mechatronics, Computer Science, Electrical Engineering or Robotics
2. Strong knowledge of two or more of the following:  sensors, machine learning, digital signal processing
3. Programming proficiency in C, C++, Python under Windows and/or Linux environments
4. Passion for innovation and problem solving
5. Initiative and interest in implement research ideas
6. Good communication skills and fluency in oral and written English 

Application process

This project has identified a preferred candidate and is no longer available.  Please review remaining scholarships projects .

Enrolment Requirements

Eligible to enrol in a Higher Degree by Research Course at Curtin University by March 2024.

Recipients must complete their milestone 1 within 6 months of enrolment and remain enrolled on a full-time basis for the duration of the scholarship.

Enquiries

The Project lead has identified a preferred candidate and is no longer accepting applications. Please click here to review remaining scholarships projects.

Scholarships Email Alert
Sign up now